PROTOZOA 



variety 9, but which are not as closely related as are varieties 4 

 and 8) show greater resemblances than either does to variety 9. 

 Further, their serologic resemblances do not seem to be as close 

 as those which exist between varieties 4 and 8. 



In spite of the profound differences between the two groups of 

 varieties, some serologic similarities are still detectable. Thus, se- 

 rotypes D and G are found in both groups. Moreover the D sero- 

 type of a strain of variety 1 may be more like the D of a strain 

 of variety 4 than it is like the D of another strain of variety 1. 

 These serotypes may be ancient and widely distributed. Preer 

 (unpublished) has found serotype G even in the distinctly differ- 

 ent species P. caudatum. 



The body of knowledge now available on the serotypes of P. 

 aurelia, although limited to only 6 of the 16 varieties, is already 

 of some practical value in identification. If one has a battery 

 of antisera against the various major known D, G, and S sero- 

 types of variety 1 and grows an unidentified strain at low, inter- 

 mediate, and high temperatures, the strain could be identified 

 with reasonable confidence as belonging to variety 1 if the ani- 

 mals are immobilized by one of the D antisera at high tempera- 

 ture, one of the G antisera at intermediate temperature, and one 

 of the S sera at low temperature. Even a less complete series of 

 responses would be highly indicative. In like manner identifica- 

 tions of strains belonging to varieties 3 and 9 could probably be 

 made with appropriate antisera; but at present work on these va- 

 rieties is less extensive and doubtless a fuller battery of antisera 

 than now available would be required. The procedure with varie- 

 ties 2, 4, and 8 would have to be somewhat different because of 

 the absence of the regular temperature relations and, conse- 

 quently, the great strain differences in serotypes which are stable 

 at any particular temperature. One would have to use all avail- 



le antisera against the whole array of serotypes known in the 

 variety. In variety 4, there are 17 such serotypes and a number of 

 variants of some of them. Moreover, new serotypes arc 1 continu- 

 ally being found as more 1 strains are studied. Because o\ this great 

 variability and the incompleteness of present knowledge, the 

 probability of success in attempts at serologic identification is 



