Danforth: Is Twinning Hereditary ? 
embryonic cell-mass and from each of 
these centers a_ distinct individual 
develops. Such twins, known as zden- 
tical, homologous, or uniovular, are always 
of the same sex and, moreover, often 
show the most striking similarity.2, The 
evidence in favor of the existence of 
this type of twins is drawn from many 
sources and seems quite conclusive.* 
It will be apparent from the foregoing 
that there are at least two fundamentally 
different classes of twins which must be 
taken into account in a study of hered- 
ity. Biovular twins, since they owe 
their existence to the simultaneous 
ovulation of two ova instead of one, 
must be explained entirely by reference 
to some attribute of the mother and 
could not possibly be due to any char- 
acteristic of the father. Uniovular 
twins, on the other hand, might con- 
ceivably be due to factors supplied by 
either the father or the mother or, like 
a recessive character, to some inherent 
peculiarity. in the germ cells of both 
parents. In one case the characteristic 
is manifested in the parental generation 
(mother), in the other it appears in the 
filial generation (twins themselves). 
This is an obvious and important dis- 
tinction. Indeed it might be argued 
that the occurrence of uniovular and 
biovular twins represents two entirely 
distinct and unrelated phenomena. Yet 
this point is frequently neglected in 
studying twins from a statistical or 
hereditary standpoint. 
Simon Newcomb’s memoir? “‘A statis- 
tical Inquiry into the Probability of 
Causes of the Production of Sex in 
Human Offspring,’’ may be mentioned 
as one illustration of such an oversight. 
In this article it is shown from very 
extensive data that twins are of the 
same sex much more frequently than can 
be explained by reference to the laws of 
‘chance. Even so careful a student as 
Professor Newcomb, proceeding from 
197 
this fact and completely ignoring the 
possibility of the existence of these two 
types of twins, tried to prove that the 
sex of twins, and therefore all embryos, 
is determined subsequent to*fertilization. 
If, however, the current views as out- 
lines above are correct it follows that, 
while the data presented by Newcomb 
are exactly what would be expected, 
his arguments from them are practically 
pointless. 
THE TWO KINDS OF TWINS 
Since embryologists insist that stu- 
dents of heredity recognize the existence 
of these two classes of twins, the next 
question that arises is as to how they 
may be distinguished. It is commonly 
assumed that twins of opposite sex are 
necessarily biovular, while those of 
similar sex may belong in either class. 
It therefore becomes a question of 
passing judgment on the degree of 
resemblance between the members of 
each pair where the sex is the same. 
This is no easy matter. On the one 
hand biovular twins may sometimes 
closely resemble each other as is shown 
by the fact that two brothers or sisters 
born several years apart are frequently 
very similar. That uniovular twins, 
on the other hand, may be very different 
is strikingly indicated by those cases in 
which one of the individuals has suffered 
from some handicap before birth, or is 
reduced to a mere parasite attached to 
its more vigorous partner. Identity is 
rarely attained. 
The relation of the foetal membranes 
in which the twins develop has been 
proposed as a criterion, and it is probably 
true that any pair of twins which at 
birth are found to be surrounded by a 
single set of membranes have come from 
a single ovum. But it does not neces- 
sarily follow that those surrounded by 
separate sets of membranes are biov- 
ular. That they frequently are not 
2 The extent of this similarity, which may reach even to the finger prints, has been studied by 
Professor Wilder and others: see H. H. Wilder in the American Journal of Anatomy, vol. i. 
3 The following observations may be mentioned: (a) in some lower forms it is possible 
experimentally to cause two embryos to develop from one egg; (b) in the case of the North Amer- 
ican Armadillo, it is definitely established that four young regularly develop from each egg; (c) 
embryos showing all grades of division from a slight bifurcation to complete separation of individ- 
uals are actually found. 
4 Carnegie Institute of Washington Publication No. 33. . 
5 This is the present view of Sobotta and others, but such data as are referred to in the text 
seem to throw some doubt upon it. 
