Conjltudion of the Spirit Level. 399 



The level which was thus grouyd is one foot in length ; and the cylindei; on which it 

 was firft wojiced is ot the fame length. When it was finifhed it was found to be too 

 fenfible. It was, therelore, worked on another cylinder of between nine and ten inches 

 long, which diminifhed its fenfibility fo far, that the bubble, which is nine inches and 

 four Liaes' long, at the temperature of 16'' of Reaumur, above freezing is carried from the 

 middle of the tube exaftly one lint-, for every fecond of a degree of inclination. This 

 degree of fenfibility was thought' fufficicnt ; but any greater degree which may be re- 

 quired may be obtained by the proccfs here defcribcd. 



It may be remarked that a glafs tube is very fubjeft to be fplit by grinding its inneV 

 furface ; the fame tube will not be endangered by grinding its external furface even with 

 coarfe emery ; and when once the polifh of the infide is ground off, the danger is over, 

 and coarfer emery may be ufed without iear. Thick glals is more fubjeft to this mif- 

 fortune than thinner. The coarfeft emery made ufe of in grinding the tube here fpoken 

 of was fufficiently fine to employ one minute in defcending through the height of three 

 inches in water. 



Annotations by the Author of the preceding Memoir, 



(1.) If the ether be not well reftified it is fubjeft to two great inconveniences in 

 this ufe. If the tube be very flightly agitated, the ether divides itfeif into feveral bub- 

 bles, which employ a confiderable time before they unite. In the fecond place, as this 

 ether is decompofed in the courfe of time, it depofits very fmall drops of oil which ad- 

 here to the tube. Hop the motion of the "bubble, and render the level very faulty. The 

 ether is befides more fluid when reftified and freed from a faponaceous matter which 

 caufes its bad effefts *. 



(2.) A level is faid to be more fenfible the fmaller the inclination required to produce 

 It given change in the pofition of the bubble f . 



Let E M F, fig. 1. pi. xvii. reprefent the internal irregularity of a tube. It may 

 eafily be underftood that when the elevation of temperature dilates the liquor in the tube, 



* Our artifts ufe ether, aiid>find that the precautions here mentioned are very neceffary. N. 



f And the more quickly and rapidly it returns to its firft fituation when the micrometer fcrew is fuddenly re- 

 turned to its zero. Thefe three qualities do not conftantly accompany each other in common levels. — It is 

 not immediately obvious why a long bubble fhould be more fenfible, as the run is exaiftly the fame whatever may 

 be the Cze of the vacant fpace. To explain this let A H B I, fig. 5; plate xviii. reprefent a hollow ring, 

 iUed with fluid to A and B on a level with the center C ; then if the vertical point H be inclined to either fide, 

 the furfaces A and B will have an apparent run, along the tube, of exaftly the fame quantity. And the fame 

 would follow if the tube were filled to E and D, or to G -and F. But the aflual rife or perpendicular elevation 

 of the fluid required to pafs over a given arc will be as the fine of half the arc left empty, or half the length of 

 the bubble, and confequently the momentum by which thecohefive attra£tion at the edges of the bubble is to be 

 overcome will be in ihe fame proportion. The longer bubble will therefore move morebrilkly and fettle more 

 accurately to its ftation.— N. 



3 F a and 



