= 
344 Metschnikoff on Germ-Layers. [April 
cism of this hypothesis, which can lead us nowhere, and which 
must therefore be rejected. 
The Gastrza theory, as is well known, has rendered great ser- 
vice in reducing the different phenomena of development to a 
primary invagination; it very often simplifies the complicated 
appearances sometimes seen in the formation of the endoderm, 
for instance, in the Vertebrates. But it is when the theory is 
called upon to explain delamination that it finds itself in the 
midst of difficulties of which Haeckel was aware when he first 
formulated his views on this subject. “The greatest cause for 
doubt,” said he in his monograph on the “ Calcareous Sponges” 
(vol. i. p. 467), “seems to lie in the fact that the gastrula may 
come from the morula by two quite different roads. In the one 
case it arises by a central hollowing out of the morula, the gas- 
tric cavity thus formed breaking through to the exterior. In 
the other case a blastosphere is formed, a hollow sphere whose 
wall consists of a single layer of cells; and the gastrula results 
from a pushing in of one part of this wall, in other words, from 
an invagination.” Haeckel, however, thinks it possible to over- 
come this difficuli by assuming a “ secondary falsification of the 
ontogeny.” In his principal paper (17) he often repeats the asser- 
tion that delamination, in case it SoA occurs in the animal king- 
dom, is a ccenogenetic process, “which has secondarily arisen 
from the palingenetic process of invagination.” As to the way 
in which such a falsification came about there is no explanation 
offered. This is the more to be regretted, as Haeckel himself 
felt the difficulty his theory encountered in this matter. Haeckel 
and his school, the Hertwig brothers in particular, long disputed 
the existence of delamination, but must surely admit it now, since 
‘a member of this very school, O. Hamann (26), has lately ob- 
served the process of delamination in the Hydroids (after it had - 
been described by several previous investigators, among whom 
were Allman, F. E. Schulze, and myself). Hamann, however, 
will recognize no difficulty in this fact, and simply declares the 
| -delaminate planula to be a gastrula which has arisen by cceno- 
geny from an invaginate gastrula. “ Delamination,” states Ha- 
mann (l. c., p. 504), “is in all cases to be derived from invagina- 
tion.” “ In view of the elsewhere universal presence of a gastrula,” 
says he, farther on, “the doctrine, according to nyes a planula is 
but a transformed gastrula, will remain current.” And yet again, 
