1898.] BEEEDTBTG OF THE DEAGONET. 291 



male's pelvic fin and so losing her hold on his person if not on his 

 affections. As the tank is only four feet deep the pair naturally 

 reach the surface soon, though the upward progress is very 

 slow, since the male, who has most of the work to do, is much 

 hampered by the impossibility of using the tail, his principal 

 organ of locomotion when unhampered by female society. Arrived 

 at the surface the pair occasionally come apart, but more usually 

 continue together, their snouts bobbing in and out of the water, 

 while they wander vaguely about, still endeavouring to ascend '. 

 In process of time they lose hold of each other and dart rapidly 

 to the bottom. Sometimes the male, more rarely the female, 

 seems unaware of his or her loss and continues to cruise f iitilely 

 at the surface. Occasionally one or the other will ascend to the 

 surface alone, so that it is possible that the female assists in the 

 ascent (when the two are together), though I certainly think that 

 she is more concerned in sticking to her partner. The superior size 

 of the male is no doubt of importance, 



I cannot positively say that I saw ova extruded. They are 

 very small and practically transparent, and difficult enough to see 

 even under the most favourable circumstances in a tank. It is 

 my impression that I saw them on one occasion. Milt was not 

 extruded by the male in visible quantities, but the milky fluid 

 common to many fishes is not always found in connection with 

 ripe spermatozoa. In any ease ova were extruded and fertilized, 

 since they appeared in a net fixed on the overflow port of the 

 tank, and duly hatched out in the jars in which they were placed. 

 Evidently the limited depth of the tank is not, as Savile Kent 

 supposed, a bar to the successful accomplishment of the matri- 

 monial enterprise. 



I was able to ascertain that the female takes sometimes the 

 right, sometimes the left pelvic of the male, but whether by 

 accident or design I cannot say. In the ripe condition the ovaries 

 cause a very conspicuous bulge on either side of the posterior part 

 of the abdomen, and it may be that the close apposition of the 

 sides of the male and female assists the latter in the extrusion 

 of the products of whichever ovary is thus subjected to pressure ^. 



The first period of reproductive activity lasted, as we have seen, 

 for eight days, commencing, as first observed, on the 11th February. 

 The activity was greatest for the first few days, and the full 

 splendour of the male was only attained during about three or 

 four days. Only one female was engaged. Pairing was observed 

 to take place from 9 a.m. to about 11 a.m. or noon. After this 

 the males ceased to sport and usually buried themselves in the 



^ There is no evidence to show how far the ascent is continued under natural 

 circumstances, but a male has been recorded by Matthias Dunn (in Day's Fish. 

 Gt. Brit. i. p. 176) from a mackerel-net at the surface of 40 fathoms in May. 

 Some examples are still breeding in this month. The species has been taken at 

 a maximum depth of 218 fathoms. 



- In Mallotus villostis, a species propagating by demersal eggs, " two males, 

 one on each side, hold the female, while she rushes with great swiftness on the 

 Bandy beach and there deposits her spawn " (' Descent of Man,' p. 331). 



19* 



