1894.] IN THE BEITISH MTJSEXTM. 117 



(6) The rotular surface is very broad and flat, and makes an 

 angle of about 90^ with the inferior intercondylar surface 

 instead of passing into it by a gentle curve. 



The intercondylar fossa is slightly marked. The upper 

 pneumatic foramen is present as in JE. titan, and, the floor of the 

 pophteal fossa being broken away, it can be seen that the inferior 

 foramina opened into a large air-chamber. There is also a pneu- 

 matic foramen about the middle of the rotular surface, which is 

 not found in ^. titan. The nutritive foramen on the posterior 

 surface of the shaft is single, in the femui* above referred to 

 ^. titan there are two ; but since this is not the case in some of 

 the more imperfect femora belonging to the same species, it is 

 merely an individual variation. 



This femur may possibly belong to ^. maximus or to either of 

 the recently named species, M. cursor and ^. Untus (4). In any 

 case, as was remarked above, the differences between it and the 

 femur referred to ^. titan appear to be of generic value ; and if 

 this be so, then there is evidence of the existence of a third genus, 

 since the recently named Mulleromis seems to include only slender 

 forms of comparatively small size. 



Until, however, it is definitely known whether the femur of the 

 type species of ^pyornis resembles that of ^. titan or the one just 

 described it would be imprudent to establish a new generic name. 

 The evidence necessary for the decision of this question is probably 

 in the hands of MM. MUne-Edwards and Grrandidier, and a full 

 description and figures of the magnificent collection recently briefly 

 noticed by them, especially of the skulls and sterna, wiU be 

 anxiously awaited. 



From the same locality there are several fragments, including the 

 distal ends of three tibio-tarsi, which appear to belong to ^. max- 

 inrns. A right tarso-metatarsus with the upper end above the 

 interosseous foramina broken away may also be provisionally referred 

 to the same species ; it is, however, sHghtly smaller, measuring 

 6-3 cm. across the narrowest part of the shaft in comparison with 

 6'9 cm. in ^. maximus. If uE. mediiis should be found to be a 

 distinct species, this bone may possibly belong to it. In form it 

 closely resembles the tarso-metatarsns of JS". maximus figured by 

 Milne-Edwards and Grandidier (3), and, like it, differs from the 

 tarso-metatarsi from the centre of the island mainly in the fact that 

 the shaft contracts from side to side above the trochleae more 

 gradually and to a slightly less extent ^ 



From Itampulu-Ve and Amboulisatra are several portions of the 

 skeleton of a much smaller form, possibly that recently named 

 Mulleromis agilis by Mine-Edwards and GJ-randidier (4). These 

 include the distal ends of several tibio-tarsi (Plate XV. fig. 1), 

 which closely resemble the same bone of the larger forms in their 



1 This specimen may belong to the species recently named ^. cursor by 

 Milne-Edwards and Grandidier. 



