HYNOBIIDAE. 473 



Mr. S. Ikeda in a previous volume I knew they were their eggs, and 

 convinced myself by rearing them that they were the same species as 

 the tadpoles I studied at Tada Lake. 



"There are two natural hatcheries of this animal at Suganuma, 

 besides the place just mentioned, one in Nahanuma, and the other 

 at the southeastern corner of Kitamata. 



" The adults of this species are found at Ojiri on Kitamata, and on the 

 south shore of Nahanuma, where the beach faces to the west. They 

 live along the lake, generally from three to five feet from the water so 

 that floods do not disturb them, under rocks or decayed trees, and 

 seem to live in colonies although they are often found singly. 



"They generally protect themselves from attacks of enemies very 

 cleverly, with more than two entrances or holes into their dwelling 

 place, and make so many passage ways in the ground that in case of 

 attack they can hide or escape. However when the rock or log that 

 covers their place is suddenly removed they simply shrink and try to 

 hide by putting only their head under rocks or in the ground while 

 their body is utterly exposed so that we can easily catch them. 



"They go out near the water at night for food, which consists of 

 worms, insects, eggs of tiny animals, fishes and their eggs, and so 

 forth, while the larva eats both animal and vegetable food; algae 

 that grow on dead leaves and bushes in the water, copepods and frog 

 tadpoles and larval insects. 



" In May after the snow is gone from the mountain sides at Nikko 

 they come out of their winter retirement to the lake to breed. The 

 breeding takes place in the darkness of night and cannot be described 

 In summer they stay around the lake or hunt for food along the valleys, 

 so that we can catch them by moving rocks or piles of dead leaves. 

 They retire late in September or early in October for a long winter rest 

 under decayed trees, in the ground or under piles of dead leaves. 

 Sometimes they go many feet deep into the ground following along 

 tree roots. 



'The hatching process is similar to that of H. naevius. Sometimes 

 a green alga gets into the eggs." 



Remarks: This animal is related on the one side to H. leechii of 

 Korea and on the other to H. rctardatus of Yezo. 



The Yezo form is sufficiently specialized to be quite distinct but the 

 Korean form is quite close, and the color is perhaps the most trust- 

 worthy feature by which they can be separated. The lichen -like 

 light patches on the sides of peropus, and the absence of marked dark 

 spotting on light ground on the back is quite different from the state 



