Acknowledgements
‘Acti iucundi labores’ (Cicero), a statement which means ‘Completed works are pleasant’ when translated to English might be signed by every PhD candidate with regard to his or her finished PhD thesis.
The elaboration and completion of this present PhD thesis was possible due the continuous discussion, untiring advice and supervision of Prof. Kulozik. As the director of the Institute for Food and Bioprocess Engineering of the ZIEL-Institute for Food & Health of the Technical University of Munich, it was him who gave me the opportunity to do research under distinguished conditions, excellent facilities, and all academic freedom possible. He gave me the opportunity to delve into the research field of dairy science and food chemistry. Using this profound background knowledge of dairy chemistry and analysis, it allowed me to unravel a small piece of the unknown. I got a deeper understanding of some correlations and causal relationships and thereby, I could figure out technological approaches to the long standing issue of the heat-induced changes of casein micelles in milk and hand it over to the scientific literature and dairy practice.
Prof. Jörg Hinrichs is thanked for his work on the heat treatment and storage stability of concentrated milk giving direction to my work. In addition, his critical review and examination of this thesis is acknowledged. Prof. Polifke is appreciated for the critical examination of this thesis. Prof. Heiko Briesen is thanked for taking over the Chair of the Examination Committee.
The shaping and implementation of my ideas, suggestions, and plans were enabled by the ongoing support of the members of the institute to the extent presented. Many a time, from the very beginning to my last days at the institute, the team of the institute’s own workshop, Christian Ederer, Erich Schneider, and Franz Fraunhofer supported me with their knowledge, expertise, and skills for the construction of parts, assemblies, and installations. The elaboration and continuous improvements of the analytical facilities and methods would not have been possible without the untiring efforts of the staff of the ‘HPLC lab’ and the ‘ELEMENTAR lab’ Heidi Wohlschläger, Annette Brümmer-Rolf, Claudia Hengst, Brigitte Härter, and Ilona Hager. The organisational talent and assistance of Mirjana Stulac enabled the large number of pilot plant trials necessary for this work. Last but not least, my students Eva Gintenreiter, Céline Lang, Sabrina Engelbrecht, Patrick Biebl, Fabian Ständer, Lena Weinzierl, Michaela Neuhauser, Alexander
Hainz, Andrea Machmüller, Laura-Elisa Helgert, Raffael Schuster, Maria Bauer, Barbara Böhmer, Margarita Isolina Valdiviezo Araujo, Rupert Wirnharter, Isabel Muthmann, Immanuele Pinto, Katja Gschaider, Felicitas Peraus, Pascal Birkle, Malou Warncke, and Marie-Therese Dörle are appreciated for their creative and scientific input, assistance, and research work performed at the institute during their seminar, Bachelor’s, and Master’s theses.
Special and very personal thanks go to the former and current members of the institute who contributed to a pleasant, creative, and relaxed work and leisure time atmosphere, especially my colleagues in the E.25 office over time, Katharina Pruss, Yu Zhuang, Ronald Gebhardt, Iris Schmitz-Schug, Elisabeth Eschlbeck, Sabine Günzkofer and Franziska Kurz.
The research project, from which this work originated, was supported by funds of the Federal Ministry of Food and Agriculture (BMEL) based on a decision of the Parliament of the Federal Republic of Germany via the Federal Office for Agriculture and Food (BLE) under the innovation support program (Grant number 313.06.01-281-74.005-11).
9.2
List of figures
Fig. 1-1: Scanning electron micrographs and transmission electron micrographs of casein micelles in milk and concentrated milk. .............................. 5
Fig. 1-2: Schematic representations of the casein micelles in milk ............................... 7
Fig. 1-3: Ribbon type models showing the 3D-structure of α-la and of β-lg using the software package UCSF Chimera ..................................................... 8
Fig. 1-4: Schematic representation of the proposed pathways of the formation of heat-induced whey protein-κ-casein complexes in heated skim milk. ...................................................................................................................... 10
Fig. 1-5: Overview of the early Maillard reaction in milk leading to the Amadori product lactulosyllysine ................................................................... 12
Fig. 1-6: Schematic representation of selected degradation routes of the degradation of lactose in heated milk. ............................................................ 12
Fig. 1-7: The partition of calcium in skim milk ............................................................. 15
Fig. 1-8: Iso-effect lines for selected microbial inactivation levels, chemical, and physical changes ......................................................................................... 22
Fig. 1-9: Different types of continuous high heating systems using saturated steam or hot water as a heat transfer medium. .............................................. 25
Fig. 1-10: Principle of operation of direct heating systems ........................................... 26
Fig. 1-11: Patented steam injector having three annular steam inlets and an adjustable mandrel to control the velocity of the product flow .................. 27
Fig. 1-12: Exemplary temperature-time profiles of direct (A) and indirect (B) UHT plants .......................................................................................................... 29
Fig. 1-13: Copper plated thermostatic paraffin bath used for the subjective heat stability test of Davies and White (1966) ......................................................... 31
Fig. 1-14: Aggregated protein and non-protein nitrogen plus proteose peptone nitrogen produced in ‘normal’ separated bulk milk and milk from an individual cow after heating at 135 °C for various periods ......................... 34
Fig. 1-15: Heat coagulation time (HCT) vs. pH profile for type A milk, type B milk, concentrated milk at 20% non-fat total solids, and serum protein-free casein micelle dispersions heated at 140 °C.............................. 37
Fig. 1-16: The relation between the logarithm of coagulation time and temperature of four different types of milk .................................................... 37
Fig. 1-17: Processing alternatives for the manufacture of evaporated milk or concentrated sterile milk in general using the UHT process for sterilisation .......................................................................................................... 39
Fig. 1-18: HCT vs. pH profile for concentrated milk at 20% non-fat total solids prepared from either preheated or non-preheated milk heated at 120 °C.................................................................................................................... 40
Fig. 3-1: Assembly used for determination of heat stability of unconcentrated and concentrated milk of different total solids content ................................ 51
Fig. 3-2: Heating curves of individual milk samples subjected to heat stability test ......................................................................................................................... 52
Fig. 3-3: Heat coagulation time as a function of the pH at 20 °C ................................ 53
Fig. 3-4: Heat coagulation temperature as a function of the pH at 20 °C .................. 54
Fig. 3-5: Heating curves of samples at various oil bath temperatures ....................... 55
Fig. 3-6: Correlation between milk sample temperature and time of coagulation of concentrated skim milk of different total solids content .............. 56
Fig. 3-7: Heat coagulation time correlated with sample temperature of individual skim milk samples over a 15 months period ..................................... 58
Fig. 4-1 Simplified P&I diagram of the pilot plant unit for direct steam injection heat treatment of concentrated skim milk ...................................... 66
Fig. 4-2: Temperature-time profiles of the direct steam injection UHT pilot plant ...................................................................................................................... 68
Fig. 4-3: Scanning electron micrographs of protein particles formed during direct steam injection heat treatment of concentrated skim milk at 145 °C.................................................................................................................... 71
Fig. 4-4: Scanning electron micrographs of supernatants of concentrated skim milk ....................................................................................................................... 72
Fig. 4-5: Logarithmic density distribution measured by laser light diffraction of concentrated skim milk heat treated by direct steam injection at different temperatures ....................................................................................... 73
Fig. 4-6: Logarithmic density distribution of the supernatant from preheated concentrated skim milk heat treated by direct steam injection at different temperatures ....................................................................................... 73
Fig. 4-7: Sedimentable aggregated protein of concentrated skim milk ..................... 76
Fig. 4-8: Iso-effect lines of the onset of coagulation of concentrated skim milk heated by direct steam injection plotted on a semi-logarithmic scale ........ 78
Fig. 5-1: Sample preparation procedure for separation of the different particle fractions and RP-HPLC analysis ...................................................................... 86
Fig. 5-2: Logarithmic density distributions of CSM heat treated at different temperatures ....................................................................................................... 90
Fig. 5-3: Average volume-based particle diameter d50,3 as a function of temperature, total solids, and holding time of DSI heat treatments.................. 92
Fig. 5-4: RP-HPLC chromatograms of supernatants (4,000 xg/10 min) of caseins and whey proteins in guanidine buffer ............................................. 93
Fig. 5-5: Relative amount of non-sedimentable caseins in CSM at 27% total solids centrifuged at 4,000xg for 10 min heated at different temperatures ...................................................................................................................... 94
Fig. 5-6: RP-HPLC chromatograms of ultracentrifugal supernatants of caseins and whey proteins in guanidine buffer ........................................................... 95
Fig. 5-7: Relative amount of non-sedimentable caseins in CSM at different total solids after centrifugation and ultracentrifugation .............................. 96
Fig. 5-8: Relative amount of non-sedimentable whey proteins in CSM at different total solids after centrifugation and ultracentrifugation .............. 98
Fig. 5-9: Relative amount of soluble β–casein and soluble κ-casein in the ultracentrifugal supernatant after heat treatment of concentrated skim milk of 27% total solids at different temperatures ............................... 99
Fig. 5-10: Schematic representation of the proposed model of heat-induced dissociation and coagulation of caseins and whey proteins in CSM ........ 100
Fig. 6-1: High performance ion exchange chromatograms of ultrafiltration permeate obtained at 50 °C and simulated milk ultrafiltrate analysed for cations and anions ...................................................................................... 108
Fig. 6-2: Ca2+-activity over pH for SMUF solutions in comparison to UF permeate ............................................................................................................ 119
Fig. 7-1: Relative amount of sedimentable protein over the logarithmic heating time depending on heating temperature and total solids fitted by the Weibullian model ....................................................................... 130
Fig. 7-2: Variation of the parameter a of the Weibullian model depending on total solids .......................................................................................................... 131
Fig. 7-3: Parameter βw of the Weibullian model obtained by non-linear regression of sedimentable protein depending on heating temperature and total solids content .............................................................
