138 THE I^B-EX ERRORS OF THERMOMETER SCALES. 



tion of liquor, that even their own officers can only procure a small 

 allowance, given as part of their annual outfit on voyages. 



"Without entering into the general question of the policy of giving 

 a monopoly of the Fur trade to one company, I cannot but record as 

 the firm conviction which I formed from a comparison between the 

 Indians in the Hudson's Bay Company territories and those in the 

 United States, that opening up the trade with the Indians to all who 

 wish indiscriminately to engage in it must lead to their annihilation. 

 For while it is the interest of such a body as the Hudson's Bay Com- 

 pany to improve the Indians and encourage them to industry accord- 

 ing to their own native habits, in hunting and the chase, even with 

 a view to their own profits ; it is as obviously the interest of small 

 companies and private adventurers to draw as much wealth as they 

 possibly can from the country in the shortest possible time, although 

 in doing so the very source from which the wealth springs should 

 be destroyed. The unfortunate craving for intoxicating drinks, which 

 characterises all the tribes of Indians, and the terrible effects thereby 

 produced upon them, render such a deadly instrument in the hands 

 of designing men. It is well kuown that, although the laws of the 

 United States strictly prohibit the sale of liquor to the Indians, it is 

 impossible to enforce them, and whilst many traders are making 

 rapid fortunes in their territories, the Indians are fast declining in 

 character, wealth, and numbers, whilst those in contact with the 

 Hudson Bay Company maintain their numbers, retain their native 

 characteristics unimpaired, and in some degree share in the advan- 

 tages which civilization places within their reach. 



A METHOD OF DETERMINING THE INDEX ERRORS 

 OF THERMOMETER SCALES. 



BY W. D. C. CAMPBELL, QUEBEC. 



Read before the Canadian Institute, January 26tk, 1856. 

 The following simple method is proposed for determining the 

 errors below 32° Ft. of Mercurial Thermometers, which have been 

 compared and corrected above the freezing point. 



If air has been carefully excluded from the tube, (as is the case in 

 most Mercurial Thermometers,) on turning it upside down, the 

 mercury will run from the bulb to the end of the tube, leaving a 

 small vacuum in the bulb ; by turning the thermometer quickly 



