November 20, 1896.] 



SCIENCE. 



747 



the complete typical formula, while the 

 permanent dentition shows a very marked 

 reduction. Also in Lepidole7nur, in which 

 all the superior permanent incisors are want- 

 ing, one incisor is preserved in the milk 

 dentition. 



''Almost without exception the milk teeth 

 of the Lemurs are smaller and weaker 

 than the corresponding permanent teeth. 

 If the permanent dentition reaches a higher 

 grade than the dentition it is explained by 

 the fact, as I have already shown, that the 

 latter has undergone a more or less pro- 

 nounced differentiation in the size of its 

 individual components ; this is the case in 

 Tarsius, Indrisince and Chiromys, without in 

 the least diminishing the original number 

 of the teeth. 



" So far as I have considered the phylo- 

 geny of the different teeth, whilst among 

 the Insectivora and the mammalia of the 

 secondary period, and in exceptional cases 

 among the living forms such as the Mar- 

 supial Clioeropus and the fossil PaleocJioerus, 

 canines are observed with double roots— a 

 character which is certainly to be regarded 

 as primitive — I have found in the Lemurs, 

 both in the milk and permanent dentition, 

 two-rooted canines. The fact that often a 

 one-rooted milk canine is replaced by a 

 two-root permanent canine, and this order 

 in other cases is reversed, requires further 

 clearing up. 



" That an elongate or more premolar-like 

 structure of the superior canine is the or- 

 iginal form of this tooth in the Lemurs, ap- 

 pears to be evident in every case in which 

 the permanent canine differs from the milk 

 canine ; for the milk tooth is always more 

 like a premolar than the permanent tooth, 

 as seen in the comparison of Chirogaleus, 

 Adapts and Tarsius. A comparison of the 

 canine of the old tertiary form, Microchoerus, 

 with that of the modern Tarsius, lead us 

 to the same results." 



H. F. 0. 



CURRENT NOTES ON PHYSIOGRAPHY. 

 ORIGIN OF THE LAURBNTIAN RIVER SYSTEM. 



IJpHAM continues his discussion of the 

 great lake problem (Amer. Geol., XVIII., 

 1896, 169-177), maintaining that during 

 Tertiary time the Mississippi-St. Lawrence 

 divide probably lay northwest of the Adi- 

 rondacks, in this differing from Spencer, who 

 regards the preglacial St. Lawrence as al- 

 ready an extensive river system. Certain 

 general relations of our larger land forms 

 and river systems would, however, seem to 

 prove the extension of the preglacial St. 

 Lawrence at least into the Ontario basiu. 

 All the Great Lakes, except Superior, lie 

 along the inner lowlands, and are enclosed 

 by the infacing uplands* of an ancient and 

 greatly denuded coastal plain of paleozoic 

 strata, whose oldland is the Laurentian 

 highland. The great Appalachian valley 

 is also an inner lowland, between the inface 

 of the Alleghany and Cumberland plateau 

 and the oldland of the Blue E.idge ; but 

 this inner lowland is complicated by the 

 mountains that have been bent up and 

 worn down along it. The normal drainage 

 of both these regions would be from the old- 

 land across the inner lowland and out 

 through the scarped uplands to the Ohio or 

 Mississippi. The Wisconsin and the Kan- 

 awha rivers are exceptional in still preserv- 

 ing this normal course. The Potomac, Sus- 

 quehanna, Delaware and Hudson are all 

 abnormal in flowing from the Alleghany 

 plateau across the inner lowland and out 

 through the oldland to the Atlantic. ISTow, 

 as these abnormal courses had been attained 

 in early Tertiary time, and perhaps sooner, 

 it is not only possible, but probable, that a 

 considerable part of the abnormal drainage 

 area of the St. Lawrence had been de- 

 veloped much earlier than Upham main- 

 tains. 



* The Spanish term cuesta might be used for this 

 unnamed form. See Hill, Nat. Geog. Mag., VII., 

 1896, 295. 



