LARViE, ETC., OF OPHIOCEPHALUS STRIATUS. 117 
General Considerations. 
So far as I am aware no other instances of eggs of freshwater 
fishes floating at the surface of the water by their own buoyancy 
have been described hitherto. The same advantages, namely, 
direct proximity to atmospheric air and to sunlight, are partly 
secured in other ways, as by attachment to aquatic plants or by 
deposition in very shallow water. The floating eggs of Ophioce- 
phalus striatus , with their accompaniment of small fragments 
detached from the surrounding plants, such as cut pieces of tank 
weed, leaves, and lengths of cut stem, rather invite comparison with 
the floating nests of Gymnarchus which were described by the late 
J. S. Budgett.* He found these nests amidst the dense grasses of 
a West African swamp floating at the surface in three to four feet of 
water ; the deepest part of a nest was only about six inches below 
the surface ; in it “ were deposited about a thousand large spherical 
amber-like eggs 10 mm. in diameter. The eggs hatched five days 
after being laid, and in eighteen days a thousand young fry of 
Gymnarchus niloticus left the nest,” having in this short time attained 
a length of three inches. 
In the same flooded grass-lands Budgett frequently found “ masses 
of white foam floating on the surface of the water.” These proved 
to be the foam-nests of another fish, Sarcodaces , a member of a 
family, the Characinidse, which is not represented in Ceylon. Thej/ 
were filled with numerous transparent ova, 2\ mm. in diameter ; on 
hatching, the larvae “ make their way through the foam in which 
they are laid, down to the surface of the water, and there the young 
larvae hang, holding to the surface of the water by a large adhesive 
organ situated on the front of the head.” There is no adhesive organ 
in the young larvae of Ophiocephalus striatus, but all the same they 
are kept at the surface by the extraordinary buoyancy of the yolk- 
sac, and in due time they strike out from the surface to the bottom ; 
whereas the hatchlings of most freshwater fishes strike out from 
the bottom towards the surface. 
Thus the larva of Ophiocephalus differs from that of Sarcodaces in 
the absence of a frontal cement organ, and from that of Gymnarchus 
in the absence of external gills. Like the Mormyridse, the family 
to which Gymnarchus belongs, and the Characinidse, the Ophioce- 
phalidse is a characteristic family of tropical freshwater fishes whose 
morphology is better known than their bionomics ; and the present 
contribution, with others which are to follow, will show how much 
remains to be done upon the subject which was so successfully and 
tragically inaugurated by Budgett. 
* J. S. Budgett. On the breeding habits of some West African fishes. 
Trans. Zool. Soc. London, XVI., 1901. Reprinted in the Budgett Memorial 
Volume, edited by J. Graham Kerr, Cambridge Univ. Press, 1907, pp. 119-30, 
plates and text-figures. 
