302 



NA TURE 



[July 28, 1904 



by the molars of the numerous types of ungulates 

 peculiar to the Tertiary formations of South America, 

 Dr. Ameghino has almost a superfluity of material 

 upon which to work. So vast, indeed, is his subject, 

 that it would be impossible, within the limits of our 

 space, to follow him in his survey from one type to 

 another, or, indeed, to discuss his general con- 

 clusions, and I shall therefore confine myself to direct- 

 ing the attention of my readers to certain points of 

 special interest in the author's work. 



In the first place, it may be noted that Dr. Ameghino 

 reiterates his opinion as to the falsity of the tritu- 

 bercular theory of molar development, tritubercular 

 molars, instead of being the primitive type, having 

 been derived, on liis view, from those with four or six 

 cusps. Whether this opinion is in any way biased 

 by the author's contention that the earlier Patagonian 

 mammals are of Cretaceous age may be worth con- 

 sideration. 



Of greater importance is the support given by Dr. 

 Ameghino to the view that the molar formula of the 

 placental and marsupial carnivores is numerically 

 identical, that is to say, that there are three true 

 molars in both (when the full series is developed) 

 instead of three in the one and four in the other. He 

 consequently regards the replacing marsupial premolar 

 as the third instead of the fourth, and the tooth behind 

 it as a persistent milk-molar. The numerical identity 



Fig. I. — Right upper molar 

 of horse. /, antero-in- 

 ternal pillar ; hy\ pos 

 tero-internal pilHr. 



of the marsupial and placental series was urged long 

 ago by Prof. Gaudry, and this view was more fully 

 developed a few years since by the present writer, 

 when evidence in favour of the above-mentioned homo- 

 logy of the marsupial replacing tooth and the one 

 behind it was likewise adduced. Although it has sub- 

 sequently received the assent of Dr. Wortman, this 

 view has not, however, yet been accepted by zoologists 

 generally. It may be added that it is a question on 

 which only those with a considerable amount of 

 palseontological knowledge are competent to form an 

 opinion. 



Another point of great interest referred to in Dr. 

 Ameghino 's memoir is the alleged occurrence in 

 Nesodon and certain other Patagonian Tertiary 

 mammals of three distinct dentitions. Since the 

 existence of this remarkable phenomenon is stated to 

 have been accepted by Dr. Scott, it may apparently 

 be taken as fully authentic. The " pre-lacteal " 

 cheek-dentition, according to the figure given, consists 

 of three very small teeth, differing somewhat in form 

 from their successors of the milk-series. Dr. 

 Ameghino, whose view is almost certainly in this 

 instance influenced by his opinion as to the age of 

 the Patagonian mammals, regards the " pre-lacteal " 

 dentition as a direct inheritance from reptilian 

 ancestors. In view, however, of the specialised 

 characters of Nesodon and its allies, and the absence 

 NO. I 8 13, VOL. 70] 



of a functional " pre-lacteal " series in any other 

 mammals, it would seem much more probable that it 

 is a superadded feature. 



The last point to which we have space to allude 

 relates to the homology of two of the cusps in the 

 equine molar. To render this point clear, two figures 

 have been introduced into this notice. 



Since the date of publication of the first part of 

 Gaudry's " Enchainements," the antero-pillar of the 

 equine molar (/> in Fig. i) has been almost universally 

 regarded as one of the primitive constituents of the 

 tooth, corresponding to the inner extremity of the 

 anterior transverse ridge (protocone) of the rhinoceros 

 or anchitherium molar. From comparison with a 

 large number of extinct forms. Dr. Ameghino comes, 

 however, to the conclusion that this antero-internal 

 pillar (which is detached in Hipparion but joined to 

 the body of the tooth in Equus) is really a superadded 

 element, derived from the cingulum, and correspond- 

 ing to the " accessory pillar " of the molars of many 

 ruminants (Fig. 2). Consequently, the protocone willi 

 be represented by a part of the anterior inner crescent 

 of the horse's molar. Apparently Dr. Ameghino has' 

 made out a very strong, if not a conclusive, case fori 

 the new interpretation ; I may add that the same, 

 opinion was independently arrived at and published by 

 Dr. Forsyth-Major so long ago as 1873, but has been 

 generally neglected in favour of the Gaudrian theory. 



Without in any way endorsing all his views, it mayi 

 be confidently stated that in this memoir Dr.j 

 .\meghino has made a very important contribution to 

 mammalian odontology. R. L. 



DR. ISAA.C ROBERTS. F.R.S. '. 



^pHERE is one class of scientific amateurs whicb 

 •'• seems to be the peculiar product of English 

 society. Dealing with astronomy alone, and confining 

 our attention to those who have passed away, we have 

 such men as Lassell, Barclay, De La Rue, &c., all of 

 whom, after amassing a considerable fortune in com- 

 mercial pursuits, have devoted the evening of their 

 lives to furthering the interests of their favourite 

 science. The latest example of this earnest attachment 

 to this particular branch of science was Dr. Isaac 

 Roberts, whose death we record with profound regret. 

 It is possible that he may be nearly the last of a dis- 

 tinguished series, for it is not unlikely that, as science 

 tends to specialise in particular directions, such 

 instances w'ill become less and less frequent. The 

 wealthy amateur, it may be, will continue to provide 

 the means for others, but the requirements for the 

 production of valuable work tend to become more and 

 more severe, and the actual prosecution will soon be 

 reserved to those who have been able to give up their 

 whole life to special study. But Dr. Roberts was 

 fortunate in finding a subject at which he could work 

 with effect personally, and his own exertions were 

 rev/arded with valuable results. 



For some years Dr. Roberts seems to have wavered 

 between geology and astronomy as a congenial pursuit 

 in his leisure hours. But possibly it was the appli- 

 cation of some form of mechanical inquiry that 

 attracted him in either direction. Among his early 

 papers are the results of investigations affecting the 

 circulation of underground water and the filtering and 

 hygroscopic properties of Triassic sandstone. In 

 studying the movements of underground waters, of 

 which the observations were carried on with great 

 regularitv, he employed mechanical contrivances which 

 he designed himself. Similarly, in practical questions 

 such as the determination of the pressure of grain 

 on the walls of lofty warehouses, when stored to great 

 height, the mechanical side of the question seems to 