132
Fig. 7-4: Arrhenius-plot of the parameter αw of the Weibullian model obtained by non-linear regression of sedimentable protein depending on heating temperature of CSM ........................................................................... 133
Fig. 7-5: Simulation of the relative amount of non-sedimentable protein depending on CSM total solids at 116 °C and 27% CSM total solids depending on heating temperature over time on a logarithmic scale ..... 135
Fig. 7-6: Iso-effect lines for CSM for a critical level of 2% of sedimentable protein of total protein depending on heating temperature, heating time, and total solids content of bulk CSM ................................................... 136
Fig. 7-7: Relative amount of soluble κ-casein and αS1-casein depending on heating temperature at 27% CSM total solids over time ............................ 138
Fig. 7-8: Regular residuals of the Weibullian model fitted to the data of relative amounts of sedimentable protein of total protein and the normal probability plots of the residuals. .................................................................. 140
Fig. 8-1: Amount of whey proteins in the MF retentate of during concentration and DF of milk with simulated milk ultrafiltrate ............... 147
Fig. 8-2: Decrease in pH during heat treatment of MCC at different temperatures and the corresponding pH after dilution of MCC with SMUF buffer ...................................................................................................... 152
Fig. 8-3: Relative amount of dissociated κ-casein, αS2-casein, αS1-casein, and β-casein analysed in the supernatant of MCC ultracentrifuged at 48,800xg/26 min after heating at different temperatures ............................ 153
Fig. 8-4: Relative amount of dissociated κ-casein and αS1-casein in MCC analysed in the supernatant of MCC ultracentrifuged at 48,800xg/26 min after heating at a temperature of 116 °C at different pH ....................................................................................................................... 155
Fig. 8-5: Heat coagulation temperature and heat coagulation time of MCC depending on pH and addition of soluble calcium ..................................... 156
Fig. 8-6: Calcium activity in MCC depending on pH and addition of calcium as calcium chloride ........................................................................................... 157
Fig. 8-7: Relative amount of dissociated κ-casein and αS1-casein in MCC analysed in the ultracentrifugal supernatant of MCC after heating at a temperature of 116 °C and addition of soluble calcium .......................... 157
Fig. 8-8: Volume based average diameter d50,3 of casein micelles in heat treated MCC at different temperatures, different pH, diffent calcium levels and ionic strength .................................................................................. 159
Fig. 8-9: The maximum d50,3 of heat treated MCC depending on pH and soluble calcium level ........................................................................................ 160
Fig. 8-10: Relative amount of dissociated κ-casein and αS1-casein in MCC analysed in the ultracentrifugal supernatant and centrifugal supernatant after heating at temperature of 116 °C, pH adjustment to 6.35, and addition of 3 mM soluble calcium ................................................. 161
Fig. 9-1: Correlation between milk sample temperature and time of coagulation of concentrated skim milk of different total solids content ................................................................................................................ 167
Fig. 9-2: Iso-effect lines for CSM for a critical level of 2% of sedimentable protein of total protein depending on heating temperature, heating time, and total solids content of bulk CSM ................................................... 168
Fig. 9-3: Visually determined heat coagulation temperature and heat coagulation time of concentrated skim milk on lab scale depending on the total solids content ............................................................................... 175
Fig. 9-4: Visually determined heat coagulation temperature and heat coagulation time of concentrated milk containing various proportions of milk fat depending on the total solids content .............................. 176
Fig. 9-5: Visually determined heat coagulation temperature and heat coagulation time of concentrated milk containing various proportions of milk fat depending on volume corrected non-fat total solids ..... 177
Fig. 10-1: Calculation of the expected relative amount of sedimentable protein based on the kinetic data for direct steam injection heat treatments ................... 181
Fig. 10-2: Iso-effect lines for CSM for a critical level of 5% and 10% of sedimentable protein of total protein depending on heating temperature, heating time, and total solids content of bulk CSM............................................................. 183
Fig. 10-3: Calculated data for the factor depending on temperature and depending on total solids content of CSM ........................................................ 184
List of tables
Tab. 1–1: Composition of one kilogram of bovine milk .................................................. 3
Tab. 4–1: The pH values of skim milk and concentrated skim milk (CSM) before heat treatment ......................................................................................... 67
Tab. 4–2: Parameters of the Gompertz equation fitted to the data of sedimentable protein after DSI heat treatment of CSM of various total solids content.............................................................................................. 76
Tab. 4–3: Parameters of the Gompertz equation of different holding times at 27, 23% total solids, and at 27% totals solids of UHT preheated milk ........ 77
Tab. 6–1: Quantities of individual salts needed for the preparation of a SMUF simulating ultrafiltration permeate at 10 °C and 50 °C .............................. 109
Tab. 6–2: Ionic composition of ultrafiltration permeate obtained at different filtration temperatures ..................................................................................... 114
Tab. 6–3: Physical properties, dry matter and lactose content of ultrafiltration permeate obtained at different filtration temperatures .............................. 115
Tab. 6–4: Total ionic composition of different milk samples determined by precipitation of proteins and solubilisation of colloidal salts .................... 115
Tab. 6–5 Calculated amounts of individual ionic species of different SMUF solutions; comparison with the SMUF solutions of Jenness and Koops (1962) ...................................................................................................... 116
Tab. 6–6 Ionic composition of different SMUF solutions as determined by ion chromatography ............................................................................................... 117
Tab. 6–7: Physical properties of different SMUF solutions with and without addition of lactose ............................................................................................ 118
Tab. 7–1: Parameters and calculated parameters derived from linear regression of the Arrhenius-plot of the parameter αw in CSM at 27% total solids .... 134
Tab. 7–2: Parameters and calculated parameters derived from linear regression of the plot of the parameter αw depending on total solids content of CSM at a heating temperature of 116 °C ....................................................... 134
Tab. 8–1: Composition of the MCC powder .................................................................. 148
Tab. 8–2: Physical characteristics and protein composition of the reconstituted MCC.................................................................................................................... 150
Tab. 10–1: Comparison of the observed and calculated temperatures of the onset of coagulation ........................................................................................................ 182
Abbreviations
Latin symbols
A Membrane constant m2 s kg-1
A Constant 0.51
AMP Advanced Maillard products
BLE Federal Office for Agriculture and Food (Germany)
BMEL Federal Ministery of Food and Agriculture (Germany)
BSA Bovine serum albumine
c Concentration mol L-1, M
C5, C6 Molecule with five and six carbon atoms in the backbone, respectively
C0 Initial concentration (of protein) g L-1, %
C∞ Concentration of protein after infinite heating time %
CCP Colloidal calcium phosphate
CD Circular dichroism
CSM Concentrated skim milk
Ct Concentration at time t g L-1
CT,t Concentration of protein at a discrete temperature and holding time %
DD Degree of denaturation
DSI Direct steam injection
Dϑ Decimal reduction time at temperature ϑ s
EA Activation energy kJ mol-1
EDTA Ethylenediaminetetraacetic acid
ESL Extended shelf life
FAO Federal Agriculture Organisation
FMP Final Maillard product
fs Scaling factor for αW depending on CSM total solids -
fT
Scaling factor for αW depending on temperature -
FTIR Fourier transform infrared spectroscopy
gal Galactose
HCl Hydrochloric acid
HCT Heat coagulation time s, min
HPLC High performance liquid chromatography
I Ionic strength mol L-1
J Flux m3 m2 s
k Rate constant s-1
k Gompertz parameter °C-1
k0 Apparent rate constant at s-1
kapp Apparent rate constant s-1
kB Boltzmann constant 1.3806 · 10-23 J K-1
KCl Potassium chloride
KOH Potassium hydroxide
kref Rate constant at reference state s-1
kT Rate constant at temperature T s-1
LOD Limit of detection
LOQ Limit of quantification
MCC Micellar casein concentrate
MPC Milk protein concentrate
n Reaction order -
N0 Initial number of microorganisms -
NA Avogadro constant 6.0221 · 10-23 mol-1
NaOH Sodium hydroxide
NPN Non-protein nitrogen
Nres Number of resistant microorganisms -
Nt Number of microorganisms at time t -
P&I Pipe and instrumentation
PCS Photon correlation spectroscopy
PDB Protein data base
PEEK Polyether ether ketone
Pi Inorganic phosphate
PIu Upper limit of the (95%) probability interval
PPN Proteose peptone nitrogen
PS-DVB Polystyrene divinylbenzene
Q10 Q10-value K; °C
R Universal gas constant J mol-1 K-1
R1, R2 Organic residues
R2 Coefficient of determination -
RO Reverse osmosis
RP-HPLC Revesed-phase high performance liquid chromatography
s Total solids content of concentrated skim milk %
S(t) Survival ratio at time t -
Abbreviations
SMUF Simulated milk ultrafiltrate
SMUF-JK Simulated milk ultrafiltrate according to Jenness and Koops (1962)
SP Sedimentable protein %
SP,max Maximal amount of sedimentable protein %
sref Reference total solids content of CSM %
T Absolute temperature K
T Temperature of the Gompertz model °C
T* Reference temperature plus 10 K K
Tc Temperature of the inflection point (Gompertz model) °C
TFA Trifluoroacetic acid
thermoph. thermophilic
tR Reliable lifetime s
Tref Temperature at reference state K
tref Reference time s
TS0 Initial total solids content %
TSnon-fat Non-fat total solids content %
TSnon-fat, vc Volume corrected non-fat total solids content %
TST Total solids content at temperature T %
UF Ultrafiltration
UHT Ultra-high temperature
v Reaction velocity s-1 Mn
vc Volume corrected
VRF Volume reduction factor
w/v Weight per volume
z z-value K; °C
zi
Greek symbols
Valency of an ion -
Slope of the regression line of the Arrhenius plot for the characteristic time αW at temperature T K
α activity mol L-1
α-la α-Lactalbumin
αW Weibull parameter s-1
β-lg β-Lactoglobulin
βW Weibull parameter -
γi activity coefficient -
Δp Pressure difference Pa
ΔpDL Presure difference of a deposit layer Pa
Δπ Osmotic pressure difference Pa
ϑ temperature °C
ρCSM Density of concentrated skim milk kg m-3
ρfat Density of milk fat globules kg m-3
Φfat Volume fraction of milk fat globules -
Summary
The necessity to determine the heat stability of concentrated milk originates from the manufacture of evaporated milk and dates back to the late 19th century. It became an issue due to observed particle formation, gelation, and sediment formation in evaporated milk during sterilisation and subsequent storage. Certain batches of evaporated milk, sterilised in cans or glass, and later in HD-PE bottles, showed these instabilities during heat treatment and therefore became defective. Hence, there was an increasing economic interest in the basic understanding and control of the heat stability of the concentrated milk. Since then, dairy science deals with the topic of the heat stability of milk and especially concentrated milk concerning factors affecting the heat stability, methods to predict, and measures to improve it. However, the mechanism of coagulation, factors that cause coagulation and modelling approaches to predict the heat stability of concentrated milk for process optimisation were still lacking.
