76 FOSSIL JAWS OF MACJiOl'ODID.f:, 



yield reliable estimates of dimensional extremes and averages 

 within the species, and accurate views of the extent of likeness 

 and difference in form and size maintained among themselves In^ 

 the species. Information of this kind has Ijeen obtained from 

 479 skulls, namely, of Macropus (jigaideas 80, M. rufns 9, M. 

 robnstus 39, Halinaturus parryi 55, //. agilis 29, IJ. dorsalis ^^, 

 H. rvficollls 50, H. coxeni 9, H. thetidis 19, H. icilcoxi 2, B. 

 stiyiiiaticus 3, //. uaiahatus 8, //. hroivni 1, Onychoyale frenata 4 

 l^t^troyale penicillata 70, Dendrolayus lumholtzi 3. Furnished 

 with this instruction and with a resolution to be chary of assuming 

 anything of a fossil which may not l^e predicated of a similar 

 living species, it may be possible to thread the maze ])efore us 

 with more confidence in the progress made than would l)e per- 

 missible were the clues less frequent. 



.Since the mutilations to which the fossil remains ha^e been 

 subjected diminish in number the available points of comparison 

 })etween them and recent jaws, those data only have been asked 

 from the latter which are given with more or less constahc}^ by 

 the former. 



As to measurements, the following are those which have ])een 

 found the most useful in practice. The length of the full series 

 of cheek-teeth and its width as represented 1)}^ that of ni'^, the 

 niolar most frequently preserved in the fossil state : the length of 

 the premolars, permanent and deciduous: the external length of 

 the mandible from the edge of the masseteric fossa to that of 

 incisive outlet: its internal length from the edge of the iiitermas- 

 seteric foramen to the symphysis: its vertical height, anteriorly 

 at the fore end of the tootli m^, and posteriorly immediateh^ 

 behind ni"^ : and the thickness of the ])one below m'^. Of less 

 frequent service are the length from the hinder end of the sym- 

 physis to the incisive outlet, the length of the diastema, that of 

 the basiocranial axis, the breadth of the palate, and the height of 

 the ah'eolar process behind the orl)it. 



The following tables, which may be of some use to others 

 enu'aa'ed in similar work, are summaries of the measurements 

 taken under the headings which seemed most important. An 



