1888.] THE CARPUS AND TARSUS OF THE ANURA. 155 



In Pseudis tadpole and Hxjla (figs. 26 & 20), less conspicuously in 

 Biifo and others, it is elongated in the manner of a finger-joint. 

 With the exception of Duges, who first regarded it (16) as a meta- 

 carpal, all investigators have looked upon it as the carpal of the 

 pollex. Examination of its detailed relationships and consideration 

 of its embryological condition {Alytes, fig. 9, po. i.) leave no room for 

 doubting the accuracy of the latter determination. 



Assume what form it may, the pollex usually ossifies more 

 early in the males than in the females. In the Ranidcs, for 

 example, it remains in the females for a considerable time simple and 

 unossified (woodcut, fig. B, p. 1/4). This precocious ossification in 

 the male sex is clearly the result of its connection with the well- 

 known " thumb-pad," for it supports that more or less extensively 

 in different genera and species. In some forms there is a great 

 tendency for it to enter into direct connection with the 2nd meta- 

 carpal, either by simple apposition or by fusion with the warted 

 crest of the same. In the male Biscoglossus, however, the reverse 

 is the case, for it there appears (fig. 6) to have undergone an increase 

 in breadth directly proportionate to its diminution in length. Ex- 

 amination of the living animal shows that this change is an ac- 

 companiment of one undergone by the " thumb-pad ; " for that is, as 

 it were, duplicated, its lesser representative being alone supported 

 by the pollex. 



In the youngest individuals examined by us (e. g. Ranidce and 

 Alytes) the pollex had the form of one elongated unsegmented 

 cartilage. It ossifies as a single element in all Banidce, Bufonidce, 

 and Discofflossidce which we have manipulated. On the other 

 hand, the like is true only of certain genera in the Cystignathidce, 

 PelobafidcB, and Hylidce. 



In all forms examined by us, not cited above, it had become 

 more or less thoroughly segmented into two or three pieces (j9o. ii. to 

 /)o. iv. of figs.). These, in some of the Tree-Frogs especially, take 

 on the characters of a metacarpal twice as long as broad and two 

 short phalanges. In them and certain other forms it undergoes an 

 elongation at the same time ; in yet others the reverse is the case. 

 It must not, however, be imagined that a converse line of modifi- 

 cation is here set up, for Leptodactylus (fig. 2')), Pelodytes (fig. 11), 

 Ceratnphrys, and Callula all show that, while shortened up, it is still 

 segmented into distinct joints ; and these, moreover, subsequently 

 undergo complete and independent ossification. 



When ossification sets in without previous segmentation, it does 

 not spread from one or more definite centres, but it is diffuse and 

 irregular. 



In the males of Rhombophryne (fig. 32) and Lymnodynastes (fig. 

 28) the pollex is quite calcariform. The characters of the rest of 

 the member are, in both genera, such as to show, beyond doubt, that 

 we have here to deal with extremely specialized forms. 



d. Ulnare{Pyramidale). — Wiedersheim has shown' that in Rana 



^ ' Anat. d. Frosches,' Abth. ii. p. 79. 



