Nov. 14, 1878] 



NATURE 



3t 



the northern quarrie?, which was largely made up of these 

 fragments, probably to the amount of hundreds of tons ! It 

 will be seen from my paper (loc. a'i.) that the Sombrero Key is 

 merely the eroded remnant of an atoll, which presented, by its 

 peculiar isolation, the most favourable refuge for this Chelonian 

 family. Their existence in this region seems to furnish an im- 

 rtant link in reference to the question of the ancient geo- 

 raphical connection between the Mascarene and the Galapagos 

 I lands. Their occmrence has never been reported on the other 

 guano keys of the Caribbean Sea — Navassa, Swan, Monk's, 

 Redonda, &c ; but I would suggest the propriety of an exami- 

 nation of the cargoes of guano from Navassa, &c., as well as 

 from all other localities, which may still be brought to England. 

 The conditions of the existence of this fauna will be discussed 

 with more detail in the remainder of my Sombrero paper, when 

 sufficiently restored and re-written for pnbL'cation. 



Alexis A. Julien 



My Dear Sir,— The collection of fossils from the Sombrero 

 Key, which you kindly placed in my hands for examination, 

 comprises the remains of birds, turtles, and Saurian reptiles. 

 Those of the first, and from the more superficial deposit, appear 

 to belong to existing species of sea-b'.rds now foimd along the 

 coast, and have midergone no change. The following comprises 

 a list of such specimens as are sufficiently well preserved to be 

 identified. There are many other specimens, but they are so 

 much broken that I have not been able to determine them. 

 Specimens numbered, from I to 17 inclusive are bones of turtles, 

 and those from I to II inclusive are all from the same, viz., the 

 middle, deposit.^ 



1. Left humerus of a large turtle ; the ends are gone and the 

 shaft alone is preserved ; it is 6 inches in length, and at the 

 smallest part has a diameter of 2*07 inches, from before back- 

 wards, and of I '57 inches from side to side. Admitting that 

 the usual proportions existed, the whole length would have been 

 about 12 inches. 



2. A fragment comprising a large part of a humerus ; nearly 

 the whole of the articular facet is preserved, its longest diameter 

 measuring 2 '38 inche-, thus indicating a large species like that 

 of I. 



3. Middle portion of a femur, l '25 inch in diameter. This 

 w ould indicate an entire bone of from 8 to 9 inches in length. 



4. Lower third of a left humerus ; the inner tuberosity is 

 broken off, and does not therefore exhibit the emargination found 

 in most tmtles. This is somewhat smaller than the same part 

 from a Galapagos turtle { T. elephantopus), the length of whose 

 carapace was 27 inches. 



5. A femur from which the upper portion is broken off. This 

 ' of the same size as the corresponding part in the Snapper (£". 



pentina), the carapace of which was 15 inches long. The 

 .ransverse diameter of the condyles was i'35 inch. 



6. Middle portion of the shaft of a humerus, 0*55 inch in 

 diameter. 



7. A similar fragment, 0*75 inch in diameter. 



8. A fragment of one of the marginal bones, i inch in thick- 

 ness. This could have only belonged to a turtle of the size of 

 the Galapago? species. 



9. A fragment of the right ilium, including the middle por- 

 tion, the longest diameter of which last is 1*55 inch, and the 

 two facets for the articulation of the ischium and pubes. 



10. The shaft of a femur 0*75 inch in diameter. 



11. Upper half of the ascending portion of the scapula; this 

 has a diameter at the articular end of 0*90 inch. 



All of the above specimens are from the same deposit in 

 which the matrix is soft and crumbling. 



12. Lower two-thirds of a humerus partly imbedded in 

 a very dense matrix, which contrasts very strongly with that 

 found in connection w ith the preceding specimen. The frag- 

 ment is 3*50 inches long, 2*35 inches across condyles ; a part of 

 the inner one is broken off, but there are some signs of the 

 lateral groove and notch at the end. The narrowest part of the 

 shaft measures 0*92 inch in diameter. 



i2- Lower end of a radius from the left side. 



14. Fragment of a carapace having a raised articular surface 

 for the articulation of the pelvis, as in some of the land turtles. 



15, 16, 17. Other fragments of a carapace. 



