INTEODCrCTION. Vll 



alone or from the skulls alone. But, in order that inexperienced 

 workers may obtain the full benefit of the cranial synopses, a few 

 words on the nomenclature used for the teeth are here necessary. 



In most of the earlier systematic works on Mammalia all the 

 teeth were simply numbered from before backwards, a system that 

 answered well enough in the case of the incisors and molars, as 

 these teeth in the gradual process of their reduction in number 

 were generally lost posteriorly, so that, for example, the first three 

 molars of an animal with four molars were severally homologous 

 with the three of an animal possessing only the latter number. In 

 the case of the premolars, however, the matter is far more com- 

 plicated, as not only are the last one or two the most stable of all, 

 but in some Placentals and in most Marsupials the first premolar 

 to be lost was at neither end of the series but in the middle of it, 

 being the second of the four normally present. The system here 

 adopted for the numeration of the teeth is that by which each 

 tooth is represented by a sign always belonging to its homologue in 

 all the different species, whatever their total number of teeth may 

 be. Thus, for example, the last premolar of the Opossums, the only 

 one that has a milk predecessor, being homologous with the fourth 

 premolar of Mammals possessing the full set of four premolars, is 

 called " p.', " although it is preceded in the jaw by only two other 

 premolars. In the same way the two premolars of Dasyurus are 

 called " p.' " and " p.^," since the missing premolars appear in this 

 case to be the second and fourth of the full set. 



The general principle of this system was first introduced by 

 Owen, and it has since been used by a large and steadily increasing 

 number of authors. It is true that there are many cases in which 

 the homologies of the individual teeth are very difiicult of determina- 

 tion, and in which mistakes are at first nearly certain to be made. 



It is hoped, however, that these mistakes wiU in time be rectified, 

 and in any case, if only to stimulate further inquiry, it seems 

 better to make an attempt to work out systematically the tooth- 

 homologies of such a group as the present, rather than to rest 

 content with the old and admittedly imperfect system of simple 

 numeration from before backwards. 



The measurements given throughout the volume are taken as 

 follows : — 



Head and body. — In stuffed specimens taken with a tape from 

 the tip of the nose to the root of the tail, following the curves. 