Individual milieu conditions that affect the heat stability of milk or milk concentrated by evaporation or reverse osmosis have been intensively investigated in the past. These factors comprise the initial pH before heating, the amount of soluble divalent cations, especially soluble calcium, the amount of reactive whey proteins, and the ionic strength in the milk serum. The heat stability of unconcentrated bulk milk at its natural pH is maximal in most cases when dilution is not considered. Attempts to compensate for changes that occur during concentration of milk by removal of water, i.e. changes in pH, ionic strength, soluble calcium, and protein content could not restore the heat stability of unconcentrated milk. Attempts made comprise the addition of bases, citrates, phosphates, and many other permitted and illegal additives for practical applications and research, respectively. Other membrane techniques using porous membranes such as nano-, ultra-, and microfiltration also lead to a reduction in the heat stability of the final concentrate as compared to the unconcentrated milk. However, the overall heat stability of the concentrate increases with increasing compositional similarity of the concentrate to unconcentrated milk. This observation seems plausible. Nevertheless, a prediction of the heat stability of these concentrates in terms of maximum temperature-time combinations for these concentrates under continuous heating conditions without coagulation was lacking.
Technological treatments such as treatments of milk before concentration and heat treatment can affect the heat stability of the final concentrate positively or negatively. Therefore, the combination of processing steps and the choice of processing parame-
ters is decisive to maintain or improve the heat stability of the concentrate. Homogenisation of milk or concentrated milk before heat treatment of the concentrate negatively affects the heat stability. Appropriate preheating of the milk before concentration increases the heat stability of the concentrate. A quantitative estimation of the heat stability expressed as a shift in critical temperature-time combinations had not been performed.
Therefore, the aim of this study was to establish a laboratory test method and a model able to describe the heat stability of concentrated skim milk of various total solids under continuous heating conditions. The limited heat stability of concentrated skim milk of various total solids content should be described quantitatively on lab- and pilot scale. The intention was to be able to design heating processes that maximise microbial inactivation in the concentrates by, at the same time, avoiding heatinduced coagulation of the concentrates. In addition, a description of selected heatinduced changes on casein micelles was also intended. Explanations for the decrease in heat stability by concentration of milk due to changes in milieu conditions should be found and attributed to single components. It was assumed that certain milieu conditions increase the rate of heat-induced changes on casein micelles reducing their colloidal stability and thereby induce coagulation.
The investigation of the heat stability on lab scale was performed using small amounts of concentrated milk filled into small glass tubes tightly screwed and immersed into an oil bath for heating until visual coagulation was observed. Heat coagulation time and temperature were recorded as many samples readily coagulated during heating-up times. The aim of the investigations was to determine the effect of pH, total solids content, preheat treatments, and milk fat on critical temperature-time combinations of the onset of coagulation on lab scale. These critical temperature-time combinations for different total solids content were established by varying the oil bath temperature, i.e. the heating intensity of the samples. The investigation showed that the heat stability of the concentrates progressively decreases with increasing total solids content. A correlation of the heat stability of bulk milk and concentrates produced thereof was possible. The influence of the pH on heat stability decreased with increasing total solids content of the concentrates. The heat stability maximum was at pH 6.7 in all cases. A linear relationship between the logarithm of the coagulation time and the heat coagulation temperature could be established for discrete total solids contents of CSM. This indicated that heat-induced coagulation of concentrated milk follows the principles of formal reaction kinetics. However, an immediate transfer of coagulation temperatures and times to pilot and industrial scale was not possible as the heating profiles of the lab scale system and continuous heating were different.
Direct steam injection was used to perform heat stability testing on pilot scale as it enables isothermal heat treatment of concentrated milk (15-31.5% total solids) due to negligible come-up and cooling times. It was possible to determine the onset of coagulation and the course of coagulation over temperature (117-153 °C) of concentrated skim milk of different total solids content at short holding times (< 13 s). A rapid
increase in the amount of sedimentable heat -induced protein aggregates was observed when certain critical temperature-time combinations were exceeded. These aggregates formed from casein micelles had a size of 3-20 μm, although secondary aggregation to particles of 10-100 μm was pronounced. The size of the aggregates enabled their quantitative separation at 4,000xg/10 min in concentrated skim milk. These trials showed a clear correlation between the maximum heating intensity in terms of heating temperature and holding time and the dry matter of the concentrates. A UHT preheat treatment of the milk before concentration resulted in an increase in the heat stability of the concentrate. Higher temperatures or longer holding times became possible without formation of casein micelle aggregates. Heating intensities without coagulation resulted in a remarkable increase in casein micelle size in heated concentrated skim milk. In conjunction with the formation of dissociated casein particles in the range of 20-100 nm, this indicates a marked structural disintegration of casein micelles as a consequence of the heat treatment. This loss of structural integrity of casein micelles could be a preceding reaction to heat-induced coagulation of casein micelles.
Therefore, structural changes of casein micelles and the kinetics of heat-induced aggregation of casein micelles needed to be further addressed. Quantitative ultracentrifugal separation of the different size fractions in direct steam injection heat treated concentrated skim milk was applied. The quantitative separation of the different size fractions was monitored by particle sizing techniques. At the same time, changes in particle size of casein micelles in heated concentrated skim milk of different total solids content were also investigated. The obtained size fractions were analysed for their relative amounts of individual caseins and whey proteins by a refined RPHPLC method from literature to obtain mechanistic insights into the course and mechanism of heat-induced coagulation of casein micelles. The results showed an accelerated increase of casein micelles with increasing total solids content of the concentrates and increased heating temperature as it was observed for heat-induced coagulation. Coagulated casein was formed by calcium sensitive αS- and β-caseins. Increasing the heating intensity increased the amount of dissociated casein, mainly κ-casein. Coagulation of casein micelles was observed when 30-35% of κ-casein had dissociated. In addition, the size of casein micelles had doubled at that point. It is assumed that more than one reaction contributes to the destabilisation of casein micelles leading to coagulation.
Furthermore, an extension of the heat treatment trials of concentrated milk and the knowledge obtained concerning the relationship of heating temperature, holding time, and total solids content of the concentrate was intended to derive kinetic parameters (rate constants, activation energies) of heat-induced aggregation of casein micelles. For this purpose, stainless steel tubes were filled with concentrated skim milk (12-33% total solids) and indirectly heated with saturated steam at temperatures from 103 to 131 °C for 0 to 5000 s. After heat treatment, coagulated casein particles were removed by centrifugation. The course of the coagulation process at 27% total solids heated at different temperatures as well as the concentrates of different total
solids content heated at 116 °C could be describe by a Weibullian model. Using this model, critical limits for the heat treatment of concentrated skim milk of different total solids content could be defined. The temperature dependency of the reaction rate could be derived. A basis for the design of non-isothermal heat treatment of concentrates by continuous integration of temperature-time effects was established. The validation of the model was performed by the data obtained from direct steam injection heat treatments at higher temperatures and shorter holding times.
A comparison of the heat stability test results on lab scale with critical temperature-time combinations using direct steam injection heating on pilot scale as well as data obtained by reaction kinetic calculations was performed. This comparison enables to determine the heat stability of different milk concentrates on lab scale and to predict the heat stability of the concentrates under continuous heat treatment.
Further insights into the mechanism of heat-induced coagulation and the relation of the heat-induced dissociation of κ-casein, the increase in casein micelle size, and the heat-induced aggregation of casein micelles in the complex system of concentrated milk appeared to be difficult. Hence, these mechanistic aspects were assessed by using a model system of micellar casein as obtained by diafiltration. This whey protein- and lactose-free casein micelle suspension was manufactured by multiple diafiltration using a simulated milk ultrafiltrate (SMUF). SMUF is a synthetic salt solution that closely resembles the natural milk serum composition. This SMUF solution was developed based on the results of the analysis of ultrafiltration permeate by analytical high-performance ion chromatography which was established for this purpose. In addition, the salt composition and the physical-chemical properties of the SMUF solution depending on temperature and pH were thoroughly characterized. An identical composition and the similarity of the main physico-chemical properties, especially at different working temperatures without crystallisation of calcium phosphate at different working temperatures could be achieved. The dependency of the calcium activity on SMUF composition and pH could be determined by a calcium selective electrode. The region of supersaturation of the SMUF could be predicted by the determination of the pH-dependency of the calcium activity in skim milk.
This diafiltered casein micelle model system using SMUF as a diafiltration medium facilitated a targeted modification of milieu conditions. The interference of the Maillard reaction with analytical determination of caseins by RP-HPLC and the mechanism of heat-induced coagulation was minimised. An increased rate of κ-casein dissociation due to thiol-disulphide exchange reactions with β-lactoglobulin was prevented. The investigations into the effect of pH on the heat-induced coagulation of micellar casein showed that at pH > 6.7 coagulation of casein was very limited, whereas the dissociation of casein, especially κ-casein, from the micelles was very pronounced. With increasing temperature, the amount of dissociated casein increased at the same heating time. Gradual decrease in pH decreased the amount of dissociated casein over holding time at 116 °C. However, there was a significant increase in casein micelle size that could also be observed when soluble calcium was added. The additional increase in ionic strength after calcium addition and a reduc-
tion in pH induced the coagulation of casein micelles to distinct particles. This indicates that in the pH-range of 6.2-7.2, several factors destabilising casein micelles must be present to reduce the colloidal stability of casein micelles and to induce heatinduced coagulation of casein micelles to particles. These factors comprise an increased amount of soluble calcium, increased ionic strength, and a particular heating intensity which are all present in heated concentrated skim milk. Low pH and high soluble calcium content result in the loosening of the internal structure of the casein micelles. The newly-created calcium-sensitive surface of the casein micelles facilitates the aggregation of the casein micelles by exposure of calcium sensitive caseins on the surface of the micelles. The dissociation of κ-casein does not appear to be directly related to heat-induced aggregation, especially at pH < 6.7. However, it should be regarded as a deviation from the original micellar structure of the casein micelles in unheated milk at its natural pH that affects the physical properties of the casein micelles and is likely to foster heat-induced aggregation of casein micelles.
