BREEDING PLACES OF THE EEL SCHMIDT 315 



species can not be defined, either zoologically or geographically, 

 without further characters than hitherto employed in classification. 



It was therefore necessary to take up the whole question anew. 



I have for several years been working on the classification of the 

 Indo-Pacific fresh-water eels, but have only lately been able to give 

 it full attention. 



Thanks to the courtesy of the various authorities, I have been able, 

 with my assistants, especially Mr. Vilh. Ege, M. Sc, to examine 

 the material of Indo-Pacific Anguilla specimens of most of the lead- 

 ing museums, and get X-ray photos of most of the existing type 

 specimens. I have also received large collections from correspond- 

 ents in various parts of the world. 



In all specimens, the number of vertebrae was determined, and 

 other numerical characters investigated; several species have al- 

 ready been treated statistically, as previously with the Atlantic and 

 Japanese.^ 



The investigation shows that the previous classification is inade- 

 quate. Take for instance, the Anguilla australis, originally de- 

 scribed by Sir John Richardson. This species, as understood by 

 the most recent writers on Indo-Pacific eels, ranges from East 

 Africa to Tahiti, and from the Philippines to Auckland Island. 

 Geographically, this looks improbable, and on applying numerical 

 characters we find that ^^ Anguilla australis " consists of at least three 

 well-defined species, each with its own geographical area. 



The introduction of numerical characters, however (number of 

 vertebrae, fin-rays, etc.), is of importance not only for classification 

 of old, adult eels. We have thereby procured the means, hitherto 

 lacking, of identifying the early stages of the Indo-Pacific fresh- 

 water eels. 



The method has already borne fruit, in that I am able here to 

 describe and identify the first larva of an Indo-Pacific Anguilla 

 species, and thus answer the question at the head of this postscript. 

 I am the more pleased to be able to do so here, since the material 

 was procured by an American expedition. 



Some years back, I received from the Bureau of Fisheries at 

 Washington a large collection of Leptocephali for investigation. 

 It was collected by the U. S. S. Alhatross on its Philippine expedi- 

 tion (1907-1910) under Dr. Hugh M. Smith. 



This collection proved to contain two specimens of a Leptocephalus 

 outwardly altogether identical with the larvae of the two Atlantic 

 Anguilla species. They were taken near the surface over consider- 



» Jobs. Schmidt : " First Report on Eel Investigations " and " Second Report on Eel 

 Investigations" ("Rapports et Troc^s- Verba ux du Conseil International pour I'Ex- 

 ploration de la Mer," vol. 18, 1913, and vol. 23, 1915). See also Figure 1 in the present 

 work. 



