MACKOPODlDiE. O 



predecessor, and is therefore the one to be regarded as the last 

 premolar, or " p.*," of the typical Mammalian dentition. 



The special characteristics that render the development and suc- 

 cession of the teeth in the Macropodidce, and especially in the 

 genus Macropus, so puzzling to systematic zoologists are :■ — firstly, 

 a general progression forwards in the jaw of the whole tooth-row, 

 comparable to that found elsewhere only in the Elephants and some 

 Sirenians ; and, secondly, the fact that before the tooth-change the 

 first tooth of the series (p.") and the single milk-tooth (m.p.*) placed 

 next to it, both of which fall out at that change, are respectively so 

 very similar in shape and size to the first and second cheek-teeth of 

 the permanent series, viz. the permanent premolar (p.*) and the 

 first molar (m.'), as to be most naturally mistaken for, or compared 

 to, them in specific descriptions. The following woodcuts (figs. 1 & 2), 



Fig. 1. 



!>*- 



, 7n3' 



..-'" 



Macropm rujicollis. — Tipper cheek-teeth of (A) immature and (B) adult, 

 showing the superficial resemblances and the true relationships of the 

 former to the latter. 



taken from two specimens of the same species, show this resemblance, 

 which is in most cases sufficiently close to render very great care 

 necessary, lest skulls in different stages of dentition should be 

 unknowingly compared together for purposes of specific description. 

 The necessary knowledge as to the stage of dentition in which any 

 skuU may be can often only be gained by cutting open the bone either 

 above and behind the first tooth of the series to see if the true per- 

 manent p.* be still buried there (in which case, of course, that first 

 tooth is only p.'), or behind the last visible molar to see if there is 

 yet another tooth behind it, showing it to be m.' and-not m.^ The 

 first plan is, as a rule, the best, as p.* is generally by far the most 

 important tooth for diagnostic purposes, and its characters have 

 therefore in any case to be taken into account. The side views 

 (fig. 2) of the same two skulls show the position at which p.* should 



