252 MR. A. K. TOTTON ON 
The caudal skeleton of an adult specimen of Plewragramma 
antarcticum consists of a number of relatively large cylindrical 
-bony centra of a papery nature. To the terminal one are 
attached two expanded hypural bones, whose development proves 
that they are compound (i. e. that they represent more than two), 
and a third ventral hypural which is not so expanded. Between 
the latter and the ventralmost of the former two hypurals, the 
terminal branches of the caudal artery and vein come forth on 
each side ina notch, This terminal centrum bears no arches, but 
lying dorsal to it and the nerve-cord (not represented in PI. IT. 
fig. 12) are two bones, the epiural apophyses (cf. Huxley, Q. J. 
M.S. vol. vu. p. 41). Development shows that there were 
originally three, and there are even indications of a fourth. The 
penultimate centrum at first sight appears to be carrying two 
neural and two hemal arches, but in the case of the neural arches, 
at any rate, development shows clearly that they have arisen 
through the division of one cartilaginous “ Anlage.” 
I should like to make a digression here to discuss at greater 
length this phenomenon of a centrum bearing two neural or 
hemal arches. I have collected a few cases from various authors, 
and have observed the phenomenon also in a few specimens in 
which it has not been described, so far as I am aware. 
Mormyrus kannume has two hypurals on the penultimate 
centrum (¢f. Whitehouse, “Caudal Fin of Teleostomi,” P. Z.8. 
1910, p. 596). I am not aware that the development of this 
fish is known. It would be interesting to see whether the adult 
condition was brought about by a division of a single cartilaginous 
‘“‘ Anlage” or by a crowding together and displacement of arches, 
as is more probable. 
Lotz (Zeitschr. f. wiss. Zool. Bd. xiv. 1864, p. 88) gives the 
following description of Cotéus gobio :—* Der vorletzte Wirbel 
hat nur das Eigenthumliche, dass sein oberer Dorn sehr breit 
und oben gespalten ist, so dass er aus zwei verschmolzenen 
Dornen zu bestehen scheint.” The figure shows a much expanded 
neural and a hemal arch on the penultimate centrum, and the 
neural arch shows a line of division into two. I think this is a 
clear case of ‘ division.” Moreover, ‘ division”? seems to be 
connected in some way with expansion, as comparison with other 
cases will show. He says of Barbus fluviatilis :— Der drittletzte 
Wirbel besitzt in der Regel zwei obere Bogen deren Dornen 
bereits als Stiitzen der kurzen Flossenstrahlen dienen, diess ist 
iibrigens nicht constant; es kann auch nur ein oberer Bogen 
vorhanden sein oder es kann der vorletzte Wirbel deren zwei 
besitzen, wihrend er in der Regel nur einen besitzt, dessen Dorn 
mit einem knorpeligen Ende versehen ist” (p. 86). On page 95 
he says of Salmo salar :—‘“‘ Am. 17-18. Tag. wird tiber dem 
Neural-canal hinter dem letzten normalen oberen Bogen (g) der 
erste falsche Dorn (4) gebildet.” The ‘‘falschen Dornen” are 
the epiural apophyses. It is to be noted that the neural and 
heemal arches have already been formed. Lotz continues: ‘“ Oft 
