Miscellaneous. 173 



verses it as far as the origin of the filaments ; then she bends into 

 a semicircle. The male fish, by a like inflexion of his bodj', em- 

 braces her, turns her over, and jjresses her side, an operation the 

 result of which is a first emission of ova. These, from their light- 

 ness, tend of themselves to rise ; but, with a foresight which cannot 

 be too much admired, the male in pressing the female forms, by 

 meaus of his dorsal fin, a concave fold, a receptacle in which the 

 ova undergo the contact of the fecundating principles. Soon after, 

 there is a new visit of the female, and a fresh approximation of the 

 male, until the ovaries are completely evacuated. 



The spawning over, the female quits the conjugal roof, leaving to 

 the male the care of rearing the family, a task of which he acquits 

 himself with a truly paternal zeal. Collecting with his mouth the 

 ova scattered through the plant, he raises them into the nest and 

 arranges them in orderly fashion ; if they are too much agglomerated, 

 he separates them by a movement of the head and compels them to 

 remain in the same plane ; then he issues from the nest, and seta 

 himself with great activity to contract the entrance. When this 

 operation is completed, he goes away and swims round his edifice 

 to examine the whole, — and not without auxiety ; for he often goes to 

 fetch fresh bubbles of air, which he places intentionally under doubt- 

 ful points or under menaced parts. 



After seventy hours of incubation the male, foreseeing that the 

 ova require fresh care and quite a different medium, ascends in the 

 nest and pierces its summit ; the air-bubbles escape and the dome 

 immediately flattens upon the water, imprisoning all the embryos,, 

 the existence of which begins to be manifest. 



Eearing lest the young should escape his care, he sets to work to 

 make a new barrier for them. For this purpose he follows and tra- 

 verses the outer margin of the floating carpet, and pulling at it with 

 force, separates it from the felt, thus obtaining a sort of pendent 

 fringe where stray young ones will not be able to j)ass ; then, having 

 got rid of all anxiety fi'om this side, he takes his young in his mouth 

 and removes them to short distances, always conveying those of the 

 circumference towards the centre. 



If some of the young fish venture to descend vertically, he goes 

 in search of them, and carries them back to the protective dwelling. 

 This surveillance lasts until the embryos, having undergone their 

 complete evolution, have acquired strength and agility. Their nume- 

 rous and frequent flights announce to the male the end of his troubles, 

 which comes about eight or ten days after the sinking of the nest. 



The same pair of fish gave me three ovipositions during the 

 summer of 1875, each consisting on the average of 150 ova. 



The embryos of the rainbow CoUsa undergo a series of trans- 

 formations analogous to those which I first indicated as occurring 

 in Macropocla chinensis. Want of time and the dread of aft'ecting 

 the existence of animals which are still rare prevented my following 

 this investigation with all the attention that the subject deserves ; 

 but I propose to resume it hereafter. 



All my observations on the Indian CoUsa were made in Paris, in 



