334 EDWIN S. GOODRini. 



Introduction. 



A GREAT (leal has been written in the la.st few years about 

 the structure, development, and origin of the paired fins of 

 fish/ yet two rival and incompatible theories are still preva- 

 lent. According to the theory put forth by Gegenbaur (14, 

 15, 16), the paired fins have been derived from gill structures, 

 the gill-arch having been modified into the limb-girdle, and 

 the fin itself, with its skeleton, having been derived from the 

 gill-flap or septum, and its supporting gill-rays. This may 

 shortly be called the "gill-arch theory." The second theory, 

 that of Balfour (1, 2), Thacher (35), and Mivart (23), holds 

 that the paired fins are of the same nature as the unpaired 

 median fins. According to this view, the limbs have been 

 derived fi'om paired longitudinal fin-folds, in which skeletal 

 supports, the radial s, or somactidia (Lankester), became 

 developed as in the median fins, and subsequently gave rise 

 to the limb-girdles. This may be called the "lateral fin-fold 

 theory." Each of these theories may claim to have among 

 its numerous supporters the names of some of the most 

 eminent exponents of the morphology of the vertebrates. 

 Dohrn (10), Haswell (20), Rabl (31), MoUier (24, 25, 26), 

 Harrison (19), Wiedersheim (36), A. Smith Woodward (37), 

 and Dean (9) have written in favour of the lateral fold theory; 

 Davidoff (8), Fiirbringer (12), Brans (3-7), and others have 

 supported its rival. It is unnecessary for me in these notes 

 to give a history of the discussions to which the question has 

 given rise; the literature has been recently reviewed by 

 Mollier and Braus, and the whole subject is familiar to 

 zoologists. But there are certain essential points which seem 

 to bo in danger of being obscured from view in a cloud of 

 controversy, and it is in the hope of clearing up some of these 

 points and of filling up certain gaps in the evidence that 

 these notes have been published. 



As I am anxious to keep this paper within reasonable limits 

 and not to overburden the already very bulky literature on 

 the subject of the origin (ill lie ])iiii-e(l limhs, only some aspects 



