568 PHILOSOPHICAL TRANSACTIONS. [aNNO 17Q5. 



Glauber's salt, be mixed with the liquor first, and then the nitrous ammoniac; 

 the powdered salts do not require stirring while cooling, like snow, for however 

 hard they are frozen, they will readily dissolve in the acid; care must be taken 

 that the partition be perfect between the salts and the liquor; and that in this, 

 and every instance where the materials are to be cooled, they be immersed below 

 the surface of the frigorific mixture. The strength of the red fuming nitrous 

 acid used in these experiments, 1 found to be 1.510, and that of the vitriolic 

 acid 1.848. 



It is very well known, that vitriolic ether will produce sufiicient cold by eva- 

 poration to freeze water; this circumstance is noticed by many, and several 

 different methods have been proposed, particularly one by Mr. Cavallo, with a 

 very ingenious apparatus for the purpose (Phil. Trans., vol. 7l): yet, as I am on 

 the same subject, and the following experiments differ, as well in the effect pro- 

 duced as in tlie particular mode of conducting them, from any I have met with, 

 I have ventured to mention them. 



June 2g, 1792, temperature of the air 71°, I sunk a thermometer (the bulb 

 being covered with fine lint tied over it, and clipped close round), by dipping it 

 in ether, and fanning it to 26°; then, by exposing the thermometer to the brisk, 

 thorough air of an open window, to 20°; and again, by using some of the same 

 ether, but which had been purified by agitating it with 8 times its weight of water, 

 applied exactly as in the last experiment, the thermometer sunk to 12°. Water 

 tried in the same manner, at the same temperature, sunk the thermometer to 56°. 

 A whirling motion was given the thermometer during each experiment. The lint 

 was renewed for each experiment, and the bulb required to be dipped into the 

 ether thrice; the first time sufficiently to soak it, after which the thermometer 

 was held at the window till it ceased to sink; then a 2d quick immersion, and 

 likewise a 3d, exposing the thermometer in like manner after each immersion. 

 In this manner a little water in a small tube may be frozen presently, by good 

 ether not purified, at any time, especially if a small wire be used to scratch or scrape 

 the sides of the tube, below the surface of the water. During the warmest 

 weather of last summer I frequently froze water in this way. 



Explanation of the figures. — Fig. 8 is a vessel in one piece, open at the bottom ; a, a, the body, 

 holding inverted 2 pints ; b, the tube, holding 3 ounces ; tlie lower or smaller part (formed by a 

 contraction, or lessening of the tube in diameter, merely for the purpose of leaving a small shoulder 

 for a temporary partition), holding rather less than \; of the whole. 



Fig. 9 is a vessel consisting of 2 parts ; a, a, the body, holding 2 pints ; b, tlie tube, holding 

 5 oz., which, together with the lid c, forms a cover to take off and on tlie vessel. This vessel may, 

 if preferred, be used instead of fig. 8, the parts corresponding with it, except in not being open at 

 bottom, and the continuation of the tube upwards just sufficient to serve for a handle. 



Fig. 10 is a vessel in one piece, open at the bottom, holding when inverted 2 pints j b, the tube, 

 holding 4^ oz. 



Fig. 1 1 a vessel open at bottom, holding inverted 1 pint. 



