8 PROF. G. B. HOWES ON THE SKELETON AND [Jan. 18, 



Fig. 2 represents the ventral aspect of the pair of fins afore 

 named, as they lay in life. They were attached to the pelvic 

 cartilage (pi.) by a fibrous buffer, identical with that described by 

 Davidoff (7, p. 124). The free end of the hip-girdle terminated 

 in front in a pointed extremity {processus impar of Davidoff), which, 

 as already observed by GUnther (14, p. 535) and that author (7, 

 p. 124), was bent towards the left side. I figure this (fig. 2 a), as 

 its distortion is here much more marked than in any specimen yet 

 drawn. 



According to Schneider (23, p. 521) "Die Curve, welche der 

 dorsale Rand jeder Flosse beschreibt, ist verschieden von der 

 Curve des ventralen Randes. Nun ist der dorsale Rand der einen 

 Flosse congruent mit dem ventralen Rande der anderen." In the 

 specimen here figured, the two fins were sickle-shaped ; the inner 

 half of the preaxial border of the left one was straight, as repre- 

 sented in the figure. It will be observed that as they lay flattened 

 out, their free ends were both directed towards the animal's right 

 side ; the excavated border, which imparts to the rin-lobe its sickle- 

 shape, was preaxial for the right fin, postaxial for the left. When 

 applied to the sides of the body, the apex of the former looked 

 dorsally, that of the latter ventrally. Tlie contour of the Ceratodus 

 fin is variable ; occasionally its opposite margins are symm.etrical 

 with respect to the axis ; but the differences in symmetry between 

 these two fins more than cover those which I have observed between 

 any two members at my disposal. Turning now to the supporting 

 skeleton, it will be seen that the second mesomere bears, as Schneider 

 has pointed out, an accessory lobe (his "anderes Stiick " referred to 

 above). That, however, instead of being symmetrical, as he claims 

 it to be, is, in this specimen, unsyui metrical to the utmost — for the 

 right fin it is postaxial, for the left one preaxial. Further comment 

 is needless, as the drawing which I give speaks for itself. Thus far 

 the characters oi the pelvic fin, as defined by Schneider, are seen to 

 be inconstant and untenable : more than that, however ; for, in that 

 the preaxial lobe of the one fin corresponds almost to a degree 

 (with the exception of one feature, to which I shall return) with the 

 postaxial lobe of its fellow and vice versa, there are embodied in the 

 two the more important differences held by him to exist between 

 the pectoral and pelvic members. 



Schneider goes on to say (p. 523), " wenn man die symmetrische 

 Stellung der vorderen und hinteren Flosse in Betracht zieht, so 

 leuchtet die Aehnhchkeit des ersten Gliedes des Hauptstrahls mit 

 Humerus und Femur des zweiten Gliedes des Hauptstrahls mit 

 Ulna-Radius und Tibia-Fibula ein." I have shown above that the 

 characters of this "zweites Glied " are inconstant for the pelvic fin. 

 Its accessory lobe is present on that side on which the parameres are 

 stoutest, be it preaxial or postaxial ; and examination of the 

 specimen under my hand suggests unmistakably that it has arisen as 

 the result of coalescence between the second mesomere and the 

 confluent bases of the two proximal parameres. The well-known 

 lobe of the pectoral member {cf. figs. 5 and 6, mt,), first accurately 



