﻿65 103 



Jordan and Starks, the dia}<nose,s slate thai llie anterior verlebræ are elonf^atcd, hut nothing 

 is added to what has been observed by the earlier, above-named authors. 



The fact alone, that ribs occur and that the anterior vertebrae arc not elongated in the 

 fossil genus Rhamphosiis Ag. shows that this cannot be nearly related to (U'litrisctis. This has 

 been generally accepted hitherto (cf. Smith Woodward (36, p. 377)), since Blainville brought 

 the oldest known specimen even into the genus Centrisciis as C. aciileatus. Agassiz retained 

 this alliance, but believed that there were "differences assez marquantes pour constituer un 

 petit genre à part" (1 b, p. 271). On reading through Agassiz' description of the, al that time, 

 only species Rh. aciileatus, we very soon see, however, that the resemblances to be found 

 with (A'litrisciis are on the whole quite sui)erlicial. The long dorsal spine, for example, 

 sliows quite different relations to the skeleton; it seems to be placed far forward just behind 

 the head and is not followed by other "spines"; the snout has nothing like the characteristic 

 lube-form with terminal mouth; the mouth lies in under a prolonged snout ("Le museau est 

 très-saillant, en forme de rostre dépassant de beaucoup les mâchoires; celles-ci s'ouvrent peu 

 et sont placées immédiatement au-dessous l'orbite"; p. 270); the ventral fins are large and are 

 placed on the thorax etc. I may add to this that, according to Agassiz' figure (PI. 32, fig. 7), 

 the rays in the dorsal, anal and caudal lins are branched or divided; the same is the case 

 in the species Rh. biserratiis, later described by Bassani (2). Every trace of the ventral armour 

 is absent and the external bony plates connected with the large, postoccipilal spine have not 

 the least resemblance to the dorsal armour in Centriscus. 



It is staled, certainly, by Vaillant (33a, p. 127; 33b, p. 339) that he had found quite similar 

 small scales in Rh. aciileatus to those in Centriscus, and ~ if I understand him rightly — he 

 is not disinclined to make one genus of those two; he writes: "II me ])araît donc hors de 

 doute que dans ce genre fossile, si tant est qu'il doive être conservé, la structure des écailles 

 était la même que dans le genre actuellement existant ". 



In spite of this and though I have not had the opportunity to examine specimens of 

 this form personally, I venture to say, that Ramphosus cannot be related to (k-ntriseiis, and 

 indeed that it can by no means be placed anywhere within the group of families, which I 

 have provisionally called "Hemibranchii" + "Lophobranchii " in the Introduction to this 

 communication. 

 18; p. 87 (49): 



Regarding the skeleton of the head the earlier literature gives us just as little as for 

 Amphisilc. Rosenthal has given the only figure known to me of the skeleton in Ceiilrisciis; 

 bul il is practically useless (I.e. PI. X, fig. 11); the few statements in the explanation to the 

 figure ipp. 36, 37) only serve to show that he has understood very little of the structure of 

 the head. Nor do Agassiz and Günther give anything more than what is superficially quite 

 obvious; Günther rightly remarks, however, that "The interoperculum is extremely narrow 

 and elongate". Recently Siebenrock has ligured the posterior portion of the skull in (k'ntriscus 

 and remarks that the parietals are wanting (29, p. 131); and Starks (30, p. 624) notes the same 

 thing, as also that the opisthotic is wanting, that the articulating surface on the basioccipital 

 is concave (in contrast to Aulostomida; as Brühl however had already remarked), that the 

 uppermost portion of the pectoral girdle, the posttemporal, is suturally connected with the 

 cranium, and that there is a well-developed "myodome". His statements "pterotic normal in 

 position" and "basisphenoid small" are however incorrect. 



Cope was the first, so far as known to me, to give information regarding the branchial 

 arches in Centriscus (6, p. 457), namely: "Fourth superior branchihyal and first and fourth 

 superior pharyngeals only wanting ". This is however quite wrong. It is rei)eated nevertheless 

 by Gill (12a, i)]). 156 and 163), Jordan and Hvermann (19, p. 742) anil Jordan and Starks (20, j). 68), 

 and by Goode and Bean (13, p. 483). 



D. K. U. VUlensk.Selsk. Ski-.. 1. Ra-kkc, naluivhleiisk. i)(i iiKithcni AW. VI. 2. 14 



