1887] Metschnikoff on Germ-Layers. 343 
formed by transverse division and in part by immigration. These 
methods of forming the endoderm are not all sharply separated ;. 
on the contrary, transitions exist between some of them. The 
unipolar type appears either (a) as an immigration of the blastula 
(or blastosphere) cells from the hinder end of the larva, or (6) as ` 
an invagination." 
The question is now, What is the stand-point from which a 
comprehensive and intelligible view may be got of these various 
methods of forming the endoderm, and which of the existing 
theories on the origin of the primitive organs is best able to ex- 
plain the facts? It is at once seen that on assuming the descent 
of the Metazoa from multinucleate Protozoa (Infusoria, or perhaps 
Heliozoa and Radiolaria), we become entirely unable to explain 
either immigration from the surface, or primary delamination, or 
invagination. It is unnecessary for me to go into a detailed criti- 
t Multipolar immigration is illustrated by Fig. 3. From various points of the 
blastula cells migrate into the interior; and, their number increasing, they here form 
a solid mass of endoderm, which is subsequently hollowed out to form the digestive 
e inner ends of the blastula cells are constricted off to form the endoderm 
They very soon arrange themselves in the shape of a hollow sphere, the cavity of 
which becomes the permanent draive cavity. Secondary delamination is the rule 
among Hydroids without a free medusa i It is essentially like mixed delami- 
nation, Fig. 1, but differs from it in the very late appearance of any histological dif- 
ference between the superficial layer of cells and the included or endoderm cells. 
Neither in this type nor in mixed delamination is there an obvious blastosphere, the 
cell immigration and transverse division commencing at a very early date. In both 
use. The 
doderm, which afterwards is hollowed out (Atlas, Plates II., III., IV.). Invagina- 
ros has hitherto only been found in the Acraspeda (Nausithoe, Pelagia, Plate X., 
tlas).—H. V. W. 