17a. An irregular cast of a part of the interior of the carapace 



and plastron, the walls of the latter (?) being broken, but portions 



of them still adhering. Three of the median bony dermal plates, 



the largest 2*25 inches in diameter. These plates have neither 



' The Red Sandrock.— A. A. T. 



ridge nor tuberosity on the n.edian line, though there are slight 

 projections over the head of the ribs resembling those of very 

 old specimens of E. serpentina. The plastron is indistinctly 

 seen, but enough remains to show that its union with the cara- 

 pace was by a broad surface, and not by a narrow one, as in the 

 marine and some of the fresh-water species. 



18. A caudal vertebra of a saurian. 



19. Anterior half of a lower jaw of a saurian which resembles 

 in size and the arrangement of the teeth that of the Iguana cor- 

 nuta, as figured by Cuvier in the Ossemens fcssiles. The points 

 of the teeth are compressed and show some signs of having been 

 serrated, though now worn nearly smooth. 



20. Fr^ment of a femur, which closely resembles in size 

 and shape that of the Iguana tuberculata. 



2.1. Another fragment of a femtir closely resembling the last> 

 but about one-fourth smaller. 



The remains of tiu-tles form by far the largest part of the 

 collection of which the above is a list. From the fact which yon 

 communicated to me, that at the present time not a single species 

 of turtle inhabits the Sombrero Key, the question at once arises 

 w hether these remains belong to species like those now inhabiting 

 the sea, or to such as live either in fresh-water or on the land. 

 After a careful comparison I do not find that any of them can 

 be considered as of marine origin. All the long bones, consist- 

 ing chiefly of arm and thigh bones, differ in a marked degree 

 from corresponding ones of the sea-turtles in having the axis of 

 the bones strongly curved instead of being nearly straight, in 

 having the shaft at its middle nearly round instead of flattened, 

 and in having the distal ends proportionally much broader. A 

 comparison of the fossil fragments with the corresponding parts 

 of fresh-water and land species is much more difficult, since 

 these two kinds, in their anatomical features, so gradually shade 

 into and so closely resemble each other that there is really no 

 well-marked line of distinction. 



It is certain, nevertheless, that the remains above noticed 

 belong cither to the fresh-water or land species, and the dis- 

 covery of them where such no longer exi?t alive indicates a great 

 revolution in \!ht prnnous history of the i-land, and is therefore a 

 marked Jact. In addition we have the interesting remains of one 

 of the species, which is certainly extinct and of gigantic size, 

 equalling the largest specimens which are found fiving in any 

 part of the world, and thus surpassing any now foimd in North 

 or South America. The nearest ia-tances of tmrtles of similar 

 size are in the Galapagos Islands, where is foimd /. dephan- 

 topnis. Specimens 7 and 10 indicate species as large as those now 

 iiiiabiting the Americas. 



Although among turtles it is almo.-t impossible to establis 

 species from fragments of bones, and these not the most charac- 

 teristic ones, yet I have no doubt that the remains here described 

 show the exi:>tence at least of three species, one, the longest of 

 which represented by specimen I, was undoubtedly an inhabitant 

 of the land. Jeffries Wyman 



Cambridge, August 14, 1865 



The Figure of the Planet Mars 



In the report of the proceedings of the Academy of Sciences 

 at Paris for October 22 (Nature, vol. xviii. p. 712), with refer- 

 ence to a communication from me relative to the flattening of 

 the planet Mars, it is stated that I confirm M. Amigues' conclu- 

 sions from independent calculations. Allow me to say that the 

 communication referred to, has clearly established by reference to 

 dates of publication, that the calculations I had been the first to 

 make were confirmed by the subsequent results of M. AnMgues. 

 A formula presented by me in February, 1870, in which the 

 mean density, surface density, and velccity of rotation of Mars 

 are expressed in connection with its ellipticity, was reproduced 

 by apparently identical methods by M. Amigues, in the Comptes 

 Rendus for Jtme, 1874. The conclusions drawTi from this 

 formula by M. Amigues were, that in order to account for the 

 high amount of ellipticity assigned to Mars by many astronomers 

 its mean density must be less than its stirface density. My con- 

 clusion was, on the contrary, that the high ellipticity alluded to 

 was improbable and that the values given by Bessel, Ondemanns, 

 Johnson, and other astronomers, whereby Mars would have an 

 ellipticity nearly the same as that of the earth, should be adopted 

 until the subject was cleared up by fresh observations. 



H. Hennessy , 



Roval Collesfe of Science for Irclatd, November 2 



