30 MR. W. K. TARKER ON THE STRUCTURE AND 



that tlie first and second extra-branchlals (Plate 4, fig. 3, ex.br^~K) have lost mucli of 

 their pouch-like shape, and are but little broader than the second and third ; they all 

 become very narrov^^ bands, and thus vanish, with the exception of a common remnant 

 that becomes fused with the permanent hyo-branchial plate. 



The third intra-branchial (chr^.) is very long, as long indeed as the counterpart 

 bars ai-e in Selachians, Ganoids, and Teleostei ; in its normal position, round the 

 pharynx, this bar just reaches halfway up the side, and in those Fishes is surmounted 

 by the epi-branchial. 



In the Dactylethra Tadpole (Phil. Trans. 1876, Plate 58, fig. 1) the whole skeleton 

 of the branchial pouches is one continuous structure ; here the intrabranchials are 

 nearly segmented oiF from the extra-bran chials, but there is an isthmus of cartilage 

 uniting them in Rana, as may be seen in the section (fig. 4, c.hr^., ex.br^.). 



I must pass again to Bana pvpiens for my next stage ; the reader will, however, 

 easily eliminate the non-essential specific variations, at any stage, from those more 

 important step by step processes that transform a Cyclostomous skull into one with 

 a widely gaping mouth and jaw-hinge far behind, like that we see in certain aberrant 

 Selachians — Notidanus and Cestracion. 



Fourth Stage. — 2 (continued). Tadpoles of Kana pipiens (C) with tail lessening ami 



all the legs free. 



This stage (Plate 3, figs. 4-13, and Plate 9, figs. 1-6), of this species, stows that 

 iu the largest and most generalised of the genus the branchial apparatus remains in 

 full development and function nnicli longer than in the lesser Frogs. They are two 

 or three years (according to Wyman) before they have finished their metamorphosis, 

 and that excellent naturalist kept them (in captivity) six or seven years as Tadpoles. 



In R. clamata (Stage B) the nasal roofs just appear as the branchial pouches are 

 dwindling away. In R. pipiens the whole nasal labyrinth is perfected before any sign 

 of diminution appears in the gill-pouches. 



This indicates that the Tadpole of the Bull Frog is well fitted out with special 

 sense organs long before it finishes its metamorphosis, and suggests the probability of 

 a still slower metamorphosis in former epoclis. Rseudis ofi'ers a remarkable confirmation 

 of this view. 



Comparing this ^\•ith the two last stages we shall see both the general progress and 

 the specific differences that ai'e found to be even in larvae of the same genus. 



The most striking change is the elongation of the palato-pterygoid band (Plate 3, 

 figs. 4-6, p.2')g.), thrusting the suspensorium outwards. Then the " waist " of the 

 cranial box is narrower, and its sides more chondrified (fig. 6). 



The roof bones are segmented across (fig. 4,/, p.), and, contrary to wont, the frontals 

 are only half the size of the parietal s. 



Below, the parasphenoid (fig. 5, pa.s.) has become more developed, and foliated 

 behind; in fiuiit, the membranous layer seen in the earlier larva (fig. 2) Is now partly 



