INTRODUCTORY 7 



the fact that it was first studied in an unstable form and dissociation was 

 taking place during the course of the study. On the other hand, some of 

 the methods employed in the hopes of inducing phase variation may 

 actually cause contamination and be incorrectly interpreted. Some of 

 these points are very adequately discussed by Frobisher (1933). 



The third source of error mentioned above (variation in methods) also 

 needs emphasis. When a species is described in such terms as one fre- 

 quently encounters in published descriptions, e.g., ''produces acid (with- 

 out gas) from glucose and lactose but not from sucrose; does not reduce 

 nitrates," one has to guess at the answers to such questions as these: 

 What basal medium was used in each instance? What indicator of acid 

 production was employed? How thorough a study was made to show 

 the absence of any acid from sucrose or of any reduction of nitrate? Or, 

 in the last instance, is it safe to assume that the author of the species 

 merely failed to find nitrite in some nitrate medium? Unless such ques- 

 tions are answered correctly, the description is meaningless; the attempt 

 to identify an unknown culture with such a description may well give 

 misleading results. 



With all these pitfalls to avoid, it is easy to see how the same set of 

 data, no matter how carefully prepared, can be differently interpreted by 

 two different bacteriologists. As a result extreme caution is urged, both 

 in determining the identity of a culture and in deciding whether or not to 

 pronounce it a new species. 



Practical Hints 



Determining the characteristic of a culture. One should always, if 

 possible, make a complete study of a culture promptly after its first isola- 

 tion while it is in a condition to display its true characteristics. When a 

 culture has been carried in the laboratory for a long period of time, it may 

 change in some respects from the original. When practical, such cultures 

 should be exposed to conditions which might bring them back to the 

 "normal." When this is done, however, the possibility should always 

 be recognized that by such manipulation dissociation may be induced so 

 that the phase subsequently studied may be quite different from the 

 original isolation. Whenever distinct evidence of dissociation is observed, 

 each phase should be studied and recorded separately, and efforts should 

 be made to reverse ihe change or to obtain the same change with other 

 strains until the possibility of impure cultures seems to be out of the 

 question. No importance should ever be attached to a single determina- 

 tion unless supported by repUcations giving the same results. In describ- 

 ing morphology, one should not be contented with one or two observa- 

 tions but should study several transfers and should follow up each of them 

 day by day for about a week. When changes are observed, a careful 



