288 M. M. Mercatr 
the ball of sticky material near the tip of the tail. These are the 
true microgametes. I have seen individuals of the first type in a 
late stage of division and have followed the division to its com- 
pletion seeing two forms with very long tails arise from one short- 
tailed form (Figs. 159, 160, Pl. XXIII). I have never seen the 
transformation of such daughter-cells with long straight tails into 
typical microgametes with bent tails bearing a ball of sticky matter 
near their tips. This transformation does not occur immediately 
after the division. It seems to be the tailed microgamete mother- 
cells which NERESHEIMER has described as the gametes of O. dimi- 
diata. I do not find either these or the true gametes of either type 
“ganz platt’, as he describes them. All are generally circular in 
cross section, though they may be broadly oval.. The microgametes 
vary in size; so do the short-tailed forms from which they arise. 
The largest of the true microgametes are nearly or quite as large 
as the smallest of the short-tailed forms. 
In my preliminary notice (Mercatr 1907) I wrote “The tailed 
gametes are of two sizes, one about twice as large as the other, the 
smaller being found from the larger by longitudinal division”. This 
was probably an error. There are larger and smaller microgametes, 
the largest being fully twice as large as the smallest, but it is 
doubtful if the latter arise from the former, both probably arising 
by division of the short-tailed forms, as described. In one case, I 
have seen, in O. dimidiata, a microgamete mother-cell, in process of 
division, attached by its unusually long straight tail to the center 
of the body of a uninucleated macrogamete no larger than the 
microgamete mother-cell (Fig. 313, Pl. XXVIII). The attachment 
was a firm one, lasting over three quarters of an hour while the 
animals were ‘actively swimming. During this time no change oc- 
curred. The individuals were then lost. This seems to have been 
probably an abnormal attempt to copulate on the part of a micro- 
gamete mother-cell. I doubt if it would have been successful or if 
it indicates that fully formed and functional microgametes are ac- 
customed to divide. 
Both sorts of gametes are often numerous in the rectum of the 
tadpole. The forms resembling macrogametes are far the more nu- 
merous, probably in part because one cannot distinguish the defini- 
tive macrogametes from forms destined to divide further. The lar- 
gest number of microgametes seen intone rectum was seventy eight; 
in the same rectum the number of pairs was forty-two; and the 
number of the larger, tailed forms (parent-cells requiring probably 