An extension of the term ‘heat stability’ of the casein micelles is likely to be necessary. A consideration of the dissociation of caseins as well as the loosening of the internal structure of the casein micelles, both related to a loss in native structure is necessary.
The key outcomes of this work can be summarized as follows.
A kinetic description of heat-induced coagulation of concentrated skim milk on lab- and pilot scale for the determination and calculation of critical temperature-time combinations without coagulation of the concentrates. Studies on the dissociation and coagulation of individual caseins in concentrated skim milk heated by direct steam injection. The development of a simulated milk ultrafiltrate, e.g. for the purification of casein micelles, to investigate targeted changes in serum composition on heat-induced changes of casein micelles.
Investigations on heat-induced changes in casein micelle structure and dissociation of caseins in heat treated micellar casein depending on ionic strength, pH and soluble calcium.
Zusammenfassung
Die Notwendigkeit der Bestimmung der Hitzestabilität konzentrierter Milch nahm ihren Anfang mit der industriellen Herstellung von Kondensmilch im späten 19. Jahrhundert. Die Thematik war aufgrund beobachteter Probleme mit der Gerinnung, Sedimentbildung und damit Unbrauchbarkeit einzelner Chargen konzentrierter Milch bei der Sterilisation in Dosen, später in Glas- und HD-PE-Flaschen, von großem wirtschaftlichem Interesse. Seither beschäftigte sich die Milchwissenschaft mit der Thematik der Hitzekoagulation von Milch, insbesondere konzentrierter Milch, deren Einflussfaktoren, Methoden zur Vorhersage und der Verbesserung der Hitzestabilität. Dabei sind trotz intensivster Forschungsarbeit bis heute noch viele Fragen zum Mechanismus, den auslösenden Faktoren der hitzeinduzierten Aggregation der Caseinmicellen und der Modellierung der Reaktion im Sinne der Prozessoptimierung offen.
Die einzelnen Millieufaktoren, die die Hitzestabilität von Milch bzw. mittels Eindampfung oder Umkehrosmose konzentrierter Milch bestimmen, sind sehr intensiv untersucht worden. Zu diesen Faktoren gehören insbesondere der pH-Wert vor der Erhitzung, der Gehalt an löslichen divalenten Kationen im Milchserum, vorwiegend lösliches Calcium, der Gehalt an reaktiven Molkenproteinen und die Ionenstärke im Milchserum. Dabei ist unkonzentrierte Milch mit natürlichen pH-Wert meist am hitzestabilsten, wenn eine Verdünnung nicht in Betracht gezogen wird. Trotz des Versuches, die sich während der Konzentrierung von Milch ändernden Millieufaktoren pH-Wert, Ionenstärke, löslichen Calcium, Proteingehalt teilweise durch Zugabe von Basen, Citrat, Phosphat und vielen weiteren Zusatzstoffen auszugleichen, bleibt die Hitzestabilität konzentrierter Milch im Vergleich zum unkonzentrierten Zustand geringer. Auch andere Membranverfahren der Konzentrierung wie Nano-, Ultra- und Mikrofiltration führen zu einer Verringerung der Hitzestabilität. Dabei wird die Hitzestabilität des Konzentrates umso höher, je weniger dessen Zusammensetzung von der unkonzentrierten Milch abweicht. Dies erscheint plausibel, war jedoch bisher nicht quantitativ in Bezug auf die mögliche Erhitzungsintensität als Funktion der maximalen Temperatur-Zeit-Bedingungen für Konzentrate unter Bedingungen in Durchlauferhitzern ohne Koagulation beschreibbar. Technologische Einflussfaktoren, d.h. vor allem die Verfahrensschritte vor der Konzentrierung der Milch und Erhitzung des Konzentrates, können sich sowohl negativ als auch positiv auf die Hitzestabilität des Konzentrates auswirken. Die Kom-
bination der einzelnen Verfahrensschritte und die Wahl der Prozessparameter sind daher sehr entscheidend bei der Herstellung haltbarer flüssiger Milchkonzentrate. Die Homogenisierung der Milch oder des Konzentrates vor der Erhitzung des Konzentrates wirkt sich negativ auf die Hitzestabilität aus. Im Falle einer geeigneten Vorerhitzung der Milch vor der Konzentrierung kann die Hitzestabilität gesteigert werden. Eine quantitative Aussage in Form der Verschiebung kritischer TemperaturZeit-Bedingungen war bisher nicht möglich. Zudem sind die Prozessparameter zur kontinuierlichen Erhitzung von Milchkonzentraten in Platten- oder Röhrenwärmeübertragern oder mittels direkter Erhitzungsverfahren bisher meist nur empirisch ermittelt und konnten nicht aus der im Labormaßstab ermittelten Hitzekoagulationszeit des Konzentrates ermittelt werden. Eine Berechnung möglicher TemperaturZeit-Bedingungen basierend auf reaktionskinetischen Parametern, wie dies für mikrobiologische Inaktivierungs- sowie chemische Effekte möglich ist, war bisher nicht möglich.
Daher war es das Ziel der hier vorliegenden Arbeit, einen Hitzestabilitätstest und eine Berechnungsgrundlage zu schaffen, durch die sich die Hitzestabilität von Milchkonzentraten unter kontinuierlichen Erhitzungsbedingungen, insbesondere von Magermilchkonzentraten, die mittels Umkehrosmose hergestellt wurden, beschreiben lässt. Die begrenzte Hitzestabilität von Magermilchkonzentraten unterschiedlicher Trockenmasse in Bezug auf Erhitzungstemperatur und -zeit sollte quantitativ im Labor- als auch im Technikumsmaßstab beschrieben werden. Dadurch sollte es möglich sein, Erhitzungsprozesse in Bezug auf erwünschte mikrobiologische Inaktivierungseffekte zu optimieren und die Koagulation der Konzentrate während der Erhitzung zu verhindern. Die Beschreibung ausgewählter hitzeinduzierter Veränderungen an den Caseinmicellen war ebenfalls ein Ziel dieser Arbeit. Daraus sollten Erklärungsansätze für die Verringerung der Hitzestabilität von Milch durch die Veränderung verschiedener Millieufaktoren während der Konzentrierung gefunden werden. Es wurde vermutet, dass bestimmte Millieufaktoren zu einer Beschleunigung der hitze-induzierten Verringerung der kolloidalen Stabilität der Caseinmicellen führen.
Die Untersuchungen zur Hitzestabilität im Labormaßstab erfolgte durch die Erhitzung von mit Magermilchkonzentrat befüllten Probengefäßen im Ölbad und der visuellen Bestimmung des Beginns der Koagulation der Konzentrate. Dabei wurde jeweils die Zeit bis zur einsetzenden Koagulation als auch die momentane Temperatur der Konzentrate registriert, da viele Konzentrate bereits während der Aufheizphase koagulierten. Das Ziel der Untersuchung der Hitzestabilität im Labormaßstab war es, den Einfluss von pH-Wert, fettfreier Trockenmasse, Vorerhitzung und Fettgehalt auf die sichtbare Koagulation im Labor-Erhitzungssystem zu bestimmen. Durch die Variation der Ölbadtemperatur sollte der Zusammenhang zwischen der Hitzekoagulationstemperatur und –zeit und diskreter Trockenmassen gefunden werden.
Die Analyse der Hitzestabilität im Labormaßstab ergab zunächst, dass die Hitzestabilität mit ansteigender Trockenmasse der Magermilchkonzentrate kontinuierlich
abnimmt. Eine Korrelation der Hitzestabilität von Sammelmilch und daraus hergestellter Konzentrate war möglich. Der pH-Wert der Konzentrate hatte mit steigender Trockenmasse einen geringer werdenden Einfluss auf die Hitzestabilität, wobei das Stabilitätsmaximum aller Trockenmassen bei einem pH-Wert von 6,7 lag. Zwischen der Koagulationstemperatur und dem Logarithmus der Hitzekoagulationszeit ließ sich für diskrete Trockenmassen ein linearer Zusammenhang darstellen. Dies deutete darauf hin, dass die hitzeinduzierte Koagulation von konzentrierter Milch reaktionskinetisch beschreiben lässt. Eine direkte Übertragbarkeit der kritischen TemperaturZeit-Bedingungen vom Labor- in den Pilot- und Industriemaßstab war aufgrund der unterschiedlichen Aufheizprofile des Labor -Erhitzungssystems und einer kontinuierlichen Anlage nicht möglich.
