258 University of California Publications in Zool^ogy [Vol. 20 



might produce a bimodal curve are non-operative. Factors involved 

 in faulty technique were eliminated as far as possible. Over the effect 

 of growth or age variation, however, the observer has no control. The 

 possibility of these variations producing such curves as the one above 

 is precluded by reference to the data from which the curves are con- 

 structed. Among the individual measurements recorded in Table I 

 it will be noted that there are small individuals measuring 27 X 50/i, 

 and others measuring 42 X 50jn; and that among the larger individ- 

 uals, some measure 50 X 120/t, and some, 90 X 150/*.. Further study 

 of Table I shows that the two types are found among all sizes of 

 individuals; consequently the variation of bod.y proportions repre- 

 sented by the graphs is not correlated with variations of size, and 

 probably not with the age or growth of the individuals. 



That the variation could be accounted for by the occurrence of 

 fission seems unlikely. Individuals might continue to elongate until 

 binary fission occurred, and then by this process they might be 

 shortened and the body proportion changed. Two considerations 

 oppose this explanation. In figure A nearly equal numbers are of the 

 respective types ; in figure B each example contained almost exclusively 

 one tj-pe of individual. In the former material very few dividing 

 individuals were found, while in the latter not a single individual was 

 seen in fission in either sample of material. But to accord with the 

 above explanation one would expect to find many dividing forms 

 among the elongate individuals represented by the broken line. In 

 the second place, if this explanation were valid one would expect 

 curves showing variation of body proportions to be continuous. Such 

 is not the case, as is shown in figures A and B where the curve is 

 bimodal due to a decided decrease of individuals having proportions 

 intermediate between the two types. 



The possibility of any effect from gametic variation, was eliminated 

 by the study of conjugating forms, through which it was determined 

 that isogamy was the rule. Likewise the possibility of influence of 

 the quality of intestinal content of the host was eliminated by the 

 frequent occun-ence of both types in the same host. Other factors it 

 would seem must be of minor importance and should give way for 

 more positivelj^ corroborative evidence which may be found in cor- 

 related morphological difference. 



Other specific characters. — In addition to the differences in relative 

 lengths of the axes, one notes a distinct difference in the points of 

 intersection, due to the variation in the shape of the tjT)es as pictured 



