704 PROCEEDINGS OF SECTION E. 
the lower end should rest on a soft cushion or bundle of blankets, 
whilst, if carried on horseback, they should be on the rider’s back, 
and the horse should not be permitted to go out of a walk. The 
writer always feels safest with them when he is on foot with the 
barometers strapped across his back. The bearer of the instru- 
ments should, of course, be very sure-footed, as, if he fall, the 
barometers are certain to be ruined. 
Aneroid barometers have the advantage of being much more 
portable than mercurials though, of course, they are less reliable. 
The writer finds that the arrangement [suggested by Admiral 
Wharton] of having a small disc placed at the end of the hand 
and perpendicular to the dial of the instrument useful in reducing 
parallax error in readings. Care should be taken to always hold 
the instrument in the same position in reading and to gently tap 
it, thus eliminating instrumental friction. It is advisable to 
keep the instrument in the shade as much as possible when in use. 
The verification of an aneroid barometer is a most important 
point and should be very carefully done. The testing-chamber 
used by the writer at the Melbourne University is arranged 
so that the instrument can be gently tapped prior to read- 
ing whilst being tested, thus the testing conditions are similar to 
the wsing conditions. The ‘“ recovery ” test is an important 
one which the writer applies by exhausting the air-chamber as 
far as possible and restoring full atmospheric pressure again, 
in about two minutes. This is repeated six times, and the varia- 
tion in the correction at full atmospheric pressure is noted. The 
writer has, through the kindness of Professor Kernot, experimented 
with an aneroid of the ‘ Goldschmid” type in which the ordinary 
mechanical magnification has been replaced by optical magnifica- 
tion, but has not found any advantage therefrom, whilst the 
difficulty of verifying the instrument is greatly increased. 
The boiling-point thermometers used by the writer were made 
by J. Hicks, London, and have proved most satisfactory imstru- 
ments. During construction they were subjected to his method of 
fixing the zero, and no trouble has been experienced from change 
of zero whilst in use, though some of the instruments were made 
three years ago. The instruments have an extremely open scale 
(about 1 inch to the degree Fahr.), and are graduated to 0-05", 
so that they can be readily read to 0-01° Fahr., which corresponds 
to about 0-006 of an inch of mercury. During a series of some 
thirty readings taken with each instrument on Mount Juliet in 
December, 1896 (extending over five days), a pair of boiling-point 
thermometers proved to be slightly more sensitive and accurate 
than two siphon mercurial barometers used on the same expedition. 
The great difficulty in connection with the boiling-point apparatus 
is the verification of the thermometers, and a method with which 
the writer is experimenting bids fair to overcome this. Immersed 
