SOURCES OF CARBON. 207 



the pentosans (C.H 8 4 ) occur in abundance as an important con- 

 stituent of the vegetable cell-wall in cereals as well, B. TOLLENS 

 and H. GLAUBITZ (I.) having found up to 8.9 per cent, (based on 

 dry matter) in barley, 11.2 per cent, in malt, 8.7 per cent, in 

 wheat, 1 1. 1 per cent, in rye, and 5.8 per cent, in maize. The 

 question whether and in what proportions the pentosans of barley 

 undergo hydrolysis during germination has not been exactly de- 

 termined, even in the exhaustive researches of CROSS, BEVAN, and 

 CL. SMITH (I.), though it is certain that this occurs during the 

 kilning of malt, varying amounts of furfural being produced, 

 according to the working conditions. Tollens and Glaubitz state 

 that about three-fourths of the pentosans of the malt are left in 

 the grains by the mashing process in the brewery, partly, how- 

 ever, no longer in the form of pentosans, but as the resulting 

 pentoses. These, though unfermentable (see chapter Ixix.), may 

 serve as sources of carbon for the yeast, provided the external 

 conditions be favourable, as in the case of certain experiments 

 conducted by H. VAN LAER (I.) and by CROSS and BEVAN (I). In 

 other cases, however, as shown by BEYERINCK (XXI.) with regard 

 to arabinose in the case of Schizosacch. octosporus, they are utilised 

 to only a small extent or not at all. Pentoses are formed, in still 

 larger quantity than in brewing, in raw-grain distillery mashes, 

 where the raw grain is dissociated by steaming for seveial hours 

 under a pressure of 3-4 atmospheres. 



BEYERINCK'S (XVIII.) proposal to divide the genus Saccharo- 

 myces into six sub-genera: Glucomyces, Maltomyces, Lactomyces, 

 Raffinomyces, Polysaccharomyces and Dextrinomyces, in accordance 

 with their characteristic behaviour toward the various carbo- 

 hydrates, is scarcely feasible (see chapter ix.). 



When more than' one assimilable source of carbon is present in 

 the nutrient medium, selective power (see p. 45, vol. i.) is exer- 

 cised. In yeasts the study of this property so far as sugars are 

 concerned is very difficult, and little progress has been made, 

 because of the intervention of fermentative action in most cases, 

 so that the separate determination of the amount of sugar con- 

 sumed for the structural purposes of the cell cannot be performed 

 with sufficient accuracy, if at all. A great influence on the ratio 

 of the quantities of two or more nutrient substances in unit time 

 is exercised by their relative diffusibility, this being also deter- 

 minative when two or more fermentable sugars are at the 

 disposal of, and being fermented by, the yeast. The selective 

 power of yeast in fermentation has already been accurately 

 tested, but the results must be postponed to chapter Ixix., where 

 also the extensive literature on selective fermentation will be 

 cited. At present we can only take into consideration the dis- 

 covery that, when two or more diffusible carbohydrates are 

 present, the one exercising the greater osmotic pressure will 

 diffuse more abundantly in the cell per unit of time, glucose, for 

 VOL. ii : PT. 2 o 



