BJOLOQY. 
Tun . 5 
of Amphioxus and the Vertebrates, is derived from both primary layers, 
ectoderm and endoderm. 
Castle found that Ciona produces both ova and sperms at the same 
time, but that self-fertilization very rarely occurs. The eggs are laid just 
before dawn, and the larva is hatched during the following night. The 
test-cells adhering to the young homogeneous test have, it is now well 
known, no connection with the cells found later on in the adult test. The 
larvae are free-swimming for from one to several days. They avoid the 
light. 
The sperm archoplasm is alone concerned in bringing the pronuclei 
together. It divides into two attraction spheres, and forms the first cleavage 
spindle. Hence, as Castle shows, the archoplasm cannot be a bearer of 
heredity. The polar globules invariably form at the dorsal pole. The 
spermatozoon enters at the ventral hemisphere, and that point determines 
the median plane and the posterior end of the embryo. The ventral is 
the animal pole. The cleavage is from the beginning bilateral. The first 
cleavage plane is vertical, and divides, the right and left halves of the 
embryo. The four smaller dorsal cells with yolk give rise to the endodermal 
hemisphere, the four larger, more protoplasmic cells form the ventral 
ectodermal hemisphere. The cells of the latter hemisphere divide more 
rapidly, and form the future aboral surface. When the dorsal hemisphere 
has twenty- two cells, the ventral has fifty-four. 
The gaslrulation is a combination of epiboly and invagination ; the 
ventral ectoderm grows over, so as to envelop the dorsal hemisphere, 
while the latter sinks down, and becomes saucer-shaped. In the centre 
of the dorsal surface ten cells form the future endoderm. Round these 
comes a ring of cells, the chorda-mesenchyme ring, from which the noto- 
chord and mesenchyme arise. Outside this ring is a row of cells, the 
neuro-muscular ring. The more anterior of these cells form the medul- 
lary plate, the more posterior the longitudinal musculature of the larva. 
The remainder of the cells (in the 112-cel.l stage) form ectoderm. By 
growth at the anterior end the blastopore gets pushed posteriorly, 
and the anterior chorda cells are covered up, and come to lie in the dorsal 
wall of the archenteron, sixteen cells in two rows, one over the other. 
The blastopore closes in the posterior part of the dorsal surface. In 
front of it is the medullary plate, with a continuation back at the sides of 
the blastopore. This region forms the trunk of the larva, the part 
posterior to it being drawn out to form the tail. The chorda cells pass 
back into the tail, while the mesenchyme cells shift forwards into 
the trunk. The muscle cells, derived from the neuro-muscular 
ring, lie behind the blastopore, and form the muscles of the tail. 
The closure of the medullary canal takes place from the blastopore 
forwards, and then the nerve cord is grown over by ectoderm. After 
closure of the blastopore the mesenchyme cells lie as lateral masses in 
the trunk ; later on they become the blood corpuscles and the mantle 
cells, &c. 
Castle also discusses some important theoretical questions. He points 
out that, in Ciona at least, the chorda-mesenchyme ring takes part along 
with the endoderm in the primary invagination, and so belongs to the 
primary endoderm ; while the rest of the mesoderm, the muscle cells of 
the neuro-muscular ring, are carried in by a secondary invagination, and 
belong to the outer layer of the young gastrula, or primary ectoderm. 
He considers that the chorda must be regarded as a mesodermal organ. 
He agrees with former observers in seeing no trace of enteroccele forma- 
tion, and he doubts whether any Chordata are Enterocoela. He does not 
believe in distinguishing those Metozoa with a “mesoderm” from those 
with a “ meseuchyme.” He considers that embryology gives no support 
to the Annelid hypothesis as to the origin of Chordates. 
Ritter gives a full account, well illustrated, of the process of budding, 
