548 EErORT OF COMMISSIONER OF FISH AND FISHERIES. [94] 



pigment granules. The pigment-cells of the body are the first to be 

 developed and are the first to become stellate, as shown in Fig. 32, where 

 those on the tail are still approximately round. Up to the time of 

 hatching they are pretty uniformly scattered over the body and less 

 abundantly on the head, as shown in Fig. 34. 



Beyond this stage a marked change in their distribution occurs which 

 cannot be explained without supposing them to possess to a certain 

 extent a power of migration or means of changing their original posi- 

 tion beneath the dermal epithelium. They aggregate in unerring regu- 

 larity in every embryo at about the same places after hatching, as may 

 be seen in Figs. 40, 42, 45, and 49. Two clusters of them are uniformly 

 aggregated on the dorsal and ventral surface of the tail, as shown in 

 Figs. 40 and 45 ntpi. In the first figure they are less densely aggregated 

 than in the last; this is due to a spreading of the pseudopodal prolonga- 

 tions of the pigmented protoplasm composing them. Some of them in 

 the cod now have, when highly magnified, a striking resemblance to a 

 flower, the corolla of which is represented by the radially arranged and 

 flattened black protoplasmic processes of the cell with the clear nucleus 

 in the dark center. On the head region they remain isolated and even 

 without marked processes, as may be seen upon looking at those rep- 

 resented on the brain and jaws in Fig. 49. 



A second and internal layer of pigment cells is developed in the 

 embryo cod. These are confined to the dorsal parietes of the abdominal 

 cavity and seem to underlie the peritoneum. These appear later than 

 those found in the skin, and whether there is any genetic relation 

 between the dermal pigment cells and those of the abdominal cavity 

 would perhaps be hard to say, though it is to be remembered that since 

 we know them to possess migratory powers in other situations, it may 

 not be impossible for the pigment cells of the abdomen to have primarily 

 originated from the same layer as those of the skin. In Fig. 49 they 

 are shown as especially well developed just over the intestine, liver and 

 air-bladder as far forward as the base of the breast fins. 



In the young fourspined stickle-back, Apeltes quadracus, a second 

 kind of pigment cells are developed, forming a row on the median dorsal 

 line and a row on each side of the body. These are of a dirty yellow 

 color, but they develope precisely like the more numerous black ones 

 which surround them and which give the very dark color to the young 

 of this species, which are as dark as young tadpoles, two or three days 

 after hatching. 



The young goldfish (Carassius) has only black pigment cells developed 

 after hatching, no sign of the bright red color being apparent in just- 

 hatched embryos known to have been spawned by red-colored parents. 

 The same remark applies to the young of Idus melanotus, another cypri- 

 noid, in which the skin of the adult is brilliantly colored with red or 

 orange- red pigment. 



To show that light has probably very little to do with inducing the 



