AMPHIBIA. G5 



outer edge of fifth toe. This species has the same habits as Rana tigerina, with 

 which it is usually found associated in the swamps, rice-fields, and banks of 

 streams. It is, however, much less common. Bryant's collection contains 

 but two specimens. These do not agree at all with specimens from Japan and 

 the Riu Kui Islands. So that there is some doubt as to just what Rana limno- 

 charis really is. The species is very rare in Java; consequently if the type came 

 from Java, it is quite likely that it was a young Rana tigerina. In which case 

 both R. limnocharis, so-called, from Java and Japan, will need new names. 



Rana papua Lesson. 



Plate 5, fig. 1.5. 



Lesson, Voy. Coquille. Zool., 1830, 2, 1, p. 59, pi. 7, fig. 1. Boulenqer, Cat. Batr. Sal. Brit, imi.s., 

 1882, p. 64. Van Kampen, Nova Guinea, 1906, 6, 6, p. 164. 



Type locality: — the island of Waigiu. Here, Lesson remarks, the natives 

 brought them almost every day to the ship as an article of food. 



Six examples were taken in Sorong, one at Manokwari, and two at Pom, 

 Jobi Island. A large female from the last-named station has been figured 

 (PI. 5, fig. 15) and the colors are accurate to the living condition. A male from 

 Pom exhibits well-developed humeral glands. Boulenger {loc. cit.) remarks that 

 they do not occur in this species, while van Kampen, on the other hand, found 

 them on a male from Ibaiso. His records show that this was the only adult 

 male which he had before him. Van Kampen thinks that this may be a chance 

 occurrence. It seems, however, that pcssibly there may be some geographical 

 variation. Boulenger's Papuan examples were all from the Torres Straits 

 region, while these two cases are from the northeast region. 



This frog is now known from Waigiu, Batanta, Jobi, the Aru Islands besides 

 New Guinea itself. It is also reported from northern Australia and Timor-Laut. 



Rana moluccana Boettger. 



Plate 5, fiK. 12. 



Boettger, Zool. anz., 1895, 18, p. 132. Abh. Sonck. nat. ges., 1900, 25, p. 366. (Rwia varians Boulen- 

 ger). 



Type locality: — "Auf ganz Halmaheira und Ternate haufig." 



While Boettger in his preliminary report described this frog as peculiar to 



the Halmahera group of the Moluccas, he considered it as synonymous with 



Rana varians Boulenger in the final report on Kiikenthal's collection. Thanks 



to Dr. Stejneger I have before me three topotypes of R. varians, U. S. nat. mus., 



Nos. 39,9G4, 39,977, 39,978, from Pancal and Puerto Princesa, Palawan. On 



comparing these with one from Gane, Halmahera, and eighteen from Ternate, it 



becomes at once evident that, though these species are related closely, yet they 



