58 CARBON NUTRITION 



hexoses (138), and furfural formation from autoclaved xylose is suffi- 

 cient to inhibit bacterial growth (120). Autoclaving of carbohydrate 

 in the medium has been reported also to cause the formation of growth- 

 promoting substances for bacteria (36) and fungi (114). 



Oligosaccharides may be partially hydrolyzed during autoclaving, 

 especially at low pH (7, 22), and probably many claims of slight growth 

 on oligosaccharides reflect merely the partial hydrolysis of a non- 

 utilizable sugar to a utilizable monosaccharide. 



It follows that sugars and other carbon sources should never be 

 autoclaved in a medium. As a general rule, they may be autoclaved 

 separately and then added aseptically to the cooled basal medium. 

 The expedient of intermittent sterilization is no better than autoclav- 

 ing (136). The ultimate recourse is to filter sterilization, and ques- 

 tionable results with a possibly labile sugar should always be checked 

 with this method. 



It is usual in carbon nutrition experiments to harvest all cultures 

 after a predetermined, often arbitrarily chosen, incubation period. 

 No objection can be made to this procedure if the only purpose of the 

 experiment is to find a practicable growth medium for a specific pur- 

 pose. If, however, it is intended to yield meaningful information on 

 the intrinsic utilizability of compounds, some error may be introduced. 

 One compound may be used with a very long lag, probably adaptive 

 in nature (p. 86); the classification of this compound as utilizable 

 or non-utilizable will depend on the length of the incubation period 

 chosen. On the other hand, if time is allowed for slow utilization of 

 compounds, growth on rapidly utilized substrates may have been com- 

 pleted and followed by extensive autolysis before the cultures are 

 harvested. Ideally, both these difficulties may be met by planning 

 periodic harvests; comparison of compounds can then be made on the 

 basis of two criteria, maximum growth and the time required to reach 

 maximum growth. The data of Lilly and Barnett (116) show for 

 many fungi the gain in understanding which flows from the use of 

 more than one incubation period. 



Some additional information, and considerably greater confidence in 

 interpretation, may be gained by analysis of the medium at the time 

 of harvest for residual carbon source. This procedure affords an ad- 

 ditional check upon growth evidence, especially if growth is limited. 



Differences in results between different laboratories are often casually 

 explained as an expression of strain differences or as the result of 

 variation. There is, of course, no doubt that strains of a given species 

 may differ in their nutritional requirements, and there are authenti- 

 cated cases of metabolic changes resulting from spontaneous genetic 



