MKMOIU XXVI.] ( 'A HLLARY ATTRACTION, ETC. 345 



among optical writers under the name of Newton's col- 

 ored rings. At the point c the lens and the plane are, 

 as Newton has shown, a distance apart of about one half 

 of the millionth of an inch ; and from this centre, pro- 

 ceeding outwardly, the distance between the glasses of 

 course increases. If anywhere at the outer portion a 

 drop of water be introduced, it extends itself instantly 

 across all the colored rings, reaching even across the cen- 

 tral black spot. 



If a tube of such diameter that it could lift water ten 

 inches be broken off so as to be only six inches long, we 

 might inquire whether the water would overflow from 

 its top or simply remain suspended there. 



Mathematical considerations, as well as direct experi- 

 ments, prove that in such a case there would be no over- 

 flow. A capillary tube under these circumstances lifts 

 the water, but does not produce a continuous current. 



But if a removal of the liquid at the top of the tube 

 take place in any manner, as by evaporation or by being 

 dissolved in another liquid, a continuous current is pro- 

 duced. 



As illustrating the production of such a continuous 

 flow we may cite the case of a spirit-lamp, the wick of 

 which may be regarded as a fagot of capillary tubes. Be- 

 tween the fibres of the wick there are interspaces that 

 answer as tubes. If the cover of the lamp be taken off, 

 all the spirit will eventually pass up the wick and escape 

 from the reservoir by evaporation. Or, in an oil-lamp, 

 the wick of which becomes readily saturated with the 

 oil, but never exhibits an overflow, on the lamp being 

 lighted, the oil is burned off, a current is established, and 

 after a time the reservoir is emptied. 



The phenomena of capillary tubes are connected with 

 the adhesion of surfaces. Clairaut showed that if the 



