274 j^ttraBton,l^c. between floating Bodies. 



concerning its figure or dimenfions. It is only known that it rifes according to the nature 

 and temperature of the fluid, and the material of which the flip itfelf is compofed. 



The two mafles of water raifed by the plate of glafs, tend to give it motion down- 

 wards by adding to its weight, and fideways by the ai^lion of the water, which tends to 

 move in the horizontal direction. M. Monge compares this to the a61:ion of a chain, haring 

 the figure of the curve, but varioufly loaded in its different parts. In the prefent figure the 

 horizontal adtions are precifely alike, and confequently equal; but it is eafily to conceive 

 that they may become unequal, and alfo to fhew the fame by experiment. FBA, fig. 9, repre- 

 fents a drop of water lying on a fmooth plane: to this the plate CA, fuppofed to be moveable 

 on the centre D, is applied. The effeiSl; of the action, which is now either wholly or chiefly 

 on the fide EF, tends to carry the extremity A towards B, and would caufe the plate to 

 place itfelf in the pofition reprefented in fig. 10, if gravity or other circumftances did not 

 interfere. 



But the attradtion is more particularly feen by fufpending two flips or plates of glafs, as in 

 fig II. Thefe will retain the vertical pofition as long as they arc more diftant. from each 

 other than the limits of this capillary attradlion; but when, as in fig. 12, the curves inter- 

 fere, other confequences muft follow. Suppofe the water to acquire the figure AHB at its 

 furface, and the quantity which the two plates fufpended before will be greater by the por- 

 tion CHD than it is at prefent. The power of fufpenfion is not, therefore, faturated,andthe 

 water will rife, for example, to IGK, and this elevated water, independent of any increafe of 

 attraction from the diminifhed diftance, is confidered by the author as a weight fufpended to 

 the chain which connedls the two plates, and producing a greater reaction on the plates 

 themfelves. The plates will not, therefore, continue in equilibrio, but will come together and 

 ftrongly adhere. 



This eiFedl of fluids upon the fides of capillary vacuities, whetlier it be well accounted 

 for or not by the comparative argument of the weight, is of confiderable moment in the phe- 

 nomena of nature. M. Monge obferves, that as thefe plates come together and are retained, 

 not by any diredl attractive force, but indire:ily by the operation of the fluid ; fo likewife by 

 extending the argument to fubmerged bodies we may account for the ftrong adhefion of the 

 parts of the cryftals of ialts in their feparation from water, fuch as the fiilphate of lime, which 

 lofes its water of cryfialJization and its tenacity by heat,_fulphate of foda, and many other hard 

 cryftals, which fall into powder by drying, &c. 



The laws of attraction, when two plates are plunged in mercury, as in fig. 13, is alfo con- 

 fidered by our author. This prefigure will be equal to the weight of a column of the fluid, 

 whufe bafc is the furface DH of the glafs, and height is half the depth between D and H. 

 If the plates be placed horizontally beneath the fluid, as i^i fig. 14, the preffure will be twice 

 as great, and conftant, provided the depth AB be at leaft equal to the limit of repulfion IH". 

 It may alfo be remarked, that if a fmall hole M, fig. 15, be made at lefs depth, the mercury 

 will not flow out, though it wouW iffue from the hole M, fig. 14, below that depth. Hence 



it 



