_ Correspondence—T. G. Bonney. 141 
represented by a point in the directions-image. A circular plate is 
ruled with concentric circles at distances from the centre equal to 
r tan 0, where @ stands for different angles at intervals of 5° up to 
the full aperture of the objective and r as a constant length, say 
50mm., and with radiating lines 5° apart; it may be placed on 
supports which fix its position above thestage. When the microscope 
is adjusted for observations of the directions-image of a mineral, the 
point, the angular position of which is to be determined, is identified 
by the end of an adjustable pointer, which is placed so as to be seen 
in focus. The microscope is then focussed up until the objective is 
accurately at a distance + from the plate, which is now placed in 
position and is clearly seen. The angular position required is then 
shown by the position of the end of the pointer relatively to the scale 
of the plate.—L. J. Spencer: A new (seventh) List of Mineral 
Names. 
CORRESPONDENCE. 
Pit Uae 
ON GLACIER LAKE CHANNELS. 
Srr,—As I have stated my case at some length in regard to the 
“overflow channel’ valleys, and as Professor Kendall and Mr. B. 
Smith in their courteous criticisms restrict themselves mainly to 
points of detail, most of which had been noted in my pamphlet, 
while omitting to discuss the fundamental difficulties to which 
I called special attention, I must not weary your readers with more 
than a few words. ach of them, I may remark, has confined himself, 
perhaps wisely, to the district which he has made hisown. Hence 
they have not dealt with the Ringstead Down ‘railway trench’ 
(which is perfect from its head to its end, besides having a lateral 
tributary of more normal form), or with the ancient and more modified 
trenches near Hawes Junction andin the Dufton Pike district, or with 
those in the Cleveland and Black Combe districts, which are occasionally 
perfect or are suggestive of decapitation. Nor have they met my 
objection that the shape of the trenches is not such as should be the 
result of an overflow from an ice-dammed lake (see my pamphlet, 
pp. 7 and 8), for what Professor Kendall does say on this subject 
seems to me hardly to meet my objection. He and Mr. B. Smith 
lay much stress on the fact that spurs are severed by channels 
transverse to the valleys by which those spurs are defined, no doubt 
a thing not easily explained, but they apparently forget that similar 
and similarly situated trench valleys occur in the Dufton Pike district 
and in that to the south of Cader Idris, where the lake overflow 
hypothesis seems to me an impossible explanation. Nor, so far as 
I can see, does the ‘‘ marginal drainage of an ice-sheet’’ really help » 
us. Of course that would erode, but so far as my experience goes 
(and it is a fairly large one) it would find the cutting of a flat- 
bottomed trench very difficult and that of an ‘in-and-out’ impossible. 
Much of what Von Engeln says of marginal drainage is familiar to 
me, and the channels cut by it are not ‘railway trenches’ but nearer 
to gorges. Also, Mr. B. Smith forgets to mention that the granite 
