592 Mi(. n. II. wTiiTKiiousE ox THE [Apr. n, 



The most, conipi-eheiisive work of all, however, wa.'^ pulili.sheil 

 by Ryder in 1886 ; on tlie whole, it contained very useful infor- 

 mation, but liis liirure of the tail of the young eel i.s wrong with 

 i-espect to the neural arches. Ily(lei''s suggestions on the origin 

 of heterocei'cy are interesting and probably correct, though some- 

 what novel. Ryder's evolutionary series of the forms of the 

 caudal fin lias come to be acknowledged as the order of appearance 

 of the different types in time. The next work of importance 

 dealing witli caudal fins of fishes was written in 1895 by J)ollo in 

 di.scu.^sing the phylogeny of the l)i))uoi : basing his conclusions on 

 geological evidences, he gives the trnn gejihyrocercy an extra- 

 onlinarily wide application ; this aii.ses fi-oni the id(\-i, sti-ongly 

 upheld by Dollo, that the present cauilal fin in many Teleosls 

 is really a posterior anal fin, brought into its present position 

 by complete atrophy of the real caudal. Dollo's is the last of 

 importiint refei'ences regarding the tail-fin. 



Boulenger, in 1901, in dealing with the fishes of the Congo, 

 gave a good summary of woi'k done. In 1907, (iregoiy, in his 

 work on the classification of Teleostomous fishes, makes continual 

 references to the caudal fin ; but it is difiicnlt to reconcile many 

 of his statements : for instance, the caudal fin of MciHtacevihelvs 

 is referred to as gephyrocercal, while in fact internally it is (juite 

 a generalized homocercial type. As an appendix to the woi-k, 

 there was given an elaborate scheme of the evolution of the 

 caudal fin, W'hich includes several new terms, consitlered by some 

 as unnecessary. 



2. Termixolooy. 



In order to avoid auN'andiiguity, I have thought it necessary to 

 modify the meaning of sevei-al terms ; some have lieen widened 

 and others have been more restricted as regards their denotation. 

 The reasons foi' such modifications are given in their respective 

 paragraphs : — 



1. Urostijle. — This term is used in the generally accepted sense 



denoting the more or less elongated cone-like termination ol 

 the vertebral column, appearing as a prolongation of the last 

 centrum. It represents tiie fusion of several centra. 



2. JTypaxial. — This adjective refers to any structure morpho- 



logically ventral to the cliordal axis. 



;>. Kjxixlnl. — This adjective refers to any .structui-e morphdlogic-a Ily 

 dorsal to the cliordal axis. 



4. Ihipiiral. — Any hypaxial element having a dirci'l coniuH-tion 

 with the choi-dal axis, and bearing one or more iin-iays 

 distally, will be called a hypural or hypural Imne. Previously, 

 this term lias been useil rather loo.sely and in such a way as 

 to leave one in d<)td)tas to its limitition ; the limitation given 

 above, concerning its direct connection witii the chortlal 

 axis—?*.'', as a general rule, with centra — .as will b(> seen 

 later, botli widens and narvows tlie nie;uiing of the li'rm, 



