150 Mr. O. Thomas on 
foramina about the length of the molar series, not reaching 
back to the front of m*. Molars light and narrow, of the 
usual Holochilus pattern. 
Dimensions of the type (measured by collector in the 
flesh) :-— 
Head and body 178 millim. ; tail 159; hind foot, s. u. 38°5, 
c. u. 42; ear 15. 
Skull: greatest length 37; basilar length 30°7; greatest 
breadth 20°6; nasals 14°5x 4:5; interorbital breadth 95; 
palate from henselion 18°7; palatal foramina 7°5 x 2°6; length 
of upper molar series 6:9. 
Type. B.M. no. 1. 6. 4. 87. 
This water-rat, whose discovery extends the known range 
of the genus some way to the northward, is no doubt very 
closely allied to H. sciureus, but differs from it by its darker 
colour (Wagner says that in his type “fehlt eine schwarze 
Beimischung ganz’) and by the more lighter build and less 
ridged skull. H. nanus, the smallest species of the genus, 
occurs in the Amazonian valley, and divides the ranges of the 
other two. 
20. Sigmomys * savannarum, sp. 0. 
Fourteen skins. 240 feet. October and November. “On 
Savannahs.” 
Allied to 8. Alstoni, Thos.+, and similarly like a Sigmodon, 
but paler in colour, with shorter foot and shorter skull. 
Size rather less than in S. Alstont. Fur straight and 
sparse, with little or no underfur; hairs of back about 
10 millim. in length. General colour above heavily lined 
and mottled brown and greyish buffy, the combination not 
easily definable, but perhaps nearest to Ridgway’s “ broccoli- 
* Genus novum. Type “ Reithrodon” Alstoni, Thos. P. Z. 8. 1880, 
. 691. 
r The many differences that distinguished #. Alstone from the true 
Reithrodons were detailed when the species was described. Further 
consideration and much further material render it clear that the Alstonz 
group, Sigmodon-like in external appearance, in skull, and in dentition, 
should be separated generically from the true LRerthrodon. In distin- 
guishing Sigmomys emphasis may be laid on the evenly tapering cranial 
outline, the broad short molars, the deep well-defined pterygoid fosse, 
and on the number of the mamme, there being, as in Stgmodon, 10 at 
least, as against 8 in Reithrodon. Altogether it seems probable that this 
form is rather a grooved-toothed Stgmodon than any close relation to 
Reithrodon. 
+ This species was described on a spirit-specimen from “ Venezuela” 
collected many years before by Mr. Dyson. But an excellent series of 
skins, agreeing closely with the type, was recently received from Cumana, 
so that that may now be looked upon as the typical locality. 
