626 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.40. 



In relation to the soil there is a notable contrast between the habits 

 of millipeds in the Central American region and in West Africa. 

 Owing, perhaps, to the great abundance of termites, the African for- 

 ests have very slight superficial accumulations of dead leaves . and 

 humus. The soil remains relatively open and noncolloidal, and is 

 inhabited by numerous species of millipeds. In the forests of tropi- 

 cal America the termites effect no such complete destruction of the 

 dead leaves and other vegetable debris, which accumulate and 

 decay on the surface. The underlying soils are generally much more 

 colloidal than in Africa and the milliped population is generally sparse, 

 or often lacking altogether, especially after the land has been under 

 cultivation for a few years. 



SYSTEMATIC POSITION OF THE HOTHOUSE MILLIPED. 



The hothouse milliped was first described in 1847 by C. L. Koch 

 as Fontaria gracilis, but has still to find a satisfactory place in generic 

 classification. In addition to the original reference to Fontaria, 

 the species figures in literature under two other generic names, 

 Paradesmus and Orthomorpha, but none of these names seems to find 

 a correct application to the animal in question. 



In 1872 Humbert and Saussure referred Koch's species to the old 

 genus Polydesmus, as a member of a subgenus Paradesmus, that had 

 been established by Saussure for other species in 1859. In Latzel's 

 monograph of the Austro-Hungarian myriapoda, published in 1884, 

 the same course was followed, except that Paradesmus was recog- 

 nized as an independent genus. 



The next change was made in 1893 when Bollman discovered the 

 preoccupation of the name Paradesmus and proposed Orthomorpha 

 as a substitute. Hence the species has been called, by all recent 

 writers, Orthomorpha gracilis. The unfortunate necessity of a fur- 

 ther change arises from the fact that Bollman's new name was pro- 

 posed as a direct substitute for Paradesmus, and is therefore limited 

 in its application to the genus to which the original type of Parades- 

 mus belongs. 



DESIGNATION OF THE TYPE OF PARADESMUS. 



The determination of the type of Paradesmus carries us back to 

 the original treatment of the subgenus by Saussure. Several species 

 were included, though none that recent authors have treated as 

 congeneric with the hothouse milliped. Pocock has recently pro- 

 posed, in the Biologia Centrali-Americana, to designate as the type 

 of Paradesmus (and Orthomorplia) , an East Indian species, 0. olain- 

 villei. This designation can hardly be accepted, in view of the fact 

 that blainvillei was placed by Saussure in one of the two divisions of 

 Paradesmus that were indicated as forming transitions to other sub- 



