150 
renals are found only in those forms in which a reduction in number 
of ripe sexual elements required has taken place. 
In the present state of our knowledge of these organs I do not 
wish to lay too much stress on the point of existence or non-existence 
of suprarenal bodies in any‘ order of Vertebrates. The matter has 
been mentioned for two reasons. Firstly, in its bearings on the theories 
of RaBL and others as to a supposed repeated loss and gain of food- 
yolk. For, if Ragr’s „tree“ be correct in respect of the ancestry of 
Ganoids from sharks — a line of descent which has also the high 
authority of Prof. GEGENBAUR on its side — I can conceive that the 
non-existence of suprarenals in the former group would be a fact 
which our author might find difficulty in accounting for — unless he 
ignored it entirely! Secondly, if the statements I am about to make 
as to the interrelationships of the sub-classes of fishes be sound, the 
presence of suprarenals in only certain orders of the Vertebrata ought, 
instead of giving rise to difficulties, to be rendered obvious and rea- 
sonable. TheMarsipobranchii, the Ganoidei, the Teleostei 
and the Dipnoi!) produce a large number of ova as compared with 
the Selachii. Their egg membranes?) also difter considerably in 
kind from those of the latter group. That is to say, the horny shell 
of Selachians has no parallel in any of the other Ichthyopsida. 
I must guard against being assumed to maintain any near rela- 
tionship between Dipnoi and Ganoidei, as many and distinguish- 
ed morphologists have done’). Indeed, I read the facts as demon- 
strating a nearer relationship of Dipnoi and Selachii than of the 
former and the Ganoidei. How near the sequel will show. 
Those points of morphology, which, in my opinion, are of special 
importance, concern the urinogenital organs, especially the pronephros, 
and the nature of the fore-brain. I will not deny that there are other 
matters, such as the division or non-division of the lateral muscles, 
but these need only be considered here as confirmatory of other con- 
clusions, or as demonstrating further subdivisions. 
I will briefly indicate: (1) the two great divisions into which the 
fishes naturally fall, (2) the subdivisions of these, and then proceed to 
1) I can deduce this fact for Protopterus from the condition of the 
ovary — full of numerous ripe ova — in many females I have examined. 
2) See on this question, for the Marsipobranchii, SHIPLEY, 
Quart. Journ. of Microsc. Sci. 1887, for the Ganoidei, BeraArn, Proc. 
Roy. Soc. Lond. 1889, and for the Dipnoi, BeEpparp, Zool. Anz. 1886. 
3) GEGENBAUR, GÜNTHER and others. 
