OF THE EYE MUSCLES IN MARSUPIALS. 339 



sicli mit ihrem Nerven, clem IST. trochleai-is, am spiitesten aus." 

 By the second stage (G.L. 13 mm.): — "Die Muskulatiiranlage 

 wandert nacli vorn gegen den N. opticus hin und verliert ihren 

 hinteren Schenkel, welcher von der Vena jiigularis nach voi'n 

 zusamniengedrangt wird." It will be remembered that in 

 Trichosurus (G.L. 7 mm.) the root of the vena cerebralis 

 anterior penetrates through the dorsal side of the larger anterior 

 portion of the abducens muscle-mass ; this does not occur in the 

 specimen of Dldelphys which has perhaps passed this stage in 

 development. 



Dasyurus viverrinus. 



A good series of embryos of Dasyurus viverrinujS was examined 

 and no signs of any head-cavities were observed. The small size 

 of these embryos, however, makes it exceedingly difficult to 

 identify any small vestiges of cavities which might ■possibly 

 persist. 



We may conclude from the preceding facts tliat the occurrence 

 of well-developed preinandibular head-cavities is chai'acteristic of 

 the Diprotodontia, large cavities being found in Trichosurus, 

 Phascolarctos, Phascolomys and Macro2nis. In the Polyproto- 

 dontia, on the other hand, cavities may be present, as for 

 example in Ferameles, but only in the form of comparatively 

 small irregular spaces, quite insignificant as compared with those 

 in the former group, or tliey are altogether absent, as in Dasyurus 

 and probably also in Didelphys, the muscles from the first somite 

 then developing from a solid mesodermal mass as in higher 

 mammals. 



Although there are variations in the mode of proliferation 

 from the walls of the premandibular cavity in the different 

 genera,, the later development of the eye muscles presents much 

 similarity and Trichosurus may be regarded as typical. 



CoxcLUDixG Remarks. 



From the foregoing observations we have further evidence of 

 the unique j^osition occupied by the marsupials in the Class 

 Mammalia. The possession of well-developed premandibular 

 head-cavities, as yet observed in no other mammals, recalls the 

 conditions existing in many reptiles. These cavities correspond 

 very closely with those found in the lizard (Corning, '99), in the 

 snake (Oppel, "90) and in the Chelonia (Filatoff, '07 & Johnson, 

 '13). 



Owing to lack of material of early stages it has not been 

 possible to determine the exact mode of origin of the cavities 

 at the anterior end of the head in marsupials, or even to establish 

 the existence of a median piece connecting the two across the 

 middle line. The connecting piece in reptiles very often forms 

 a wide cross-canal, which may persist until a comparatively late 