Zur Untersuchung der Hitzestabilität im Pilotmaßstab wurde die Direkterhitzung mittels Dampfinjektion verwendet, die es ermöglicht, bei vernachlässigbaren Aufheiz- und Abkühlzeiten die Konzentrate isotherm zu erhitzen. Damit konnten die Koagulationspunkte sowie der Koagulationsverlauf für unterschiedliche Trockenmassen und kurze Heißhaltezeiten (< 13 s) in Abhängigkeit der Erhitzungstemperatur (117-153 °C) bestimmt werden. Das Überschreiten bestimmter kritischer Temperatur-Zeit-Bedingungen führte zu einem starken Anstieg sedimentierbarer hitzeinduzierter Proteinaggregate. Diese Aggregate aus noch deutlich erkennbaren Caseinmicellen hatten etwa eine Größe von 3-20 μm, wobei die Sekundäraggregation zu Partikeln von 10-100 μm ausgeprägt war. Dadurch waren die hitzeinduzierten Proteinaggregate bereits bei 4000 xg/10 min vollständig in den erhitzten Konzentraten sedimentierbar. Eine sehr deutliche Korrelation zwischen der maximalen Erhitzungsintensität als Funktion der Temperatur und Zeit und der Trockenmasse der Konzentrate konnte auch im Pilotmaßstab gezeigt werden. Durch eine UHT -Vorerhitzung der Magermilch vor der Konzentrierung konnte die Hitzestabilität der Konzentrate deutlich gesteigert werden, sodass höhere Temperatur-Zeit-Bedingungen zur Erhitzung ohne die Bildung von Proteinaggregaten möglich waren. Erhitzungsintensitäten ohne Koagulation führten zu keiner Sedimentbildung, jedoch bereits zu einer deutlichen Vergrößerung der Caseinmicellen im erhitzten Magermilchkonzentrat. Zusammen mit der Bildung von dissoziierten Caseinpartikeln im Bereich von 20100 nm deutet dieses auf einen deutlichen Strukturverlust der Caseinmicellen infolge der Erhitzung hin, der die Vorreaktion zur Koagulation der Caseinmicellen darstellen könnte.
Die strukturellen Veränderungen an den Caseinmicellen sowie die Reaktionskinetik der Caseinaggregation sollten daher noch näher untersucht werden. Dazu wurde differentielle Zentrifugation zur quantitativen Trennung der einzelnen Größenfraktionen nach der Erhitzung der Magermilchkonzentrate mittels Direktdampfinjektion verwendet. Die quantitative Trennung der Fraktionen wurde mittels Partikelgrößenanalyse überprüft und die Größenänderung der Caseinmicellen über die Erhitzungsintensität für unterschiedliche Trockenmassen untersucht. Die gewonnen Fraktionen wurden mittels einer weiterentwickelten RP-HPLC Methode auf die relativen Gehalte an einzelnen Caseinen und Molkenproteine analysiert. Daraus sollten Erkenntnis-
se zum Mechanismus der Koagulation und deren Verlauf erhalten werden. Die Untersuchungen zeigten, dass die Vergrößerung des hydrodynamischen Radius der Caseinmicellen ebenso wie die Neigung zur hitzeinduzierten Koagulation mit steigender Trockenmasse der Konzentrate und der Erhöhung der Erhitzungstemperatur zunahm. Koaguliertes Casein bestand vorwiegend aus calciumsensitiven αS- und β-Casein. Mit steigender Erhitzungsintensität stieg der Anteil an dissoziiertem Casein, wobei vor allem κ-Casein dissoziierte. Ab einem dissoziierten Anteil von 30-35% des κ-Caseins setzte die Koagulation der Caseinmicellen ein. Dabei hatte sich jedoch auch die mittlere Größe der Micellen verdoppelt, sodass mehrere Destabilisierungsmechanismen der Micellen wahrscheinlich sind, die zur Koagulation führen.
Die Erhitzungsversuche von Magermilchkonzentraten und die daraus gewonnenen Erkenntnisse zum Zusammenhang zwischen der Trockenmasse (TM) der Konzentrate und den kritischen Temperatur-Zeit-Bedingungen der Erhitzung sollten noch bis zur Ableitung einer Kinetik der hitze-induzierten Aggregation erweitert werden. Dazu wurden dünnwandige mit Magermilchkonzentrat unterschiedlicher Trockenmasse (12-33% TM) befüllte Edelstahlröhrchen mit Sattdampf auf Temperaturen zwischen 103 und 131 °C zwischen 6 und 5000 s erhitzt und jeweils der koagulierte Anteil an Casein durch Zentrifugation quantitativ abgetrennt. Der Koagulationsverlauf des Konzentrates mit 27% TM, das bei unterschiedlichen Temperaturen erhitzt wurde, sowie der unterschiedlichen Trockenmassen, die bei 116 °C erhitzt wurden, ließ sich zeitabhängig durch ein Weibull-Modell beschreiben. So konnten kritische Grenzen für die Erhitzbarkeit von Magermilchkonzentraten mit unterschiedlichen Trockenmassen, die Temperaturabhängigkeit der Reaktionsgeschwindigkeit abgeleitet werden. Eine Berechnungsgrundlage für nicht-isotherme Erhitzung der Konzentrate durch stetige Integration der Temperatur-Zeit-Effekte wurde erstellt. Die Validierung des Modells erfolgte anhand der Daten zur Hitzeaggregation von Magermilchkonzentraten unter den Bedingungen der Direkterhitzung bei höheren Temperaturen und deutlich kürzeren Heißhaltezeiten.
Ein Vergleich der Ergebnisse der Hitzestabilitätstest im Labormaßstab mit den erhaltenen kritischen Temperatur-Zeit-Kombinationen von Magermilchkonzentraten der Direkterhitzung im Technikumsmaßstab sowie den reaktionskinetischen Berechnungen ermöglicht, die Hitzestabilität von beliebigen Milchkonzentraten im Labormaßstab zu ermitteln und eine Vorhersage bezüglich der Hitzestabilität unter kontinuierlicher Erhitzung zu treffen.
Weitere Erkenntnisse zum Mechanismus der Hitzekoagulation und dem Zusammenhang der hitzeinduzierten Dissoziation von κ-Casein und der Micellvergrößerung mit der Aggregation der Caseinmicellen erschienen in dem komplexen Gemisch Magermilchkonzentrat schwierig. Dies sollte daher mithilfe eines Modellsystems, eines micellaren Caseins, das durch Aufreinigung mittels Diafiltration gewonnen wurde, untersucht werden. Diese molkenprotein- und laktosefreie Caseinmicellsuspension wurde durch mehrfache Diafiltration mit simuliertem Milchultrafiltrat (SMUF), einer synthetischen Salzlösung, die dem natürlichen Milchserum sehr nahe kommt, hergestellt. Das SMUF wurde basierend auf den Ergebnissen der dafür etab-
Another random document with no related content on Scribd:
In the Customs of the Augustinian priory of Barnwell, written towards the end of the thirteenth century, the following passage occurs: “The press in which the books are kept ought to be lined inside with wood, that the damp of the walls may not moisten or stain the books. This press should be divided vertically as well as horizontally by sundry partitions, on which the books may be ranged so as to be separated from one another: for fear they be packed so close as to injure each other, or to delay those who want them.”
The catalogue of the House of the White Canons at Titchfield in Hampshire, dated 1400, shows that the books were kept in a small room, on shelves called columpnæ, and set against the walls. A closet of this kind was evidently not a working place, but simply a place of storage. By the beginning of the fifteenth century, the larger monasteries had accumulated many hundred volumes, and it began to be customary to provide for the collections separate quarters, rooms constructed for the purpose. The presses in the cloisters were still utilised for books in daily reference.
In Christ Church, Canterbury, where as early as the fourteenth century, the collection comprised as many as 698 books, a library at Durham was built about 1425 by Archbishop Chichele: the library at Durham was built about the same time by Prior Wessyngton. That at Citeaux, which was placed over the scriptorium, dates from 1480, and that of St. Germain des Prés from 1513. The collection of the latter foundation was one of the earliest in France, and as early as the beginning of the thirteenth century, there is record of its being consulted by strangers. At the time of the French Revolution, it contained 7000 manuscripts and 4900 printed books.[218]
The Queen of Sicily, who in 1517 visited Clairvaux, one of the two great Cistercian foundations in France, describes the library as follows: “On the same side of the cloister are fourteen studies, where the monks do their reading and writing, and over these studies, one mounts by a broad spiral staircase to the new library. This library is 189 feet long by 17 feet wide. It contains 48 seats (bancs) and in each banc four shelves (poulpitres) furnished with books on all subjects, but chiefly theology; the greater number of the said books are of vellum and are written by hand, richly storied and illuminated.”
The phrase “written by hand,” indicates that the Queen was already acquainted with books produced from type, some of which had in fact been produced in Italy as early as 1464.
Another description, written in 1723 by the author of the Voyage Littéraire, speaks of “the fifteen little cells, all in a row, where the Brethren formerly used to write books, for which reason they are still called the writing rooms. Over these cells is the library, the building for which is large, vaulted, well lighted, and stocked with a large number of manuscripts, fastened by chains to desks; but there are not many printed books.”
The provisions of the statutes affecting the library imposed upon the colleges of Oxford and Cambridge, were evidently borrowed directly from the customs of the monasteries. The statutes of Oriel College, Oxford, dated 1329, present an example: “The common books (libri communes) of the House are to be brought out and inspected once a year, on the feast of the Commemoration of Souls (November 2d) in the presence of the Provost or his deputy, and of the scholars (Fellows). Each one of the scholars, in the order of seniority, may select a single book which either treats of the science to which he is devoting himself, or which he requires for his use. This he may keep until the same festival in the succeeding year, when a similar selection of books is to take place, and so on, from year to year. If there should happen to be more books than persons, those that remain are to be selected in the same manner.”
A statute of Archbishop Lanfranc, for the English Benedictines, dated 1070, and based, as he tells us, on the general monastic practice of his time, gives the following regulation: “On the Monday after the first Sunday in Lent, before Brethren come into the Chapter House, the librarian [here called not armarius but custos librorum] shall have a carpet laid down and all the books got together upon it, except those which the year previous had been assigned for reading. These the Brethren are to bring with them, when they come into the Chapter House, each his book in his hand. Then the librarian shall read a statement as to the manner in which Brethren have had books during the past year. As each Brother hears his name pronounced, he is to give back the book which had been entrusted to
him for reading; and he whose conscience accuses him of not having read through the book which he had received, is to fall on his face, confess his fault, and entreat forgiveness. The librarian shall then make a fresh distribution of books, namely a different volume to each Brother for his reading.”
