Correspondence. 521 
the 4 anterior ones on each side to the premolares (muelas falsas), 
and the 4 posterior ones on each side to the true molars (muelas 
verdaderas); the true molars being distinguished by their larger 
size, quadrangular crowns, and by having three roots instead of one. 
The number of true molar teeth 4—4 here ascribed to the upper 
jaw of Macrauchenia is not only an exception to the dental formula 
of Perissodactyles, but of all other Ungulates, and indeed of all 
diphyodont-placental mammalia: it would be an anomalous resump- 
tion in the Ungulate order of a character of the Marsupial one. 
Again, the reference of the fourth grinder, counting from behind 
forwards to the series of true molars, gives a distinctive simplicity 
to the alleged four premolars in advance, which has hitherto been 
seen only in the Artiodactyle Ungulates, and which would be a more 
striking example of the tendency of the Perissodactyle Macrauchenia 
to the Artiodactyle order, than its cameloid cervical vertebre, or its 
coalesced antibrachial bones. 
After a careful study of the figures of the jaws and teeth in PI. 1 
of this interesting monograph, I am led to offer a different explana- 
tion of the phenomena. In the upper jaw of the incisive series the 
outermost only are in place, viz., 2 3—7z 3; these, with the alveoli of 
22,7 1,271,722, form, as in the horse, a convex curve at the anterior 
boundary of the upper jaw. 
After a short interval or ‘diastema’ behind 7 3, there is either a 
single alveolus for the bifid base of a canine tooth, or two small 
confluent alveoli for two distinct small, simple-rooted teeth. Prof. 
Burmeister adopts the latter view, ascribes the anterior depression 
to a small single-rooted canine, and the posterior one to a similar 
premolar, which is accordingly the first of that series. Immediately 
behind the empty socket is the first of the premolars in place, with 
a crown equalling in antero-posterior extent the antecedent double- 
pitted alveolus. To judge from the socket of the mandibular tooth 
answering to the first maxillary premolar in place, and from the 
appearances in the side view of the same tooth, in pl. 1, fig. 3, I infer 
that the maxillary premolar, with the mandibular one, was implanted 
by a partially or wholly divided fang, in a two-holed socket; the 
same is more plainly the case in the second premolar in place; 
whilst the third, having acquired a greater transverse thickness of 
crown, may have also a third fang or rudiment of one, on the inner 
side of the two principal fangs. The fourth molar, in place, with a 
further increase of transverse diameter of crown, resembles the three 
succeeding true molars in the general pattern of the grinding surface, 
having an antero-posteriorly extended enamel-lined depression on 
the outer half of the crown, and two round enamel islands, one 
behind the other, on the inner half of the crown. 
Now, my interpretation of the foregoing appearances is, that the 
upper canine was implanted by a compressed, antero-posteriorly 
extended fang, pinched in the middle so as to approach to a division 
of it into an anterior and posterior root, and with a correspondingly 
partially divided socket: it may be that the base of such fang of 
