PROBLEMS OF METHOD 59 



change. Nevertheless, in the absence of adequate evidence, these phe- 

 nomena should not be accepted as sufficient explanation of the failure 

 of one laboratory to confirm the work of another. It is clear that ex- 

 tensive recourse to such "explanation" would mean that no study 

 could be confirmed or contradicted by later evidence. 



From this review of methodological problems, a number of recom- 

 mendations may be formulated. Although not of equal importance 

 and not necessarily applicable to all cases, they should be considered 

 carefully in designing experiments on the carbon nutrition of fungi. 



1. Complex natural products which supply carbon should if pos- 

 sible not be included in the basal medium. 



2. A control series with only the basal medium should be included 

 in every experiment. Failure to include such a control makes it almost 

 impossible to detect limited utilization of the test carbon compound, 

 since it is not known to what extent impurities in the basal medium 

 support growth. 



3. Adequate replication of cultures is necessary, the number to be 

 determined by the variance found in pilot experiments. Probably 

 for most fungi a minimum of four replicates for each harvest is re- 

 quired. 



4. The carbon source should always be autoclaved separately from 

 the basal medium and added aseptically to it. If it is found to be 

 labile to heat, it should be sterilized by filtration. 



5. The pH of the medium should be determined electrometrically 

 after all ingredients are added, and periodically during growth. 



6. The inoculum should be of minimum size, to avoid carry-over 

 of materials from the parent culture. In doubtful cases, serial sub- 

 culture on the same medium may be necessary. All cultures should be 

 inoculated with the same amount (mass or number) of cells. 



7. The measurement of growth must be objective and quantitative, 

 e.g., by dry weight, total nitrogen, or turbidity (Chapter 1). 



8. For fungi which have rapid rates of growth and autolysis or 

 which grow adaptively on some substrates it may be necessary to 

 construct a growth curve for each carbon source or at least to make 

 more than one harvest. 



9. Analysis of the spent medium for residual carbon source is 

 highly desirable, although not always practicable. 



10. Possible interactions between carbon and nitrogen sources 

 should be considered; in some cases it may be necessary to study all 

 putative carbon sources in media with different nitrogen sources. 



11. When applicable, the distinction between initiation of growth 



