84 Professor Barrett, 



vapour, and partly on the proximity of the two surfaces. A 

 proximity not to be estimated absolutely, but with reference to 

 the length to which a molecule of the gas will travel in the 

 intervals between its encounters with other molecules. Hence 

 there are obviously three modes whereby the excess of pressure, 

 this Crookes' force, may be developed or augmented : — 



1st. By lengthening the paths of the molecules between the warm 

 and cool surfaces, accomplished by attenuating the gas. 



2nd. By bringing the hot and cold surfaces very near together. 



3rd. By increasing the difference of temperature between the 

 two surfaces. 



Now if the support of the spheroidal drop be due to this 

 Crookes' force a difference of temperature must exist between the 

 drop and the surface over which it stands, and the greater this 

 difference of temperature the larger the drop that ought to be 

 supported, and the more persistent the phenomenon. Mr. Moss 

 has shown (Proc. R D. S., Dec. 1877) that by securing a continual 

 difference of temperature a globule of ether may be supported on 

 the surface of its own liquid for upwards of an hour, until in 

 fact some accidental derangement occurs. The conditions of the 

 two theories being thus defined, it is easy to see that several 

 crucial experiments might be devised which should help to 

 decide the question at issue. 



The following experiment the author has made with this object 

 in view. Upon the surface of the ordinary petroleum of commerce, 

 liquid globules of transient duration can readily be formed, simply 

 by removing a small quantity of the liquid in a pipette and care- 

 fully depositing a drop on the surface of the liquid. These drops 

 are clearly in the spheroidal condition, and they are easily and 

 abundantly formed by dipping a vibrating tuning fork into the 

 liquid, or by drawing a fiddle bow over the edge of the vessel 

 containing the liquid. According to the ordinary explanation 

 the drops are supported by the elastic force of the vapour of the 

 liquid, which would, of course, be greater the higher the tempera- 

 ture of both liquid and drops. According to Mr. Stoney's theory 

 the drops are supported by the Crookes' force, generated by the 

 proximity of the drop and liquid, and by the fact that they are 

 at different temperatures. Evaporation rapidly cools the drops 

 jerked up from the liquid, and thus a slight difference of tempera- 



