1904] HOLFERTY: ARCHEGONIUM OF MNIUM 111 
worts; the others declare it to be peculiar to mosses. In the 
full accounts some authors state that this process of basal seg- 
ment cutting ceases very early; one that it continues late in the 
growth of the archegonium; and a third implies this, though it 
is not stated. The disagreement as to time is thus as evident as 
the disagreement as to method. 
It may be argued that the disagreement in these conclusions 
is to be accounted for by the fact that they are drawn from. 
observations on different forms; yet it would seem that such a 
process in closely related forms would not be radically different. 
There seems to be no alternative but to examine the develop- 
ment of the archegonium in a sufficient number of forms to dis- 
cover whether there is any such thing as general agreement, and 
as a contribution to this the record of events in Mnium cuspidatum 
is here presented. 
In this species the terminal cell continues to cut off successive 
peripheral segments laterally (jigs. 13-20, 24, 30 34,37), and 
these increase the length of the archegonium neck. Trans- 
verse divisions in the peripheral cells and in the canal cells go on 
simultaneously (fig. 75). The presence of mitotic figures in the 
same phase side by side in the two rows leaves no room for 
doubt on this point, and establishes the fact that part of the 
growth in length of both neck and canal is intercalary. The. 
terminal cell now cuts off a second cell from the lower face. 
This statement rests on the presence of the mitotic figure shown 
in fig. 20. The search for a figure in this particular cell was a 
long one, and its value should not be underestimated, for it 
proves that the terminal cell contributes to the growth of the 
axialrow. The process would seem to be unique, differentiating 
the mosses from all other archegoniates in this particular. 
In this connection it may be stated that careful search was 
made in the literature bearing upon this and related points for 
confirmatory evidence of the relationships of the members of the 
axial row based on the presence of mitotic figures, but no such 
evidence was found. Hundreds of figures illustrating all stages 
of archegonial development in both mosses and liverworts were 
examined, but in none of them are spindles shown. Conclusions 