It would appear from this reference as if Lanfranc’s monks were under obligations to read through but one book each year, which was certainly a very moderate allowance. It is also to be noted that the books appear not to have been distributed according to the preferences of the readers, but to have been assigned at the will of the librarian. There must certainly have been no little difference in the character and extent of the duty imposed of reading through one book (even with so long an allowance of time) according to the particular volume which the custos saw fit to assign. The worthy Archbishop writes, however, as if a book were a book and one as good for edification or as fitting for penance as another.
It is evident that there were two classes of volumes, one utilised for distribution for separate reading, and the other reserved for reference and placed in a separate room (first called armarium and later bibliotheca) where they were fastened with irons chains to lecterns or reading-desks.
In the various details concerning the distribution of books, the arrangement of the lecterns for the chained books, etc., the practice in the early colleges was evidently modelled on that of the monasteries. The system of chaining, as adopted in England, would allow of the books being readily taken down from the shelves and placed on the lectern for reading. One end of the chain was attached to the middle of the upper edge of the right-hand board or cover; the other to a ring which played on a bar which set in front of the shelf on which the book stood. The fore-edge of the books, not the back, was turned to the front. A swivel, usually in the middle of the chain, prevented tangling. The chains varied in length according to the distance of the shelf from the desk.[219]
In a copy of Locke’s Treatise on the Epistles, printed in 1711, Maitland found inscribed the following “advertisement”: “Since, to the
great reproach of the nations and a much greater one of our Holy Religion, the thievish disposition of some that enter into libraries to learn there no good, hath made it necessary to secure the innocent books, and even the sacred volumes themselves, with chains (which are better deserved by those ill persons who have too much learning to be hanged and too little to be honest), care should be taken hereafter that as additions shall be made to this library (of which there is a hopeful expectation), the chains should neither be longer nor more clumsy than the use of them requires, and that the loops whereby they are fastened to the books may be rivetted on such a part of the cover and so smoothly as not to gall or raze the books while these are removed from or to their respective places.”[220]
Isidore, Bishop of Seville (c. 560-636), possessed probably the largest collection of books at that time in Europe. It was contained in fourteen presses or armaria, each of which was ornamented with a bust and inscribed with verses. The series of verses concludes with the following notice addressed ad interventorem, a term which may be interpreted a talkative intruder:
Non patitur quemquam coram se scriba loquentem; Non est hic quod agas, garrule, perge foras.
(The scribe allows no one to speak in his presence; there is nothing for you to do here, chatterbox; you had better go outside)—a motto which would serve very well for a reading-room of to-day.
In Rome the Church had, from an early date, preserved a collection of manuscripts which related more particularly to church matters, but which included also some specimens of the Roman Classics. In 855, Lupus, of Ferrières, writes to Pope Benedict III., begging for the loan of certain texts from which to make transcripts. He specifies the Commentary of S. Jerome on Jeremiah, Cicero de Oratore, Quintilian, and Terence.[221]
In the centuries following, however, as the Roman Church sank into a condition of ignorance and strife, and Italy was continuously upset by invasions, the library in Rome and the collections which had been instituted in certain churches outside of Rome were either
seriously lessened or entirely destroyed. As late, however, as 1276, a few valuable manuscripts were still to be found in the church collections. Wattenbach speaks of the collection in Verona, in the library of the Town Hall, as one of the most important of those in Italy in which old manuscripts have been preserved to the present time. Next in importance among the older collections, he mentions that of Hexham in England, which had been originally collected by Bishop Acca in the year 700, and which is referred to by Bede.[222]
With this is to be mentioned the library of York, which is first described by Alcuin.[223]
Among the earlier important library collections was that of the monastery of Vivaria, which had been founded by Cassiodorus; the writings were classified according to their contents, and were arranged in a series of armaria.
After the beginning of the seventh century, the most noteworthy collection was that of Bobbio, a portion of which remained as late as 1618, and was taken by Paul V. for the Vatican Library. Another portion found its way to Turin.
The literary activity of the monastery of Corbie has already been mentioned, and the library there continued in existence during the entire lifetime of the monastery After 1350 the monks appear to have themselves given up the work of writing. Étienne de Conty is recorded as one of the special benefactors of the library. He collected books for it, and he employed special scribes to add to the collection.[224]
In Germany, the monastery of Reichenau was noted as early as 821 for its excellent collection of manuscripts. The librarian Reginbert prepared in 821 an exhaustive catalogue of the collection. Not a few of the manuscripts were, as appeared by the notes in the catalogue, the work of his own hands. Of these manuscripts, which he had prepared with so great zeal and labour, there have remained but five sheets of one book, with a portion of the catalogue.
Of nearly as early a date is the first catalogue of the library of St. Gall, previously referred to; in the catalogue of this there are beneath
the titles various critical notes. There is record of the loan of books to the Emperor Charles III., to Frau Rickert, and to Liutward, Bishop of Vercelli.[225]
In the monastery of Pomposa, in Lombardy, Abbot Jerome brought together in the eleventh century (in spite of certain grumblings on the part of the monks, the ground for which is not clearly explained) a great collection of manuscripts.[226] A certain Henricus Clericus, writing in 1093, describing this collection to a friend, says that in no church, not even in Rome, could so wonderful a group of books be found. Henricus prepared a catalogue of the library, and at the close of the catalogue he finds it necessary, as a matter of consistency, to apologise for the abbot who had ventured to include in the collection heathen books. The presence of such books, known at the time as libri scholastici, was, however, by no means exceptional in monastery collections, and in many of these were to be found copies of Virgil, Ovid, and particularly Cicero. While this was more frequently the case in Italy, it occurred also in Germany. An inventory made in 1233 of the monastery of Neumünster, near Wurzburg, includes in a special list the titles of a number of the Classics.
A similar separate catalogue of libri scholastici was made in 1297 for the collection in the cathedral library of Lübeck.
While the principal increase in the monastery libraries had been secured through the work of scribes and through exchanges, and occasionally through purchases, a considerable proportion of the books came to them through gifts or bequests. The gift that it was customary for a novice to make on entering a monastery very frequently took the form of books.
In 1055, the priest Richlof, in placing his son with the Benedictines, gave as an accompanying present a farm and some books, and his mother gave a copy of a treatise of S. Ambrose.[227]
Léon Maitre says that in Fleury, each new scholar was expected to present at least two codices. Towards the end of the eleventh century, a noble cleric, who entered as a monk the monastery at Tegernsee, brought with him so many books that, according to the
account, when placed by the principal altar they covered this from top to bottom.[228]
In what was known as the Scottish Monastery, near Vienna, there was kept in the thirteenth century a record of gifts, which record includes a long list of presents of books. In the latter part of the century, the monastery appears to have degenerated, the library fell into disuse, and the presents of books ceased. In 1418 the so-called Scottish monks were driven out, and the foundation was taken possession of by Germans. From this date the record of gifts of books again began.
In 1453, the monastery received as a bequest from Dr. Johannes Polzmacher his entire library. The library came to include a considerable list of works on jurisprudence together with a series of classics, including several copies of Ovid. The latter appears to have been a special favourite in the monastic collections. The books on jurisprudence were utilised for the profit of the monastery by being loaned out to the jurisprudence Faculty of the university. They were, it appears, also occasionally loaned to the students for transcribing. In the chance of the manuscripts suffering damage while out on hire, the borrower was compelled to deposit an adequate pledge in the shape either of money or other valuable property.[229]
The monastery in Bobbio received books from wandering Irishmen, as is indicated by the following inscription:
Sancte Columba, tibi Scotto tuus incola Dungal, Tradidit hunc librum, quo fratrum corda beentur.
Qui legis ergo, Deus pretium sit muneris ora.[230]
(Holy Columba of Scotland, thy votary Dungal has bestowed upon thee this book, whereby the hearts of the brothers may be gladdened. Do thou who readest it pray that God may be the reward of thy labour.)
In the monastery of St. Père-de-Chartres the Abbot Alveus, who died in 955, presented to S. Peter a book Pro Vita Æterna. [231]
Dietrich Schreiber, a citizen of Halle, who, notwithstanding his name, is said not to have been a scribe, gave, in 1239, for the good of his soul, to the preaching Brothers of Leipzig, a canonistic manuscript, with the condition that either of his sons should have the privilege of redeeming the same for the sum of five marks, in case he might require it in connection with his study of the law.[232] Robert of Lille, who died in 1339, left in his will to his daughters a certain illuminated calendar, with the condition attached that after their death the calendar was to be given to the nuns of Chikessaund.[233]
It is also the case that bequests securing an annual income were occasionally given with the specific purpose of founding or endowing monastery scriptoria and libraries. The Abbot of St. Père-de-Chartres ordered, in 1145, that the tenants or others recognising the authority of the monastery must take up each year for the support of the library the sum of eighty-six solidos.[234]
His successor, Fulbert, instituted a new room for the collection and kept the monks themselves at work, so that in 1367 a catalogue, inscribed in four rolls, gives the titles of 201 volumes.[235]
Also in Evesham, in Worcestershire, England, a statute enacted in 1215 provides that certain tenths coming into the priory should be reserved for the purpose of buying parchment and for the increase of the library. During the following year the amount available for this purpose was five solidos, eighteen deniers.[236]
The account books of the monks of Ely showed that in the year 1300 they purchased five dozen sheets of parchment, four pounds of ink, eight calf-skins, four sheep-skins, five dozen sheets of vellum, and six pairs of book clasps. In the same year they paid six shillings for a Decretal and two shillings for a Speculum Gregor. In 1329, the Precentor received six shillings and seven pence with which he was instructed to go to Balsham to purchase books. In the same year, four shillings were paid for twelve iron chains (used, of course, for fastening the books safely to the reading-desks). Between 1350 and 1356, the purchases appear to have included no less than seventy dozen sheets of parchment and thirty dozen sheets of vellum.[237]
Prince Borwin, of Rostock, in 1240 presented the monastery of Dargun with a hide of land, the proceeds of which were to be used for the repairing and preservation of the books in the library.[238] Adam, treasurer of the Chapter of Rennes, in 1231, presented his library to the abbey of Penfont, with the condition that the books were never to be diverted from the abbey, and that copies were to be lent only against adequate pledges.
