JUNE 14, 1912] 
found in the dogfish, contain sufficient stored food 
to account for the amount of growth that occurs, 
the conclusion seems to be justified that the grow- 
ing tissues receive their food supply from other 
tissues transplanted with them. Even when the 
culture medium contained both proteins and carbo- 
hydrates there was much more growth than when 
more than one kind of tissue was present. 
In both my own experiments and in the pub- 
lished accounts of the work of others many tissues 
differentiate abnormally in the medium used, and 
in many other cases their identification with normal 
body cells seems at least to be doubtful. In other 
instances tissues which can be identified have been 
observed, but so far as I know these have never 
been found except when plasma was used or when 
more than one kind of tissue was present. The 
effect, therefore, either of the activity of these 
other tissues, or of their specific decomposition 
products, has not been excluded. This fact, taken 
in consideration with the instances of abnormal 
differentiation, while it does not demonstrate, 
seems to indicate that the differentiation of a cell 
depends not only on its own inherent properties, 
but also on the nature of its environment. 
Some Recent Discoveries in Paleozoic Vertebrates: 
S. W. WILLIsToN, University of Chicago. 
Some Early Embryonic Stages that Conclusively 
Demonstrate Polyembryonic Development in the 
Armadillo: J. T. Patterson, University of 
Texas. 
The investigation reported in this paper had as 
an object the demonstration of polyembryonic de- 
velopment in the North American armadillo. It 
was shown that from late cleavage stages until 
after the two primary germ layers are laid down 
the differentiation of this egg represents a typical 
mammalian development. Upon becoming attached 
at the animal pole to the placental region of the 
mucosa the vesicle undergoes the process of ‘‘ germ 
layer inversion,’’ producing two secondary, incom- 
plete vesicles, one lying within the other. The 
innermost of these soon becomes completed and 
may be called the ectodermic vesicle; the outer- 
most remains incomplete on the placental side and 
is the entodermic vesicle. The region of the 
trophoblast which is in contact with the mucosa 
forms the Trager, while the distal or free part 
sooner or later sloughs off, leaving the entodermic 
vesicle directly exposed to the uterine cavity. The 
mesoderm arises as two pouches, situated at the 
right and left sides of the ectodermic vesicle. 
These expand and fuse together to form a char- 
SCIENCE 
937 
acteristic extra-embryonic body cavity, which occu- 
pies the space lying between the Trager and the 
proximal side of the ectodermice vesicle. The 
rudiments of the embryos arise from the ecto- 
dermic vesicle, and at first appear as two blunt 
outgrowths (the primary buds), arising from the 
right and left sides of the vesicle. Hach primary 
bud, which lies directly above a mesodermic pouch, 
grows laterally and ventrally along the inner sur- 
face of the entodermic vesicle, and soon bifurcates 
at its distal end to produce two secondary buds, 
each of which represents the rudiment of an em- 
bryo. The secondary buds of the right side pro- 
duce the pair of embryos previously designated 
the ‘‘dorsal’’ (III.) and ‘‘right-lateral’’ (IV.), 
while those of the left side form the ‘‘ventral’’ 
(1.) and ‘‘left-lateral’’ (II.). As a result of the 
failure of the primary buds immediately to com- 
plete division at their proximal ends, the two sec- 
ondary buds of each side remain connected at 
these points. The further differentiation of each 
secondary bud to produce an embryo consists in 
its rapid extension as a finger-like process along 
the inner side of the entoderm towards the Trager, 
with which a placental connection is eventually 
established. The cavity of each secondary bud 
forms the amniotic cavity for the embryo, and the 
four cavities communicate in pairs with the orig- 
inal cavity of the ectodermic vesicle. This latter 
space remains small, and has been named the com- 
mon amniotic cavity. 
From these facts it is concluded that polyem- 
bryony in the armadillo is the result of a preco- 
cious budding, which apparently can not be cor- 
related with the formation of the blastomeres of 
the four-celled stage. 
Factors Controlling the Rate of Regeneration im 
the Frog Tadpole: C. ZELENY, University of 
Tilinois. 
(1) Spermatogenesis in the Gryllde. (2) The 
Puget Sound Marine Station: W. J. Baum- 
GARTNER, University of Kansas. 
(1) The germ cells of the male cricket show 
division figures which do not bear much resem- 
plance to the illustrations shown by Vom Roth. 
There is no circle of four balls as he shows them, 
although he is correct in interpreting the tetrad 
as dividing longitudinally and transversely. The 
number of chromosomes is 23 and is reduced to 12 
and never to 6 as Vom Roth has it. 
One other important observation was made. One 
of the tetrads divides unequally in the first divi- 
sion and the larger half always goes with the 
