1894.] NATURAL SCIENCES OF PHILADELPHIA. 411 



and Celebes. 20 I believe, that this occurrence may be considered as 

 a re-adaptation of these two species, as they are found also in fresh- 

 water. Since the genus Caridina is not a primitive one, while the 

 genera of the Xiphocarince are so, and live exclusively in fresh- 

 water, it is very probable, that the fresh-water habit of the family 

 must be regarded as the original manner of living. I believe, there- 

 fore, that the Atyidce, even of the Indian Archipelago, are not im- 

 migrants from the sea, as stated by Weber (1. c. , p. 543), but "true 

 localized fresh- water animals, forming an old element of the fresh- 

 water fauna." 21 



The main differences of the Atyidce and their supposed ancestors, 

 the Acanthephyridce, are morphological as well as biological, the 

 Acanthephyridce being true marine, and essentially abyssal animals. 

 To all appearance the morphological differences are causally connected 

 with the change of habits. The peculiar pencil of hairs at the distal 

 extremities of the fingers is adapted for securing the special food 

 required, as described by F. Müller in Atyoida potimirim. 22 No 

 doubt the other species of Atyidce feed in the same manner. I cannot 

 say whether the absence of the synaphipod of the mandible is due to 

 the same cause, since the function of the synaphipod is unknown, but 

 it may be in connection with it. On the other hand the habits of 

 the Acanthephyridce are wholly unknown, so that we cannot compare 

 this family with the Atyidce, but it is very probable that the mor- 

 phological differences of the Acanthephyridce correspond to differences 

 in the habits, especially in securing food. 



We can state, briefly, that the Atyidce are closely allied to the 

 most primitive Eucyphidea, forming a peculiar branch of develop- 

 ment very early separated from the main stem, now represented by 

 the Acanthephyridce. Their several characters are connected with a 

 change of habit, and with the immigration to fresh-waters. 



The geographical range of the Atyidce embraces the whole of the 

 circumtropical parts of the world, members of the family being re- 

 corded from all the localities explored within these limits. Only in 

 two localities does the range exceed the true tropics : in Japan, where 

 it extends as far north as Tokio, and in the Mediterranean province, 



20 See de Man, ibid., pp. 387, 399, 400. 



21 Weber, 1. c., p. 533: "ecbte regionale und locale Süsswassertiere, die 

 einen alten Bestand der Süsswasser Fauna bilden." 



22 Kosmos, IX, 1881, p. 117 ff. 