In 1345, a library was founded in the House of the German Brothers of Beuggen, near Rheinfelden, through the exertions of Wulfram of Nellenberg. He directed that all books left by deceased Brothers throughout Elsass were to be brought to this library, and the living Brothers were also earnestly urged to present their own books to the same collection.[239]
The great library of the monastery of Admunt was catalogued in 1380 by Brother Peter of Arbonne. The Chapter of S. Pancras, in Leyden, received in 1380, through a bequest of Philip of Leyden, a collection of eighty manuscripts, the catalogue of which has been preserved.[240]
As before indicated, the Monastery Reform, which was instituted with the beginning of the fifteenth century, exercised a very decided influence upon the interest in books and upon the development of libraries. In Tegernsee, where the once noted library had fallen into ruins, the Abbot Casper (1426-1461) reorganised it, restored such of the old manuscripts as were still in existence, bought new codices, and put to work a number of hired scribes. His successor, Conrad V., carried on the work actively and purchased for the sum of eleven hundred pounds heller no less than 450 volumes, in addition to which he secured a number of gifts or devout presents.[241]
In Salzburg, the Archbishop Johann II. (1429-1441) caused a new library building to be erected, and collected for it many beautiful manuscripts. In the monastery of Bergen, near Magdeburg, the Abbot Bursfelder (1450-1478) organised a library, and utilised for his books an old chapel. In 1477, the Prior Martin instituted a library in Bordesholm, and Brother Liborius, who was a professor in Rostock, gave over, in 1405, to this library, for the good of his soul, his works
on jurisprudence, with the provision that they were to be placed in chains and to remain forever in the reading-room. A catalogue of this collection, which was prepared in 1498, and which contains more than five hundred titles, has been preserved.[242] The library of the Benedictine monastery of St. Ulrich, near Augsburg, retained its early importance until the invention of printing, and in 1472, as before mentioned, a printing office was instituted in connection with the monastery, by the Abbot and the Chapter, in which active work was carried on. Abbot Trithemius presented to the monastery of Sponheim, in 1480, the sum of fifteen hundred ducats for the enlargement of its library.
As before stated, the Brothers of Common Life planned their collections of books expressly with reference to the service of the students in their schools, and these libraries contained, therefore, a much larger proportion of books in the vernacular than were to be found in other monasteries. In some of the Brotherhood Homes, the library was divided into the collection for the Brothers and the collection for the students. It was ordered that at least once a year all books that were not out on loan should be called in and should be inspected in the presence of the Brothers.
Public Libraries.
—Of the libraries of antiquity, only a single one, and that the latest in foundation, the Imperial Library of Constantinople, continued in existence as late as the Middle Ages. This library, founded in 354 by the Emperor Constantius, was largely added to by Julian the Philosopher Under the Emperor Basiliscus, the original library, which at that time was said to have contained no less than 120,000 volumes, was destroyed by fire. It was afterwards reinstituted by the Emperor Zeno, the prefect of the city, Julian, having given to the work his personal supervision. References are made to this library in 1276, and again early in the fourteenth century, when John Palæologus was able to present from it certain manuscripts (probably duplicates) to the well known manuscript dealer Aurispa of Venice. It is probable that the manuscripts of the imperial collection had been to some extent scattered before the fall of the city in 1453. Such manuscripts as had escaped destruction during the confusion of the siege of the city were hidden away by the
scholars interested, in various monasteries and in out-of-the-way corners, from which they were brought out by degrees during the following two or three centuries.
Large quantities of these manuscripts found their way, however, very promptly to Italy, chiefly through Venice, and, as is described in another chapter, not a few of the Greek scholars who were driven from the Byzantine territories, or who refused to live under the rule of the Turk, brought with them into Italy, as their sole valuable possessions, collections of manuscripts, more or less important, which they used either as texts for their lectures or for transcribing for sale.
The collections in the monasteries of the West, brought together in the first place simply for the requirements of the monks and restricted (at least in theory) to devotional or doctrinal books, were, in large measure at least, placed at the disposal of scholars and readers outside of the monasteries, as the interest in literature came to extend beyond the class of ecclesiastics. With this extension of the use of the libraries, there came a natural development in the range of the books collected.
Long after the monks or ecclesiastics had ceased to exercise any control over the books or to be themselves the only readers interested in their preservation and use, the most convenient space for the collection was to be found in the church buildings. Many of the collections came, in fact, to be known as cathedral libraries.
In certain cases, books or money for the purchase of books was bequeathed in trust to ecclesiastical authorities with the direct purpose of providing a library for the use of the general public. The cathedral Prior of Vercelli (in Piedmont), Jacob Carnarius, who died in 1234, left his books to the Dominicans of S. Paul. He made it, however, a condition of the bequest that under proper security of deposit or pledge, the books should be placed at the disposal of any scholars desiring their use, and particularly of instructors in the Theological Faculty of the University of Vercelli.
Petrarch’s library was bequeathed in 1362 to the Church of S. Mark in Venice, with the condition that the collection was to be for
the use of the general public. The books were neglected, and for some time disappeared altogether, and it was only in 1635 that a portion of them were recovered. The famous library of S. Mark dates from 1468, when Cardinal Bessarion presented to the city eight hundred manuscripts, assigning as his reason for the gift the generous hospitality extended by Venice to the refugees from Constantinople. These books were to be for the use of any qualified citizens of the city, a pledge of double the value being deposited for any manuscript borrowed. The library of Boccaccio, who died in 1375, was bequeathed to the monks of the Holy Ghost in Florence. This library was afterwards added to by the collection of the famous theologian, Luigi Marsigli, and that of Niccolo Niccoli.[243]
To Florence, which stood at the front of the intellectual development of Italy, belongs the credit of instituting the largest and most important of the earlier public libraries of Italy. Niccolo Niccoli, one of the most energetic of the scholarly book collectors, specified in his will, made in 1430, that his manuscripts should be placed in the Camal-dulensian monastery of S. Maria, where his friend Traversari was prior, and that these manuscripts were to be available for public use. In 1437, however, the day before his death, he added a codicil to his will, under which the decision as to the abiding-place for his manuscripts was left to sixteen trustees.
He died in debt, however, and the books would have been seized by his creditors if they had not been redeemed by Cosimo de’ Medici. Cosimo placed them in the Dominican monastery of S. Mark, the collection in which, in 1444, comprised four hundred Latin and Greek manuscripts. Cosimo gave much care to the further development of this collection. As has already been mentioned, he used for the purpose the services of the great manuscript dealer, Vespasiano. After the earthquake of 1453, he caused the library building to be restored with greater magnificence than before. The care of the library was continued, after the death of Cosimo, by his son Pietro, and the collection finally became the foundation of the famous Laurentian library, which is in existence to-day.
Pietro took pains to send the Greek grammarian, Laskaris, twice to the Orient to collect further manuscripts. From his first journey,
Laskaris brought back no less than two hundred works, of which eighty had not heretofore been known in Italy. On his second journey, Laskaris died.
The library suffered much during the invasion of Charles VIII., but a large proportion of the books were redeemed from the French invaders by the Dominican monks, who paid for them three thousand gulden.
Cardinal Giovanni de’ Medici (later Pope Leo X.) took the collection from the monastery with him to Rome, but it was afterwards returned to Florence by Pope Clemens VII.
Clemens gave to Michel Angelo the commission to build a hall for the library, but both Pope and architect died before the work was completed, and the building took shape only finally in 1571, the plan of Michel Angelo having been carried out in substance.
The library of the Vatican passed through various vicissitudes according to the interest or the lack of interest of the successive popes, but under Pope Nicholas V. (1447-1455) it became one of the most important collections in the world for the use of scholars. In 1471, Sixtus IV. completed the library building and the rooms for the archives and added many works, and it was under this Pope that the use of the books was thrown open (under certain conditions) to the general public.
Frederick, Duke of Urbino, is reported to have spent as much as 40,000 ducats on the ducal collection in Urbino, and Vespasiano rendered important services in the selection and development of this library. The books were, in 1657, under the papacy of Alexander VII., transferred to the Vatican.
During the fourteenth and fifteenth centuries, there was very considerable interest in literary work in Hungary and some noteworthy collections of manuscripts were there brought together. The collectors in Italy found in fact some of their richest treasures, particularly in manuscripts in Greek, in the monasteries of Hungary and of Transylvania. The cause of literature was much furthered by King Matthias Corvinus, who brought together a very valuable
collection in Ofen. He kept four scribes in Florence preparing works for the Ofen library, and thirty were continually at work in Ofen itself. His wife, Beatrix, who was a daughter of King Ferdinand of Naples, and a grand-daughter of Alfonso the Good, is said to have exercised no little influence upon the literary culture of the Hungarian Court. At her instance, many Italian scholars were brought to Hungary, and their aid utilised in completing the library. The codices Budences came to be well known in the scholarly world, and secured fame both for the beauty of their script and the richness of their adornment. Wattenbach says of these, however, that their text is very largely inaccurate, giving the impression that the transcripts had been prepared hurriedly and to order. After the death of King Matthias, a number of his books came into the possession of Emperor Maximilian, who used them for the foundation of the Court Library of Vienna. This was the only portion of the original Hungarian collection which escaped destruction at the hands of the Turks.
Among the public libraries in France is to be noted that of Louis IX., which was open for the use of scholars, but which, being limited almost entirely to devotional books, could not have been of any great scholarly service. In the middle of the thirteenth century, Richard de Furnival, chancellor of the Cathedral of Amiens, instituted a public library, and himself wrote, as a guide for the same, a work entitled Biblionomia.
