if Even supposing the vestiges of such 
ATAVISM. 161 
otomized unequally, and is thus more likely to be 
erroneously interpreted from a superficial examination. 
Bardeleben has recently attempted most strenuously 
to prove that the ancestors of modern mammals were 
heptadactyle rather than pentadactyle, and he bases this 
opinion on a supposed discovery in the human embryo 
of some remnants of the missing digits, on the pre- and 
post-axial side of the foot respectively. In addition he 
draws attention to the existence in some mammals, 
chiefly rodents, of an ossicle in the foot, usually regarded 
as asesamoid. This Bardeleben regards as 
vestiges of the missing digits on the inner 
side of the manus and pes, and terms 
prepollex and prehallux respectively. On 
these grounds he would urge that super- § 
numerary big-toes and thumbs are ata- 
vistic. 
FIG, 85.—Super- 
numerary digit 
digits to exist as Bardeleben believes, it in the manus 
eta infil b litt] | eee of a horse due 
would influence but little the views now  ¢6 dichotomy, 
held regarding the nature of bifid thumbs  (Modifiedfrom 
f : Cheaveau.) 
and big toes, for by an overwhelming amount 
of evidence it can be shown that in man they arise by 
dichotomy of the digits. 
A full study of the question of supernumerary digits 
in all its bearings, and the examination of a very large 
number of specimens convinces me that an excess in 
the number of digits, in pentadactyle mammals, is due to 
dichotomy, and can in no sense be regarded as reversion. 
Supernumerary digits in animals with fewer digits than 
five are, in some few cases, due to atavism, but in many 
I2 
