T. O. Bosworth—Wind Erosion in Mull. 303 
of each genus the distinction holds good. The very fact of the over- 
lapping would tend to confirm the supposition that this ‘ plate-crushing’ 
is of genetic value, and that parallel development takes place among 
the several genera. 
In the case of Conulus albogalerus, Leske, from the Upper Senonian, 
the crushing is markedly earlier in its origin. So much so that, 
whereas in Holectypus and Discoidea the ambulacra on the adapical 
surface of the test seem almost entirely composed of primaries, in 
Conulus the great bulk of each ambulacrum is built of crushed plates. 
The first indication of crushing in set ¢ appears at plate 9, and in 
set 6 at plate 20. But while in set ¢ the plates become small demi- 
plates inserted in the outer angles of the others by the time plate 15 
is reached, in set 4 the primary character lingers on until plate 42. 
Thus the greater part of the median ambulacral suture is made up of 
the edges of plates belonging to sets a and b. This perpetuation of the 
primary plates in set 6 would appear to be a special feature developed 
in the course of evolution of the genus, for in C. subrotundus from the 
Cenomanian, while the plates of set ¢ begin to diminish in size at 
number 12, and those of set 6 at number 20, the latter set of plates 
becomes definitely separate from the median suture at number 43. 
There seems, therefore, to have been in the history of the development 
of the Holectypoida a tendency to introduce ‘plate-crushing’ in the 
ambulacra to an increasing extent. As this feature may very likely be 
connected with two other noticeable. characters of the progressive 
development (increase in the height of the adapical surface of the test 
and narrowing of the ambulacra), it should serve as a useful index to 
the genetic relationships of the various genera and species. As I have 
not sufficient material to investigate the ambulacral structures of the 
other genera of the order, nor the opportunity of collecting sufficient 
numbers of specimens for working out specific relations within the 
various genera, I have thought that the above indication of the 
possibility for important evidence on this branch of evolution might 
perhaps induce others, more favourably situated in these respects, to 
continue and amplify the study. 
TV.—Wuinp Erosion on tHe Coast oF Mutt. 
By T. 0. Bosworrn, B.A., B.Sc., F.G.S.1 
(PLATES XXVIII AND XXIX.) 
(J\HE Ross of Mull is a comparatively low-lying peninsula exposed 
to the storms and swept by gales from the Atlantic. On these 
rocky shores are occasional stretches of sand which frequently have 
been driven inland, forming dunes and patches of blown sand at 
intervals around the coast. Sometimes the sand is forced up the cliffs- 
and deposited more than 100 feet above the sea, and there are places 
in the centre of the peninsula where the blown shell-material is quite 
a noticeable constituent in the soil. 
In one instance, the peculiarities of which are here to be described, 
1 With permission of the Director of the Geological Survey. 
DECADE V.—VOL. VII.—NO. VIII. 23 
