730 The American Naturalist. [September, 
ured and these, as seen in the accompanying plate, leave much 
to be desired regarding the details of the bifurcations. 
The period in which these monstrosities arose is not well 
‘known; whether they were present in the embryo or were _ 
formed in the maturer period of the individual’s existence. 
Yet there is little support for the former supposition, while for _ 
the latter we have in two cases good evidence and in many 
others considerable presumption towards this conclusion. 
Thus the case shown in fig. 6 can hardly have arisen other- 
wise than as a consequence of the loss of a normal large head — 
in which the normal pharynx could function. Again, the 
somewhat doubtful double-headedness of the annelid figured by 
Bonnet, figs. 7 and and 8, is explicitly stated by him to have 
arisen after the normal anterior portion had been cut off. 
Moreover a considerable number of the observed cases have — 
been found while experimenting upon the power to reproduce 
lost parts. 
Granting for the present that these monstrosities have arisen 
in late life after removal of parts of the main axis or after _ 
injuries, we may next enquire how far the two new ends are 
of equal or unequal value, whether, as figs. 2 and 4 would 
indicate, the two new parts are equal in origin or whether one, 
as in fig. 10, etc., is to be regarded as a subordinate part or 
lateral outgrowth from the main trunk. In both cases of © 
double-headedness, figs. 6, 7, 8 the left is the more complete of 
the two heads; amongst the cases of bifid posterior ends three 
have the left more developed, one the right and two others an 
undetermined side exceeding the other. Only two, figs. 2 and- 
4, are known to have undoubtedly equal ends. 
As far as the evidence goes (and it is too scanty to warrant 
binding conclusions), there is some indication that one of the 
heads or tails is a supernumerary part growing out, often on 
the right, as a somewhat imperfect duplication of the normal — 
end or continuation of the main axis. 
` Whether the two branches are at first equal, as in fig. 2, 
and subsequently become unequal or, as seems probable 1m- 
many cases, one is at first only a side bud on the main axis, 
cannot be determined as yet. | 
