370 
APPENDIX. 
a mercurial barometer are self-evident. I managed to carry 
a mercurial barometer from Lahore to Yarkand, and got it 
safely back as far as Le, the entire distance being about 
1500 mileSj but on previous occasions I never succeeded in 
keeping my barometer safe for many days on the march. 
It may be objected to the self-registering arrangement 
that the thermometer is not quite so accurate as it would 
otherwise be^ but this objection is not considered of any 
weight in the case of clinical thermometers,, which are 
usually made self-registering. And in estimating heights 
roughly for the march, either by a mercurial barometer or 
by a thermometer^ one must always allow for a possible 
error of one or two hundred feet. 
It seems almost unnecessary to remark that boiling-point 
thermometers and aneroids sent to India should be capable 
of measuring heights up to at least 20,000 feet. 
B. 
A sample of water taken from the western end of the 
Pangong Lake was very kindly examined for me by Dr. 
Frankland^ who gives the following as the results of his 
analysis : — 
100^000 parts by weight contain — 
Carbonate of lime 2*10 
Carbonate of magnesia .... 22'38 
Silica 3-60 
Alumina, with a trace of iron . 5*40 
Sulphate of lime 4*40 
Magnesia 50*88 
Sulphate of magnesia .... 99*38 
Chloride of sodium 591*39 
Sulphate of soda 410*31 
Chloride of potassium .... 107*84 
Chloride of lithium . . . . . a trace 
Organic matter not determined . 
1297*68 
THE END. 
