SPECIES NO. 25. 45 



longer ones spreading fan-like towards the dorsal fold ; the third encircles 

 the body, its anas radiating downward from a dorsal centre, and upward 

 from two lateral centres (the lateral centre is seen in Fig. 9 before expan- 

 sion) ; the fourth (still in a contracted state in Fig. 9) also encircles the 

 muscular axis of the tail. A few yellow pigment-cells are now seen on the 

 yolk-sac, as well as on the oil-globule. Granular cells are also seen over 

 the whole surface of the fish, a few of which are indicated on the median 

 fold of Fig. 9, and more extensively on Fig. 1(1. The yellow spots seen in 

 Fig. 9 contract, after passing through this stage of expansion, and again (at 

 50 hours) appear in the form of clusters of rounded cells. 



In the next stage (Fig. 10 is about seventy-four hours old) we find the 

 yellow spots much reduced, and apparently about to disappear altogether. 

 A line of black chromatophores has developed along the ventral side of the 

 tail, and round the eye. 



Species No. 25. 



Plate XIX. Figs. 1-7. 



The stages here described were all obtained in the early part of July 

 (1884). The species could not be determined ; but the stages cover a very 

 interesting period in the development of the pigment, extending from the 

 time of closing of the blastopore until the young fish is four days old. 



The largest number of eggs found at any one time was six. The egg 

 varies in size from .SI to .90 nun., and the oil-globule from .22 to .23 mm. 

 The globule is proportionally larger than that of any other egg described in 

 this memoir, and its absolute size is exceeded only by tin' globule of the 

 egg of Lophius and of No. 10. Two or three oil-globules were of frequent 

 occurrence among the earlier stages; but they invariably coalesced in the 

 interval of eighteen hours, which separates the stages of Figs. 1 and 2, 

 forming a single globule of the ordinary size. The three globules repre- 

 sented in Fig. 1 coalesced while the drawing was making. The two on 

 the left side coalesced first. The act was not seen, but it must have been 

 quite sudden, as it occurred in the course of an interval of live minutes. 



In the early stage of Fig. 1, just after the closure of the blastopore, 

 we find two kinds of pigment-cells already fully developed in the embryo. 

 Not a single pigment-cell is to be seen on the yolk-sac. in any of the stages 

 before hatching; but the globules are well covered with chromatophores 

 in all stages. The brownish yellow pigment-cells of the body are small, 



