VOL. XXXIII. J PHILOSOPHICAL TRANSACTIONS. 51 



down to D, by way of index, and then set ui)on any table, or other place, 

 scarcely a yard higher, one may observe that the liquor is risen sensibly. Dr. D. 

 has seen it rise a quarter of an inch, when the bottle was set but a yard higher 

 than where it stood before ; so that the column of atmosphere, that pressed 

 down the tube, while the machine was on the ground, being shortened only 3 

 feet, was so overbalanced by the expansion of the air in the bottle at b, that 

 the liquor rose a 10th of an inch above d. There is, indeed, a great un- 

 certainty in this instrument ; for since it is a thermometer, as well as a baro- 

 meter, the warmth of the hand that touches it, or even comes near it, will 

 make it rise, if the air in the bottle was cold before. Mr. Gray therefore con- 

 trived to put the bottle cb, into the vessel fe, which he filled with sand ; that 

 in raising the instrument, and moving it up and down, the air in cb might con- 

 tinue in the same state, and the machine be only a barometer during the experi- 

 ment. 



This seems to bid fair for an instrument, by which the different levels of 

 places may be taken; but on a nice examination, it will be liable to error. For, 

 in the first place, though sand is not suddenly altered in its heat or cold; yet in 

 2 or 3 hours, as it is carried into a warmer or a colder place, it will become 

 hotter or colder, and the least degree of heat or cold, communicated to the air 

 CB, will alter the height of the liquor at d, when the instrument is made so 

 sensible as beforementioned. Then if, in carrying the instrument, it should be 

 accidentally inclined, as in fig. 8, so that the liquor in the bottle should not 

 cover the bottom of the tube at b, some liquor may fall out of the tube at b, or 

 some air may get into it : each of which accidents will quite spoil the experi- 

 ment. But if this machine be made portable, without any inconveniency, and 

 be secured against the action of heat and cold; or, which is the same, if the 

 alterations by heat and cold be exactly allowed for, it will be of very great use 

 and certainty, in taking the levels of distant places; provided they be not so far 

 distant from each other, that it requires above 6 hours time to carry the instru- 

 ment from one place to another ; nay very distant places, even at 2 or 3 days 

 journey from each other, may be taken tolerably well with two instruments, 

 nicely adjusted to each other, if they be noticed by two observers at the same 

 hour, in fair and calm weather. 



Now such an instrument Dr. D. thinks he has contrived, by which the differ- 

 ence of level of two places, which could not be taken in less than 4 or 5 days 

 with the best telescope levels, may be taken in as few hours. 



To the ball c, fig. g, is joined a recurve tube ba of a very fine bore, with a 

 small bubble at top at a, whose upper part is open. It is evident from the make 

 of this instrument, that if it be inclined in carrying, no prejudice will be done 



H 2 



