518 DUNN. 



'The eggs are protected from shock by stiff jelly. "When larvae 

 emerge the egg opens in the same direction as the opening of the bag. 



" The jelly bag remains attached to the rock empty and flat after all 

 the animals have gone. 



" The bags vary a great deal in size and in number of eggs contained. 

 Some are 2-3 vershok (4-6 inches) long, and some are even \ arshin 

 (14 inches) long, and the number of eggs varies accordingly. I have 

 a bag shrunk in alcohol which measures 2\ duima (2§ inches), and con- 

 tains 25 eggs. 



"The egg at hatching measures 1 cm. in diameter, and the larva is 

 1.75 cm. long. They have well developed front legs and perceptible 

 hind legs. The dorsal fin reaches the head. In the next stage, 2.5 

 cm., the dorsal fin reaches the middle of the body, and the larva has 

 developed hind feet. 



" The young are lighter in color than the adults. The background is 

 yellow with many brown dots which later become the background. 



"Many specimens have well developed gills, but otherwise do not 

 differ from those without gills. No very large ones were found with 

 gills, so probably those with gills are young ready to transform. 



"203 and 210 mm. were the lengths of my largest adults, the medium 

 sized examples were 150 mm. long, and the smallest were 90 and S5 mm. 



"The only specimen with atrophied gills is the smallest of all and is 

 81 mm. long, while two with gills measure 92 and 95 mm. 



"They normally live under stones, and their motion is halfway 

 between a swim and a crawl. But in deep water they put their legs 

 to their sides and swim with their tails. They probably feel uncom- 

 fortable for they immediately try to find a stone and go under it. 



" Both Kirghiz and Cossacks believe that these animals when dried 

 in the sun and powdered are a good medicine for broken bones. Dried 

 salamanders can be found in every Kirghiz or Cossack village. In 

 places they have been exterminated by catching them for this pur- 

 pose." 



Remarks: Ranodon is an isolated species whose relationships are 

 remotely with Hynobius, perhaps with its recently described com- 

 patriot //. turkestanicus. It certainly has nothing to do with Batra- 

 chupcrus. The two are independent more aquatic derivatives of 

 Hynobius, and the similar dentition, due to weakening of the pre- 

 vomers, is probably a parallelism. 



Ranodon kcsslcri does not seem to be a distinct form. 



Strauch (1870) stated that this animal occurred in Northeast China, 

 but as his only Chinese locality was Kuldja, which is in the extreme 



