September 9, 1909 J 



NA TURE 



305 



LETTERS TO THE EDITOR. 



[The Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of Nature. 

 No notice is taken of anonymous communications.] 



The Systematic Position of Mceritherium. 



Perhaps of all the groups of mammals at present exist- 

 ing, no two are more dissimilar in general form than the 

 Proboscidea and the Sirenia, but, nevertheless, the sugges- 

 tion made long ago by de Blainville, that these animals 

 are nearly related, has of late years been shown to be 

 almost certainly correct, the discovery of early members 

 of both groups in the Lower Tertiary beds of Egypt having 

 to a great extent bridged the gap between the two orders. 

 The animal that perhaps most nearly approaches the 

 ancestral type of the two groups is Mceritherium, which, 

 however, has hitherto been regarded as already far 

 advanced along the proboscidean line, and if not directly 

 ancestral to the undoubted primitive elephant Palaeo- 

 mastodon, at least closely related to its ancestor, and re- 

 presenting a stage of evolution through which it must have 

 passed. 



In an article " On the Feeding Habits of Mceritherium 

 and PaliEomastodon," published in Nature of July 211, 

 Prof. H. F. Osborn seems to dissent from this view to 

 some extent, to emphasise the Sirenian characters of 

 Mceritherium, and to regard it rather as a belated 

 Sirenian which had not undergone the specialisation 

 for a purely aquatic life already attained by some 

 Sirenians contemporary with, or even earlier than, it. 

 Prof. Osborn to a large extent bases his conclusions 

 on peculiarities shown in some .more or less con- 

 jectural restorations of the heads of Mceritherium and 

 Palaeomastodon, his chief reasons for the separation of 

 Mceritherium from the Proboscidea being (i) the small 

 size and high anterior position of the eyes and the high 

 position of the ears, both said to be characteristic of the 

 Sirenia ; (2) the difference in the arrangement of the 

 anterior teeth and mouth-parts from that found in 

 PaUeomaslodon. 



As to the characters of the eyes and ears, they seem 

 to be purely adaptive, and are simplv the result of the 

 admittedly semi-aquatic habits of Mceritherium, and would 

 not be expected to exist in purely terrestrial members of 

 the group. The apparent height of the ears is, moreover, 

 mainly the consequence of the small development of the 

 occipital region of the skull compared with that found in 

 the heavier-headed Paljeomastodon. As to the arrangement 

 of the jaws and the anterior teeth, it seems to represent 

 exactly such a stage as a mammal with the normal 

 Eutherian dentition would be expected to pass through 

 before attaining the condition found in Palreomastodon. 

 Certainly the anterior dentition of Mceritherium is quite 

 unlike that of any known Sirenian ; thus in Eosiren, a 

 contemporary of Mceritherium, it is the first pair of upper 

 incisors, not the second, that is enlarged, while the other 

 incisors and the canine are alreadv in a fair way to the 

 complete disappearance characteristic of the later Sirenia. 

 In the lower jaw the differences are greater still, Eosiren 

 possessing the strongly deflected symphysis probably already 

 partly covered with a horny plate, and in which the 

 incisors are undergoing reduction : in Mceritherium, on the 

 other hand, there is no deflection of the svmphvsis, and the 

 first pair of incisors is small, while the second form large 

 procumbent tusks, very similar to those found in many 

 primitive Proboscidea. Mceritherium is further dis- 

 tinguished from the contemporary and even earlier 

 Sirenians by the possession of a well-developed pelvis, which 

 was supported by a strong sacrum composed of three 

 fused vertebra;. The hind limb is, unfortunately, not 

 completely known, but the femur was large and straight, 

 being similar in many respects to that of Palaeomastodon. 

 In several other ways, also, Mceritherium differs from 

 the Sirenia and approaches the Proboscidea ; thus the 

 position of the auditory meatus with reference to the 

 zygomatic process of the squamosal and to the neighbour- 

 ing bones is as in PaliEomastodon, and quite unlike what 

 Is seen in the Sirenia. ."Xgain, the cervical vertebrae, 

 NO. 2080, VOL. 81] 



though short, show no traces of the extreme shortening 

 already present in Eosiren. 



Another argument in favour of the relationship of 

 Mceritherium to Pateomastodon is the existence of forms 

 like Moeritherium trigonodon and Palaeomastodon minor, 

 which, though unfortunately at present very imperfectly 

 known, appear, both in their size and in some respects 

 in their tooth structure, to be annectant forms. 



On the whole, it seems that the weight of evidence is 

 in favour of regarding Mceritherium as a proboscidean, 

 though perhaps not on the direct line of ancestry of 

 PaIa?omastodon, and retaining some characters of the 

 original Proboscideo-Sirenian stock. 



Chas. W. Andrews. 



Remarkable Halo of August 21. 



The accompanying diagram is a sketch of the Danzig 

 phenomenon as observed at Blackpool on Saturday, 

 .August 21, between 11.45 a.m. and 0.30 p.m. There 

 had been heavy rain and stormy winds on the previous 

 night and in the early morning up to 10 a.m. At 

 U.45 a.m. thunder cloud and cumulus extended all around 

 the horizon, but in the vicinity of the zenith, where the 

 22° halo and the western portion of the mock sun ring 

 now appeared, there was no trace of cirrus or other cloud 

 form, although a very pale milky tint might be discerned. 



At noon the sky was overcast, but by 0.25 p.m. it was 

 again clear at the zenith, and the complete phenomenon 

 stood out very prominently, the 22" halo very vivid and 

 brilliant with what would appear to be its arc of upper 

 contact, the mock sun ring being very clear and of a silvery 

 hue, and the mock suns ; the two furthest from the real 

 sun being apparently at the intersection of the mock sun 

 ring with the 90° halo. This halo, however, was nowhere 

 else visible. 



There was nothing like an image of the sun at the 

 N. point of the 22° halo ; but this was a position of very 

 great brilliancy, with the usual reddish colouring on the 

 edge nearest the sun and the gradual shading off outwards 

 to a bluish-white. The S. point was also a position of 

 maximum brightness, though less intense than the N. 

 point, while the E. and W. portions of the halo were 

 fainter and untinted, but quite discernible. 



The two nearer mock suns at .A and B likewise could 

 hardly be called images of the sun, but resembled the N. 

 point of the 22° halo, as though they were the inter- 

 sections in the mock sun ring of another halo, nowhere 

 else visible, of about 33° radius. They were thus some- 

 what elongated, projecting slightly on either side of the 

 ring. The two further mock suns were of a similar 

 character, but with no colour. 



Some special peculiarities seem to have been : — (i) The 

 absence of cloud in the region of observation where the 



