452 
BULLETIN OF THE NATURAL HISTORY SOCIETY. 
Instruments. — We had with us, ‘and read constantly, three 
aneroids, a small one belonging to Mr. Furbish, another of my own 
which has been used in making the measurements communicated in 
previous years to the Society, and a new Watkin Mountaineering 
Aneroid made by Hicks of London (No. 117, 4| inches in diameter.) 
This instrument, possessing a new device which entirely overcomes 
^he “creeping” error inseparable from the older forms, is the best 
aneroid now manufactured. It was tested for me just before the trip 
by Dr. Harrison at Fredericton, and, although at first it seemed to 
give a considerable error, it was later found that this was due to an 
improper mode of reading it, and when correctly read it gave no 
measureable error. After the trip it was compared by Mr. Hutchin- 
son with his standard instrument at St. John, and found to be without 
appreciable error. Moreover, it has since been examined by its 
maker, Hicks, who reports it in perfect order. It was found to be 
more sensitive than either of the other instruments, and hence its 
readings alone have been used in making the following calculations, 
though the readings of the others have been used as a check. 
Comparison Base. — For this I have used the Fredericton station 
from the beginning of the trip until Little Southwest Lake was 
reached, i. e. August 2nd to 19th, and the Chatham station from 
August 14th to September 2nd. For lists of readings from the two 
stations, I wish to express my thanks to Dr. Harrison and to Mr. J. 
F. Connors, the dominion observers at the two stations. Both stations 
are too distant to form satisfactory bases (both being about sixty to 
seventy miles from the principal places measured), this distance allow- 
ing of considerable error due to difference of weather conditions 
between the places measured and the stations. To lessen this error 
(or indeed to eliminate it), I have devised the following method, which 
applies, however, only where several measurements are made of the 
same locality : The station readings were plotted as polygons (curves), 
in which the abscisste were the dates and times of reading, and the 
ordinates (made long to bring out slight variations) were the barome- 
tric readings. Over these my own readings were plotted upon the 
■same scale, but with the first reading superposed over the first base- 
station reading. It is now obvious that, if the w’eather conditions at 
base-station and the places measured are identical, and the barometers 
read alike, the curves should be coincident throughout, and that the 
