ON THE ZOOLOGICAL STATION AT NAPLES. 455 
This is the only Teleostean, so far as I am aware, in which the true 
nature of the third segmentation process can appear. It is masked, for 
instance, even in G, paganellus, though this latter species is very instruc- 
tive in the light of Gobius b. as affording a passage from the regular or 
amphibian type of segmentation visible in this last to the types that 
obtain in other Teleosteans. If we observe a number of eggs of Gobius 
paganellus that have undergone their third segmentation we can find 
some that resemble these of Gobius b. but for the fact that the four 
protoplasmic swellings on the surface of the yolk that represent the four 
lower cells have been, as it were, forced apart, two to one side and two to 
the other of the square of four cells that form the upper half of the egg ; 
and, on the other hand, other eggs will show us the eight divisions 
arranged in the two rows of four, which seems the most common form 
with Teleosteans of this stage. 
Much ingenuity has been expended in explaining the third segmen- 
tation process in individual species of Teleosteans. However, a comparison 
of what I have been able to see for myself and what I have read makes 
me think that these early segmentation processes in Teleosteans can be 
best and most consistently explained as due to a masking of the simple 
regular geometrical plan of, for instance, Amphioxus and the frog by the 
presence of a greater or less amount of food yolk, and that all attempts to 
explain the plane of the third segmentation more definitely than by say- 
ing that it is trying to be horizontal, and to separate four less yolky cells 
from the other yolk-laden and imperfectly divided part of the egg, will be 
unsatisfactory, as wanting generality and only explaining what occurs in 
particular cases. 
After I had studied the segmentation in these species of Gobius I dis- 
covered Rauber’s paper (‘Neue Grundlegungen zur Kenntniss der Zelle,’ 
Morph. Jahrb. 1883), part of which happens to be devoted to a compari- 
son of the early segmentation stages in ana and a species of Gobius. As, 
however, Rauber was not fortunate enough to examine the species of the 
latter that fell to my lot, he has, as I think, missed the real and simple solu- 
tion that they might have indicated, viz. that here, as in so many other 
cases in comparative embryology, it is the different amount of yolk in the 
eggs which causes processes that are really essentially the same to appear 
so very different. 
I have been able during the last month to examine the complicated 
structure of the floating egg-case of Mierasfer acus, briefly described at 
p. 68 of the monograph on this genus in the Naples ‘Fauna und Flora.’ 
Shortly, this consists of an aggregate of hexagonal tubes, like elongated 
bee cells, open at both ends and grouped symmetrically side by side, so 
as to form an oval hollowed out at one side, on which and on the opposite 
side the tubes open. Tach of these prismatic cells has a number of eggs 
attached to its internal walls by short stalks. To make out the whole 
structure it is best both to examine in the natural state and to coagulate 
the egg-case with alcohol. Corrosive sublimate causes the whole case 
to disappear, and Perenyi’s fluid is not good. The egg-case seems also to 
disappear shortly before hatching. I hope to be able to examine the egg- 
case of Scorpenw more carefully. This is considerably larger than that 
of Fierasfer, and appears to consist of a single large sac to the inner wall 
of which the eggs are attached, but Ihave not yet been able to verify this 
from a hardened preparation. It would apparently be interesting to make 
a comparative study of the various modes in which Teleostean eggs are 
