•ON HYDROSTATIC INSTRUMENTS, AND HYDRAULIC ARCHITECTURE. 311 



proper place, unless the instrument be very accurately horizontal, or in tlie 

 same position in which the mark was adjusted. The surface of the bubble, 

 especially when it is small, cannot, in a strict sense, be called perfectly hori- 

 zontal, since its form approaches nearly to that of a sphere; but in order that 

 the centre of gravity of the water may attain the lowest possible situation, the 

 bubble must necessarily occupy the highest point of the tube. (Plate XXI. 

 Fig. 279.) , ' 



The principles of hydrostatics have been employed in various ways for sup- 

 plying lamps with oil. It is found that a lamp will burn, without consuming 

 any considerable portion of its wick, as long as it is amply supplied with oil ; 

 hence it becomes desirable that it should always Ijife level with the surface of the re- 

 servoir, and this may be effected sufficiently well by placing the wick at the ■ 

 edge of a very large vessel,or atthe end of a tube projecting from such avessel,or 

 from a vessel closed above, and opening only by an orifice below, which lets in 

 the air as the oil escapes through it. But all these methods are often attended 

 with inconveniences of various kinds, especially where the lamp is to be em- 

 ployed like a candle, and placed on a table. A French artist has applied a 

 little pump, which is worked by means of a spring, for raising the oil from * 

 vessel under the lamp; but this refinement is too complicated to be practi- 

 cally useful. Mr. Keir's lamp contains a divided cavity, one part of which is 

 filled with oil, and the other with a saline or saccharine fluid of greater density, 

 so that when the oil contained in the upper part of the tube is exhausted, its 

 place is partly supplied by a fresh portion, which is forced up in consequence 

 of the descent of the denser fluid in a much larger vessel. Still, however, the 

 surface must be lowered by degrees; but by combining the invention with 

 Dr. Hooke's semicylindrical counterpoise, a little modified, the height of this 

 fluid may be so regulated, that the surface of the oil may remain almost in- 

 variable, until the reservoir is quite exhausted. For this purpose, the centre 

 of gravity of the counterpoise must be a little higher than the line which bi- 

 sects it; and its specific gravity must be about three fourths as great as that of 

 the fluid ; and in this manner it may be made to raise the surface of the hea- 

 vier fluid, in proportion as a greater quantity of it escapes, to supply the place of 

 the oil; and to keep it always at a sufficient height above the surface which 

 separates it from the oil, so that the wick may be amply and almost uniformly 

 supplied. (Plate XXI. Fig.-280.) 



