[59] THE EVOLUTION OF THE FINS OF FISHES. 1039 



Idonotseethatitnecessarilyfollowsthatthecentiumoftheseventeeutli 

 vertebra of 3Iola consists of two consolidated vertebral bodies, because 

 it has three elements above and three below it, which Putnam, it seems 

 to the writer, erroneouslv regards as neural and htsmal arches. The^' 

 are evidently nothing but iuterspinous pieces, shoved into this strange 

 position by the process of development already described. They appar- 

 ently disappear as soon as the fish becomes adult. We are not told 

 whether they are forked proximally, as neural and haemal arches should 

 be when in contact with so large a centrum. In the adult there is good 

 reason to suspect that these " floating" interspinous elements referred 

 to above have been co-ossified so as to form apparently a single element, 

 if we may place full confidence in Wellenbergh's figure* of the adult 

 skeleton. They seem in Putnam's figure of the skeleton of the young 

 of Mola to be simple bars or rods, separated by an interval from the 

 last " floating vertebra." It is much more probable that they are merely 

 interspinous elements, like the remainder of the chain. They cannot, 

 at any rate, be homologized with the epural and hyi)ural processes of 

 normal Teleosts, which consist, as we have shown elsewhere, of neural, 

 haemal, interneural, interhsemal and basilar interneural and basilar in- 

 terhaemal elements. 



Measurements of the height and length of the principal stages of 

 growth are interesting, and show that the metamorphosis from Osiracion 

 loops to 3Iola proceeds progressively, as shown by the gradual changes 

 in the proportions of these two dimensions. The proportions of the 

 length to height in Osiracion hoops, measuring from between the anal 

 and dorsal forwards and above the eye for the length, and for the 

 height obliquely across the eye, are in — 



Fig. 1, length to height as 1 to 1^. 



Fig. 2, Molacanthus, length to height as 1 to 1^. 



Fig. 3, Molacanthus, length to height as 1 to 1|. 



Fig. 4, 3Iola, length to height as If to 1. 



Fig. 5, Mola, length to height as IJ to 1. 

 This will conclude the arguments which we have presented in favor 

 of regarding Molacanthus as merely the young of Mola. What dif- 

 ferences the corresponding stages of growth of such a form as Ranzania 

 or Masturus might present in contrast with the known stages appertain- 

 ing to Mola, it is impossible to say in the absence of sufiicient material 

 for comparison ; but it is safe to assert that no amount of additional 

 evidence will be at all likely to break down the reasons here given for 

 the incorporation of Molacanthus nummularis (Walb.) Gill, with Mola 

 rotunda, Guv., as one and the same species, for the reason that it is per- 

 fectly sale to predict that when such additional evidence is forthcoming 

 it will probably confirm the position which has been here assumed. 



*Observ. AnatomicBe do Orthagorisco Mola. Diss. Inang. P. H. J. Wellenbergh. 

 Lugdnni-Batavorum, MDCCCXL. 



