114 The Philippine Journal of Science 1922 
Were I to treat this form as a variety of M. carteri I would 
be assuming that the female coenobia of the variety contain a 
much larger number of reproductive cells than are contained in 
the asexual coenobia. This would involve the assumption that 
no female coenobia are shown in the figures or described, and 
that the so-called eggs are in reality very large gonidia. Grant- 
ing that the so-called fertilized eggs are actually such, then the 
variety has no place in M. carteri, but stands as a separate 
species of the genus. 
Male coenobia of the kind found and described in Migula’s 
preparations were observed by Klein (’90, p. 27) in the collections 
of the following season almost to the exclusion of another form 
of male coenobium which Klein had previously described for 
what he considered to be the same species, namely, Volvox aureus. 
After recognizing the occurrence of two forms of male coenobia, 
he observed that, though either form might be met with at any 
time during the vegetative season, as long as sexual coenobia 
were being formed, from May to November, nevertheless the two 
forms almost never occurred simultaneously in the same habitats, 
and when, in exceptional cases, they were so found together, it 
was in each case only for a short time. The male coenobia that 
he had described in his first paper he called, in drawing a 
distinction in his third paper, “normal, typical, large Sphaero- 
sirae.” He called them normal and typical because he took them 
to be like the material on which the former genus Sphaerosira 
was founded by Ehrenberg (’38).1! The male coenobia of the 
kind found in Migula’s preparations he called “small Sphaero- 
sirae’”’ and he further distinguished them by the name Endo- - 
sphaerosira. The normal Sphaerosira averaged 350 to 450 » in 
diameter and was found to reach even 750 ». The androgonidia 
in this form were hardly distinguishable from the vegetative 
cells at the time of birth, and the development of the antheridia 
was correspondingly late. The Hndosphaerosira, on the other 
hand, was invariably smaller, usually between 100 and 150 » in 
diameter, and commonly ripened its antheridia before birth, or 
at least had its androgonidia segmented at that time. It also 
contained a much smaller number (120 to 140) of vegetative 
cells than the normal Sphaerosira, though the number of anthe- 
ridia (60 to 70) was relatively larger. Klein’s figure of an 
Endosphaerosira (’90, fig. 32) shows a coenobium very similar 
* Klein (’90, p. 82) called attention to the fact that De Toni (’89, p. 536) 
erroneously assigned Sphaerosira volvox Ehrenb. to Volvox glabator Ehrenb., 
whereas it is the male coenobium of V. aureus Ehrenb. 
