I OO URODELA chap. 



rod. The latter is then withdrawn as quietly as possible, the 

 hook and bait being left. As soon as a jerk of the string is 

 noticed, a pull is made, which generally ends in the capture of 

 the unfortunate animal. If tlie first pull should fail, the bait is 

 replaced as before, and a second opportunity is offered, which the 

 unwary creature accepts as readily as the first. Tlie fisherman, 

 having obtained one bite, is sure of ultimate success, as the sala- 

 mander does not learn by experience to refuse the proffered 

 morsel. When captured, it emits a peculiar slimy secretion, 

 having an odour much like that of the leaves of the Japan 

 pepper {Xantlioxylon peperitum). This secretion hardens into a 

 gelatinous mass after a short exposure to the air. 



Temminck and Schlegel state that the act of inspiration is 

 ordinarily performed once every 6-10 minutes. This is true 

 for specimens kept in tubs ; but Sasaki is inclined to think 

 that they perform this act less frequently in their native brooks. 

 The eyes are so small that they are obviously of little import- 

 ance ; the salamanders capture their prey not by pursuing, but 

 by waiting for its near approach, wliereupon they seize it with 

 their teeth by a swift lateral movement of tlie head. The eggs 

 are said to be laid in August and September, and they form a 

 string resembling a rosary. Each egg floats in a clear fluid, 

 inclosed in a bead-shaped gelatinous envelope, and this is con- 

 nected witli the next by means of a comparatively small string. 

 The egg measures about 6 mm. by 4 mm., and is yellow 

 everywhere except at the upper pole, where it is whitish. All 

 attempts to make Cri/2:>tohranchus breed in captivity have failed 

 hitherto, owing no doubt to the difficulty of obtaining the cool 

 temperature of its mountain streams. Sasaki's smallest specimens 

 measured 19 to 20 cm. These had three pairs of very short 

 branchial processes, from 3 to 5 mm. in length, attached just 

 inside the branchial orifice. Each process was somewhat 

 flattened and tapering, most of them still with branchlets. In 

 another specimen, 20-5 cm. in length, the gills had almost 

 wholly disappeared, but the branchial slits were still visible. 

 One of 24-5 cm. length showed no trace of gills, and the 

 branchial orifice was completely closed, but still marked by a 

 licfht streak. 



Aiiiphiuma means s. tridactyla. — The limbs are very much 

 reduced, and end in two or tliree little fingers or toes. Just in 



