1894. NATURAL SCIENCES OF PHILADELPHIA. 131 



It moreover seems hardly advisable to make so many genera out of 

 the old genus Porzana as is done in this scheme. 



Two of the generic names adopted by Dr. Sharpe ( Corethrura and 

 Rallina) are clearly untenable as is shown below. 



Although the Rails of the New World have been excellently 

 monographed by Messrs. Sclater and Salvin, P. Z. S. 1868, p. 442, 

 no recent attempt has been made at a systematic arrangement of the 

 Old World members of the family except in the paper by Dr. Sharpe 

 just referred to. 



The satisfactory arrangement in a lineal sequence of the genera of 

 any family, especially such a one as the Rallklce, is well nigh im- 

 possible, and there must necessarily be breaks in the series. 



From the typical Rails (Rallus and Hijpotaenidici) we can run in 

 one direction through jE'H/(^6eo>v;/s and the G'yjmiorTf.r group to the 

 Woodhens {Ofijdromus) unA again through DryoUmnas, Rougetius 

 and Euryzona towards the Crakes. In any case the genera Ocydro- 

 mus and Himantornis are extreme forms and had better stand 

 separately at the end of the series than be interpolated in the middle. 



Cabalus is evidently allied to Hypotuenidia and Canimllus to 

 Euryzona though they are both aberrant forms. 



With these ideas in mind I have begun my list with the genera 

 most nearly approaching the Woodhens and followed with the true 

 Rails passing from them to the Crakes. As Dr. Sharpe says the 

 Crakes merge on the one hand into the Rails and on the other into 

 the Gallinules, Amaurornis being the connecting link with the 

 latter. 



Of the Crakes I have recognized eight genera as follows : Crex, 

 Porzana, Lhnnohaenua, Limnocorax, Sarothrura, RalUcnla, Porz- 

 anula and Pennula. 



The Gallinules and Coots have not been included as I have 

 limited my paper to what are generally known as the Rallince though 

 it is an exceedingly difficult matter to draw a sharp line between the 

 several so-called subfamilies of the RalUdce. The genera Amauror- 

 7iis and Oenolhnnas^ have also been omitted as they seem to belong 

 with the Gallinules, and also the genera Ocydromiis " and Himantor- 



^Dr. Sliarpe arranges this genus (type J?a/lt7ta isabellina Schl.) with the 

 Crakes but it seems to me closely allied to Amaurornis. 



2 Monographed in BuUer's Birds of N. Zealand. See also Ibis, 1893, p. 261. 



