MUSEUM OF COMPARATIVE ZOOLOGY. 93 
According to his conclusions the reversion is produced by the longitu- 
dinal expansion of the dorsal region. It should be borne in mind that 
previous to reversion the ventral plate extends nearly around the egg, 
bringing the head end and the tail end near together on the dorsal side, 
and that the narrow region separating these two structures represents 
properly the whole of the dorsum. By expansion of the dorsum the 
head and tail are removed further and further apart, and, according to 
his notion, the embryo naturally bends upon itself ventrally, since it 
cannot straighten out, and since, as he erroneously states, it does not 
become shorter. 
I have already shown (Pl. VIII. figs. 50-52) that the tail-lobe be- 
comes separated from the rest of the body during reversion, a condition 
that has not hitherto been recognized, but which is very apparent in 
sagittal sections. This fact alone is sufficient to show that Balfour’s 
explanation as stated is untenable, for were reversion produced by a sim- 
ple expansion of the dorsal region, — unaccompanied by shortening of 
the embryo, —at the end of the process the tail would still exist as an 
elongated conical appendage, instead of being shortened almost to oblit- 
eration. In reality, however, a shortening of the ventral band does 
take place, which is at least equal to the pre-existing tail-fold, and the 
tail is in consequence drawn forward ventrally. This shortening would 
tend to make the bands assume the position which would make the dis- 
tance between head and tail least. There are further to be accounted 
for in this period, the wide divarication of the nerve bands and the ven- 
trad movement of the passive yolk mass. The cause for the latter must 
be found in the relative pressures exerted upon it by the dorsal area on 
the one hand and the ventral area on the other hand: that area which is 
increasing most rapidly in extent would exert a constantly diminishing 
pressure ; however, the shape of the area is of the utmost importance. 
Although the dorsal region is changing its proportions most, it changes 
rapidly from a very broad and short condition to a long and narrow one. 
The principal force, then, that pulls the nerve bands away from the ven- 
tral surface is the one which tends to reduce the width (not as Balfour 
will, the one that increases the length) of the dorsal region. The evi- 
dent cause for this reduction in width is the dorsad concentration of the 
ectodermic elements which accompanies the formation of the so-called 
terga, and this is also the cause for the descent of the yolk mass, for the 
cells that are brought close together to form the thickened ectoderm of 
the tergal region put the remaining cells of the dorsal region to a tension, 
the force of which is exerted upon the yolk mass. It will be remem- 