The libraries of S. Ulrich and Afra in Augsburg have already been referred to.[244]
According to Savigny, there were before the time of printing no university libraries in Italy. The stationarii provided both instructors and students with such books as were prescribed in the courses, and the demand for others appears not to have been great. In Paris, on the other hand, a collection of books for the use of the students was instituted as early as 1270, the first benefactor being Stephen, Archdeacon of Canterbury. Stephen gave his books to the church of Notre Dame to be loaned to poor students of theology. In 1297, Peter of Joigny, in continuation of the same work, gave a collection of books in trust to the university directly for the use of these poor students of theology. The famous College of the Sorbonne probably
dates from 1253. The librarium of the college was instituted in 1289, and it was specified that the books were for the common use of the instructors and students. The catalogue of this collection, prepared in the following year, is still in existence and contains 1017 titles.[245]
Each socius of the college had a key to the library rooms and was permitted to take guests in with him. The books were all fastened to the wall or to the reading-desks by chains, so that the risk of abstraction was not a serious one. The statutes of 1321 prescribed that of every work issued, one copy in the best form must be preserved for the Sorbonne collection. This is probably the first statute of the kind having in view the preservation, in a public collection, of copies of all works produced. It is to be borne in mind, however, in the first place, that it could have had reference only to books produced under the direct supervision of the college, and secondly, that there was here no question of original literary production, but merely of copies of the older works accepted as possessing doctrinal authority. The books in this library (and probably in other similar libraries) which were not protected by chains were called libri vagantes, and these could, under certain restrictions, be loaned out. Wattenbach is of opinion, however, that no books other than duplicates were placed in this class.
Another library of importance was contained in the College of Narbonne, which had been founded in 1316, and which was itself a continuation of an earlier foundation instituted in 1238 by the Archbishop Peter, at the time he was about to take part in the Crusade. The books were to be open for the use of students as well in Paris as of Narbonne.[246]
In the College of Plessis, the statutes of 1455 described that all books, with the exception of the Missals, must be chained, and that no unchaining should be permitted except with the authorisation of the master of all the bursars. In the College of the Scots, the loaning of books outside of the building was absolutely forbidden.
To the College of the Sorbonne belongs the credit of taking the initiative step in inviting the first printers to Paris. In 1469, the prior and the librarian made themselves responsible for finding work and
support for two printers, called to Paris from Mayence. The fact that the Prior Johann Heynlin was himself a German was doubtless of influence in bringing to the college early information concerning the importance of the new art.[247] The first book which was printed in Paris was the letters of Gasparin of Bergamo, which appeared in 1470 (twenty years after the perfecting of the Gutenberg press), and bore the imprint in ædibus Sorbonnæ.
In England, the foundation of the Franciscans in Oxford took, early in the thirteenth century, active part in furthering library facilities for the clerics and the students. They appear to have had two collections, one called libraria conventus, doubtless restricted to theological and religious books, and one described as libraria scholarium or studentium, which contained a number of examples of the classics. It was to the Franciscans that Bishop Grosseteste, who died in 1253, bequeathed all his books.
The interest in literature of Richard de Bury, the friend of Petrarch, has already been referred to. He was the instructor of King Edward III., and exercised later, important official responsibilities. He served as a foreign representative more than once, and was for a time chancellor of the kingdom. At the time of his death in 1345, he was Bishop of Durham. He had a passion for the collecting of books, and with the exceptional advantages of wealth, official station, and knowledge of distant countries, he had advantages in this pursuit possessed by no other Englishman of the time. It is said that the other rooms in his house having already been crowded with books, these were massed in his bedroom also in such quantities that he could get to his bed only by stepping upon them. His library was bequeathed to Durham College in Oxford, which had been founded by himself. The college was discontinued by Henry VIII., and the books were scattered, not even the catalogue, which Bury had himself prepared, having been preserved. In confiding his books to Oxford for the use of the students, Richard gives various earnest injunctions as to the proper respect in which they should be held and the care with which they should be handled. A reader who should handle the books with dirty hands or while eating or drinking, could, in Bury’s opinion, be fitly punished with nothing less than
banishment. The collection of Durham College was to be open not only to the use of the members of the college itself, but of all masters and students in Oxford, but no books of which there were no duplicates were to be taken out of the building.
The earliest university library of Germany was that of the College Carolinum in Prague, instituted by Charles IV. The next in date appears to have been that of Heidelberg, where as early as 1386 the Faculty of Arts had a library for itself in addition to the general collection belonging to the university. As before stated, there was also a collection in the Castle which was open for the use of all readers, students, citizens, or strangers. The university library in Vienna dates from 1415, and that in Erfurt from 1433. The town library in Leipzig had for its origin a collection possessed by the Augustinian monks in the monastery of S. Thomas, which collection was thrown open for the use of the public in 1445. Additions to the library were to be made only under the inspection and supervision of the monastery authorities.
The most noteworthy library which had no connection with any university was instituted at Alzei (in Hesse Cassel) in 1409. Its founders were Johannes of Kirchdorf, Prebendary of the Cathedral of Worms and chaplain of King Rupert.
The books were given in order that the clerics and other scholarly people who belonged to the city of Alzei “could use the same for entertainment and instruction, and could spread among the community at large the learning contained therein.”[248]
In Hamburg there was, as early as 1469, a collection comprising forty volumes of medical books, for the use more particularly of the city physician and his assistant, and also for general reference. In 1480 the burgermeister Neuermeister left a considerable legacy for the foundation of a city library. In Frankfort, the library of the Carmelite monastery was taken over in 1477 for the use of the city, in order that the “books could be made of service for the enlightenment of the community to the greater glory of God and of the Mother of God.”
Collections by Individuals.
—Among the laity (outside, at least, of Italy) it was particularly the kings who from time to time interested themselves in collecting books. Pepin received from Pope Paul I., at his own urgent request, a collection of books which included certain Greek manuscripts. The latter could, however, hardly have been of any particular service either to the King or to any members of his Court.[249]
The collection formed by Charlemagne has already been referred to, and also the provision of his will, under which, after his death, the books were to be sold and the proceeds given to the poor. Charles the Bald, with whose name it is not easy to associate intellectual activity, appears to have been a great collector of books. After his death his library was, under his directions, divided between St. Denis, Compiègne, and his son.[250] It is recorded of William the Great of Aquitaine, who died in 1030, and who was the father of the Empress Agnes, that “he had many books and read zealously therein.”[251] Count Baldwin of Guines, who died in 1205, brought together a collection of books which he had translated into the Romance tongue. Louis IX. of France was interested in the idea of bringing together a collection of devout books, and, although he did not live to carry out his plan, the manuscripts which were left by him served for the scholar Vincennes of Beauvais in the preparation of his great encyclopedia.
Louis heard, during his crusade, of some sultan who had caused to be prepared transcripts of all the noted works of philosophy. This example incited the zeal of Louis, who gave directions that all the “authentic, useful, and devout books” which were to be found within his realm were to be transcribed, and the transcripts placed in the Royal Library. The collection was, however, not allowed to remain complete, as in his will Louis directed that the books should be divided between the preacher monks and the Minorites of Paris, the Abbey Royaumont, and the Dominican monks of Compiègne.[252]
John, Duke of Berry, son of the Good King John, and brother of King Charles V., found opportunity, even during the troublous times which culminated with the battle of Poitiers and the imprisonment of
his father, to bring together a noteworthy collection of books. It was this collection that made the beginning of the library of the Louvre, instituted by Charles V., a library for which Gilles Mallet prepared in 1373 a very complete catalogue. Barrois published in 1830, in Paris, a work devoted entirely to a description of the books collected by Prince John and his brother Charles.
David Aubert, whose translation of the History of the Emperors was published[253] in 1457, makes, in the preface to this history, special mention of the literary tastes of Philip, Duke of Burgundy. He says that Philip made a daily practice of having read to him ancient histories and that he kept employed a great number of skilled translators, learned historians, and capable scribes who were busied in adding to his great library. This collection of Philip appears later to have been scattered as there is no record of its preservation.
The Duke of Bedford found time, between his frequent campaigns, to interest himself in the collection of manuscripts, and more particularly of works which were beautifully illuminated. He purchased, for 1200 francs, a portion of the library of Charles V., which had been captured, and, these books being taken to Oxford, finally found place in the Bodleian collection.
Philip of Cleves, who died in 1528 and who was connected with the Burgundy House, shared the passion of his relatives for magnificent manuscripts.
An inventory of Margaret, Duchess of Brittany, contains the descriptive titles of eleven books of devotion and four romances, “all bound in satin.”[254]
The name of Anne of Brittany, the wife of King Charles VIII., and later of Louis XII., has long been famous in connection with her fondness for books of devotion and with the great collection which she succeeded in making of these. An inventory of 1498 gives the titles of 1140 books as belonging to Anne’s collection.[255]
In Italy, it was not until the time of Petrarch that there came to be any general interest in the collection of books. This interest was naturally associated with the great Humanistic movement of which it
may be considered as partly the cause and partly the effect. The development of literary interests in Italy during the fourteenth and fifteenth centuries will be considered in the chapter on the Renaissance.
In Germany, the collections outside of the monasteries appear to have been less important than in Burgundy and in France, the difference being probably in part due to the narrower cultivation of the German noblemen, and probably also in part to their smaller resources. In fact, the more important collections do not appear to have been in the possession of the nobility at all, but to have come into existence through the public spirit of citizens of lower degree. The library of two hundred volumes brought together as early as 1260 by Hugo Trimberg, a schoolmaster of St. Gangolf, has already been referred to.
Duke Ludwig of Brieg is described as having had as early as 1360 a considerable collection of books, and as having had written, in 1353, by some scribe whose name has not been preserved, the Hedwig legends.
The Electors of the Palatinate interested themselves in the formation of libraries, having possibly been influenced to some extent by their relations with their neighbours on the other side of the Rhine. Authors such as Matthias of Kemnat and Michel Behaim worked at the instance of the Electors and under pay from them. The books of Kemnat and Behaim were either originally written in German, or were promptly translated into German for the use of the Electors and of their wives. A number of books in this series are also ornamented with pictures, but, according to the descriptions, the art work in these illustrations was much inferior to that done at the same time in Burgundy.
The most important group of the Heidelberg manuscripts was collected by Ludwig III., who died in 1437.[256] His daughter Mechthild, whose first husband was the Count Ludwig of Wurtemberg, and whose second, the Archduke Albrich, retained in her widowhood in her castle at Rotenburg a collection of ninety-four volumes of the mediæval poets, whose works were written in the
