632 



NATURE 



[Oct. 30, 1! 



distinguished by the presence of well-developed dicotyledonous 

 leaves" ; and further on these are said to occur at "the very 

 base" of the formation. Now though such statements are 

 sometimes loosely made, it should be understood that American 

 geologists generally acknowledge that the base of their Creta- 

 ceous is, in some localities at least, only equivalent to the base 

 of the Upper Cretaceous of Europe. In Canada, at least, the 

 strictly Mesozoic flora of the Lower Cretaceous has been clearly 

 distinguished from the angiospermous flora of the middle ami 

 upper parts of the series. 



The oldest Cretaceous beds known in Canada are, I believe, 

 those of the Queen Charlotte Islands, referred by Mr. Whiteaves, 

 on the evidence of animal fossils, to the Neocomian age. The 

 flora of these, consisting of cycads and conifers only, with- 

 out any trace of dicotyledonous leaves, was described by me in 

 the Report of the Geological Survey for 1S73, and I remarked 

 at the time on its decidedly Mesozoic aspect. It will be seen 

 by reference to my memoir on the Cretaceous floras of British 

 Columbia and the North-West Territories, in the Transactions 

 of the Royal Society of Canada for 18S3, that the oldest an- 

 giospermous flora known at that time in Western America is 

 that of the Dakota group, described by Lesquereux and supposed 

 to be of Cenomanian age. We have not yet found any dicoty- 

 ledonous leaves quite so old in Canada. Our oldest angio- 

 spermous flora occurs in beds referred by Dr. G. M. Dawson 

 and Mr. Whiteaves to the Niobrara group, which is approxi- 

 mately of the age of the Chalk Marl of Kngland, in so far as 

 can be judged by its animal fossils. A detailed table of the 

 beds is given in the memoir above referred to, and the facts are 

 stated in general terms in the "Descriptive Sketch" of the 

 geology of Canada which was distributed to the members of the 

 British Association (p. 51). 



It will thus be seen that, though our angiospermous flora may 

 possibly have appeared somewhat earlier than that of Europe, 

 the discrepancy is by no means so great as stated in the abstract 

 referred to. The correct statement would be, in so far as 

 Canada and the western parts of the United States are con- 

 cerned, that the oldest angiosperms known in America are 

 probably of Cenomanian age, and that the older Cretaceous 

 contains only, so far as known, a flora of Mesozic character. 

 Concerning the limits of the Cretaceous and the Eocene on the 

 one hand, and the limits of the Cretaceous and Jurassic on the 

 other, there are no doubt some unsettled questions ; but these 

 do not affect the facts above stated. J. Wm, Dawson 



Montreal, October 9 



Sir J. W. Dawson's correction only applies to the published 

 abstract of my paper. The editor of the Geological Magazine 

 having kindly offered to publish the full text, it will be seen that 

 its scope was limited to Cretaceous dicotyledonous floras, and 

 the older ones, to which Sir John calls attention, were pur- 

 posely excluded. The title " Cretaceo-Eocene " was intended 

 to imply that the subject was the border-land of these two forma- 

 tions ; but I am greatly obliged for the note and the copy of the 

 work which accompanied it. J. S. G. 



Palaeolithic Implements from Cambridge 



Only two implements of Palaeolithic age have been recorded 

 from the neighbourhood of Cambridge. One of these is a rude 

 form picked off a heap of gravel near the Observatory, and the 

 other was bought from some workmen, and was said by them 

 to have come from the Barnwell gravel. There is therefore 

 considerable interest attached to the discovery of an implement 

 of this age on the plateau between Upper Hare Park and the 

 Cambridge Newmarket Road. This plateau is part of one of 

 the old river terraces which formerly abutted against the hills 

 on the east, but is now cut off from them by the valley along 

 which the railroad to Newmarket runs. It belongs to an earlier 

 period than that of the Barnwell gravel. 



Further to the south, near Lark's Hall, in gravels which pro- 

 bably belong to the same set of river terraces, remains of rhino- 

 ceros, &c, have been found, but hitherto no implements or other 

 traces of the existence of Palaeolithic man have been brought to 

 light in that district. 



The plateau near Upper Hare Park is all unfenced arable 

 land, and the implement which I found buried in the surface soil 

 with only a small part of its thicker end visible, had probably 

 been turned up out of the gravel by the plough, its surface 

 having the same general appearance as the flints derived from 

 the gravel. It is of the tongue-shaped St. Acheul type, and has 



a fine patinated surface. It measures $\ inches in length, 3 

 inches across its broadest part, and nearly 1^ at its thickest. 

 One end is rounded so as to be easy to hold in the hand, and 

 from this it tapers gradually with a sharp cutting edge to the 

 point. On each side of the implement the edge is curiously 

 rough and shattered, owing to the original quality of the flint 

 and the way in which the flakes broke off when it was being 

 made. 



Not very far from the spot where it was found, skirting the 

 carriage-drive which leads up to the house, are several small 

 hollows from which gravel has most likely been dug. In these 

 hollows and all round their margin the ground is covered with 

 flints belonging to the gravel, and amongst them I found 

 quantities of flint chips, one or two very nicely-dressed flakes, 

 and s< veral scrapers. 



Similar dressed flakes and scrapers are found in the caves of 

 the South of France associated with implements of Palaeolithic 

 age ; but in the case of those found near Upper Hare Park 

 there is not sufficient evidence to prove whether they belong to 

 the period of the gravel or were manufactured on its surface at 

 some later period. M. C. Hughes 



The Recent Lunar Eclipse 

 1 wonder whether any of the readers of Nature who were 

 witnesses to the almost total annihilation of the moon on Saturday 

 night, October 4, noticed a rather strange peculiarity which was 

 visible at about 10.50 p.m., both before and after second internal 

 contact with shadow. When the peculiarity first appeared I 

 cannot myself say, but I noticed it first at 10.43 when I went out 

 to look for the almost invisible moon with the aid of a good 

 opera-glass. In the accompanying diagram, which I have con- 

 structed from the data given in the almanacs, the moon is 

 represented as just having emerged slightly from the shadow 

 at 10.50 or so, when the peculiarity showed very distinctly, 

 the moon having the appearance which is roughly represented 



in the diagram, being apparently divided into two halves 

 by a tolerably distinct line of demarcation (be) running north 

 and south (or towards the celestial pole), the right hand or 

 westerly half appearing much darker than the left or easterly 

 half. It is evident that an appearance like this, so striking when 

 once noticed, could be produced in two ways, first, by the 

 western hemisphere of the moon being actually darker than the 

 left or eastern half ; in which case the moon would have exhi- 

 bited this appearance more or less throughout totality ; but it 

 did not, as 1 noticed nothing of the sort at 10.15, when looking 

 through the same glasses, so that the second explanation must 

 be resorted to. In the diagram the larger outer circle repre- 

 sents the border of the earth's shadow (in the case of this eclipse 

 about 5750 miles in diameter) which is cast by the earth, irrespective 

 of its atmosphere. The inner circle represents the border of an 

 inner and darker shadow of the earth, cast by those of the sun's 



