1897.] SKELETON OF PELODYTEiS PUNCTATUS, 585 



proximal ends and the consequent angle which they make with 

 the plane o£ the hypobranchial plate, it requires very careful 

 observation to demonstrate the fact in this region. The only 

 fibrous connection is that between the first and second ceratobran- 

 chials just anterior to the second spicule. The spicular cartilages are 

 definitely continuous with the cartilage of the ceratobranchials, and 

 do not give one the impression of their having arisen independently. 

 These spicules (I have adopted Graupp's name for these processes) 

 of the Anuran larva have been the subject of a good deal of contro- 

 versy by reason of the interpretation which Parker put upon them 

 in his third contribution to our knowledge of the skull of Batrachia 

 (12). I have already in an earlier communication (16. p. 113) 

 expressed my views on the subject and do not propose to discuss 

 the matter farther. My observations on the specimens which 

 form the subject of this paper tend to confirm the view of Gaupp 

 (5. p. 408), that the spicules have no morphological value, but 

 that they are merely processes of cartilage developed from the 

 proximal ends of the ceratobranchials to support the horizontal 

 fold or velum which covers over the anterior half of the branchial 

 cavity, and which undergoes reduction at the same rate as the 

 spicules. 



The mandible is quite diminutive, and consists of four subequal 

 cartilages. The median pair of these (m, fig. 1) carry the lower 

 horny beak, the lateral pair (??i') are articulated with the palato- 

 quadrate cartilage. 



Stage 2. Distance from, snout to root of tail, 20 mm. Tail at 

 its maximum, development, 28 mm. in length. Length of hind limb, 

 extended, 21 mm. Fore limb not yet extended. Horny beaJcs still 

 jyresent. (Plate XXXV. fig. 2.) 



But few differences are to be noted between this and the 

 preceding stage. The whole skeleton is considerably larger, the 

 hyoglossal notch is more rounded than before, and the four 

 constituent cartilages of the mandible are beginning to straighten 

 out. Except for these features the description already given of 

 the hyobranchial skeleton in the first stage would apply equally 

 well in this Stage 2. In fact, my chief object in introducing this 

 into the series was to show that the hyobranchial skeleton attains 

 its maximum larval dimensions without any appreciable change in 

 shape. 



Stage 3. Distance from snout to root of tail, 16 mm. Length of 

 tail, 20 mm. Length of hind limb, extended, 19 mm. Length of 

 fore limb, extended, 8 mm. Horny beaJcs no longer present. 

 (Plate XXXV. fig. 3.) ' 



The metamorphosis of the hyobranchial skeleton is now 



^ I have here given the njeasurements of the specimen the hyobranchial 

 steleton of vi^hioh is shown in fig. 3. It will be observed that tbe specimen is 

 smaller than the average ; but if all the above measurements be multiplied 

 by J the figures will be seen to be intermediate between those of the specimens 

 chosen to represent Stages 2 and 4. 



