290 



./. K. T/iiichn' — Median a ml I'dircd hlnx. 



Amil of Eugompliodua Ulorulis, 1*1. L\'l and LVIl, (i<;'s. 17-5(1. 



The number of rays is 21 in figs. 47 iiiul -49. In 18 \\v liavo 'I'l, and in 50, 'JO rays. 

 We may take then 21 as the normal nunilicr. 



Extra sogmoiitations amount to 1. 



ConcTOSoonco is estimated at ■](). Hetippin^- amounts to •:{. 



Ratio of proximal to middle piooo of middle ray i. 



Ratio of distal to middle pieee of middle ray 5. 



Wo may sum up the results of this investigation of the reseiul)lani'es and dillercnces 

 of the forms so far examined in the following table. 



No. of Kxtra ( ion- 



rays. seKliU'iitatioiiH. (■iH'HC'ciicc. Brli|i|iiii; 



Katio of Hatlo ot 

 |.ro\. to (Uwtal to iiild. 



mid. i>ic('(^ pk'co of mill. 



)riiii(l.rHV. rav. 



First 

 Dorsal. 



Second 

 Dorsal. 



Anal. 



i'lugomphodus . . 17 



Mnstelus 21 



(lalooeerdo 25 



K\damia 29 



Spliyrna :M 



i^yugomphodua _ . 17 



Mnstelus 21 



(lalooeerdo 1 :i 



Eulaiuia II 



Sphyrna -II 



' Mugomphodus ..21 



Muslelus 18 



Oaleoeordo 12 



Mnlamia 18 



Spliyrna 27 



Tlie eh:ini]fos whicli ;ire presenUHl consist chieHy in concrescence. 

 This takes place in various ways. The proximal portions more fre- 

 quently vmite tlian the distal, but we may have conci-escence of the 

 distal points while the proximal portions are separate. The reduction 

 of rays is exhibited in all deoreos from the slightest shortenino- t() 

 the extreme degree shown in PI. LVTI, tig. 50, ray 0. 



It is perhaps noteworthy that the changes which would have suf- 

 ficed to differentiate the fin of one species from another, if they had 

 continuously advanced for a few generations, are changes which in 

 no slight degree are now taking place ln'twcen parents and childi-cn. 

 I mean changes of number, segmentation, concrcsccni'c, and relative 

 lengths of parts of rays. 



I exhibit SOUK' figures of the dorsal tins of Sqin/li/.^, 1*1. LVIT, 

 figs. 51-r)0, and PI. LVIII, iig. 57, liditt, PI. LVIII, tigs. 5S, 55), and 

 Mt/liohafiti, PI. 7>VIIT, fig. (iO. It is perfectly certain that they have 

 been derived from a sei-ies of parallel rays, and that the jjrincipal 

 process of change has beiMi in tlu' way of concrt'scence. They tell 

 their own story as far as it can be told without the investigation of 

 other closely related forms. 



The last ray in the first dorsal of /xa/'o, PI. lAIII, tig. 5S, is 

 remarkable, and it is represented in the second doi-sal. fig. 50, by a 

 serii'S of se])arale nodules. 



