138 
such as Plagiaulax of the Purbeck and Microlestes of the 
Trias (!). The author appears, indeed, to consider that, 
with the exception of Pyrotherium (which, despite its re- 
markable resemblance to Diprotodon, he places in the 
proboscidean line), all mammals with a diprotodont type of 
dentition are related to one another. And he endeavours 
to show that the dentition of one type passes by imper- 
ceptible degrees into that of another. But such gradations 
may be traced between the dentition of almost any groups, 
and no allowance whatever is made for parallelism in 
“development, which has undoubtedly been an important 
factor in evolution. Moreover, no account whatever is 
taken of the undoubted resemblance existing between the 
cheek-teeth of the polymastodonts and the reptilian 
Tritylodon. 
Then, again, according to the author’s scheme, the true 
diprotodonts of Australia have no relationship with the 
polyprotodont marsupials of the same region, which is, on 
the face of it, an absurdity. It may also be pointed out 
that Dr. Ameghino takes no account of the work of other 
paleontologists. It is very generally accepted, for instance, 
that an intimate relationship exists between marsupials (as 
a whole) with the extinct creodonts, and so with the modern 
Carnivora (see Wortman, Amer. J. Sci., vol. xiv., 144, 
1902), while Prof. Osborn (Bull. Amer. Mus., xvi. p. 203, 
1902) has indicated the probability of the descent of the 
rodents from the Holarctic Eocene Mixodectide. Obviously 
both these phylogenies must be demonstrated false before 
there is even a primé facie possibility for Dr. Ameghino’s 
scheme. It will be interesting to learn what the United 
States paleontologists have to say on the subject when | 
the groups in question come to be treated in the working 
out of the Hatcher collection. R. L. 
GEOLOGICAL NOTES. 
(OBSERVATIONS have been made by Mr. R. D. Oldham 
on the growth of sandhills, which threaten to cut off 
communication between the town of Karachi and the suburb 
of Clifton, two or three miles distant (Mem. Geol. Surv. 
India, xxxiv., part iii.). He traces out the growth of 
dunes from small oval patches of sand which begin to 
accumulate on irregular tracts of the stony surface, point- 
‘ng out that even a slight accumulation may cause an 
upward bend of the air currents whereby a space of com- 
Fic. r.—Sandhill near Clifton, Karachi, showing change ot form and scuur 
by wind. 
parative calm is produced, and sand more readily comes to 
rest. In course of time the oval patches of sand are heaped 
up with a sharper slope to leeward, down which the sand 
grains fall. Here a hollow is produced by an eddy of the 
wind, and this eddy serves to maintain and increase a 
crescentic form with a crater-like opening. The principal 
winds at Clifton blow from W.S.W., and form the main 
features in the sandhills; but winds from the E.N.E. blow 
during the winter months, causing a reverse slope and a 
bank of sand to be formed near the summit of the long 
gentle slope which faces the W.S.W. winds. There is a 
NO. 1780, VOL. 69] 
NATURE 
[DECEMBER 10, 1yo3 
good deal of scour of the original steep leeward slope, but 
no complete reversal of the shape of the sandhill. 
Mr. Oldham points out that the original hollow is well 
shown in the accompanying view. The sandhill was first 
shaped by W.S.W. winds, then a period of E.N.E. winds 
caused a partial modification of form, heaping up the sand 
from that side and producing the steep slope facing to left 
of the picture. The sandhill was afterwards attacked by a 
S.W. wind, which commenced to reshape it, and this 
alteration at first led to the formation of notches in 
the crest, in which the wind became concentrated, leading 
to a violent scour and to the excavation of deep pits to 
leeward. The furthest of the notches has been cut down 
nearly to the foot of the steep slope. Eventually it and the 
other notches will be widened, and the intervening pinnacles 
will be lowered until the crest is reduced to a smoothly 
rounded outline. Mr. Oldham discusses the means which 
may be taken to arrest the progress of the sandhills, and 
concludes that much may be done by encouraging the 
growth of local grasses. 
In an essay on the deformation of rocks, Mr. E. H. L. 
Schwarz (Trans. S. African Phil. Soc., xiv., part iv.) dis+ 
cusses their crushing strength, and remarks that this is 
less when the specimen tested is soaked in water. In 
natural circumstances in the earth’s crust the crushing value 
of a column of rock, which would crush the layer at its 
foot, must be estimated by the weight of the material in 
water, and the author calculates that a column of sand- 
stone must be from about two-thirds of a mile to five miles 
in height, one of granite from four to seven and a half 
miles, and one of felsite from seven to nine miles. The 
actual zone of mass deformation seems to be much nearer 
the surface, judging by the ‘‘ creep ’’ in mine-levels, and by 
the fact, in the case of deep bore-holes, that a cylinder of 
rock gradually rises from the bottom. The author alludes 
to the effect that crushing would produce along the bases 
of deep gorges, and he points out that the line of inquiry 
indicates that there must be a limit to the height of 
mountains and to the thickness of ice-sheets. He further 
discusses the deformation of rocks at great depths by the 
action of water. 
In the Proceedings of the Royal Society of Victoria 
(n.s. vol. xvi. part i.), Mr. F. Chapman describes some 
new species of Silurian Polyzoa and Brachiopoda.  Prof- 
J. W. Gregory discusses the formation of the Henty pene- 
plain in N.W. Tasmania. In places it is 1300 feet or more 
ubove the sea, but is lower towards the north, west, and 
south. It appears to have been due to river-action in pre- 
Glacial times, when western Tasmania stood a few hundred 
feet lower than it does now. Its comparatively recent up- 
lift is shown by the King River, which, east of Mount 
Lyell, flows through a very ancient flat-floored valley, and 
then traverses the peneplain in a sinuous narrow canyon. 
An elaborate memoir on the Jurassic Trigoniz of Cutch 
has been contributed by Dr. F. L. Kitchin to the Memoirs 
of the Geological Survey of India (Pal. Ind., ser. ix., vol. 
iili., part ii., No. 1). Most of the species of Trigonia have 
been obtained from the Putchum-Charee series, which, on 
the evidence of Cephalopoda, has been grouped -with the 
Europ2an Bathonian, Callovian, and Oxfordian strata. In 
no case has Dr. Kitchin been able to identify any of the 
Cutch Trigoniz with European species, but while they 
afford no definite evidence of the correlation above men- 
tioned, they present no obstacles to its acceptance. They 
flourished in a different zoological province, but the Lower 
Charee (Callovian) forms bear the imprint of a facies which 
characterised a slightly earlier age in Europe, a fact 
suggestive of migration into the Cutch area. No Trigonize 
have been obtained from the Katrol (Kimeridgian) strata, 
but in the overlying Oomia beds, which appear to be 
transitional between Jurassic and Cretaceous, there are 
Trigonia that approximate in adult characters to forms 
found in the Uitenhage strata of South Africa. There is 
other evidence which suggests community between the Jura- 
Cretaceous faunas of the two areas, but as the forms in 
question differ widely in their youthful characters, Dr. 
Kitchin regards them as indicating homceomorphous deri- 
vation from separate stocks. Evolution of this character 
may have taken place under similar conditions, but it does 
not imply contemporaneity. The subject is of great im- 
