May 29, 1903.] 



SCIENCE. 



869 



better than this. The most superficial con- 

 sideration will show that p can not be the 

 ordinary density; but, if what we have desig- 

 nated as molecules have any physical signifi- 

 cance whatever, p must be proportional to the 

 number of molecules in unit volume. 1£ M 

 is the molecular weight of the liquid, and D 

 is its density, then D/M is the number of 

 gram molecules per unit volume, and we must 

 have 



mD 

 P = -M 



where m is the factor of proportionality and 

 depends upon the nature of the molecule. 

 Substituting this in the equation for the sur- 

 face tension we iind 



Benzene 



Toluene 



Cymene 



Methyl alcohol 



Ethyl " 



Propyl " ' 



Ainyl " 



Acetic acid 



Butyric " 



Ethyl formate 



" acetate 



" propionate.. 



" butyrate 



Methyl acetate 



Ethyl " 



Propyl " 



Ethyl ether 



Carbon bisulphide. 



Water 



Sulphur 



^6^6 



C7H8 



CH^O 



CjHsO 



CsHsO 



C'sH^O 



CoHA 



C.HsOj 

 CsHjoO, 



CjHgOj 



C5K1A 

 CiHioO 



cs. 



2.3 

 2.3 

 2.3 

 33. 



26. 



23. 



16. 

 9.7 

 3.0 

 9.1 

 6.5 

 6.0 

 5.3 

 7.8 

 6.5 

 6.3 

 4.4 

 1.81 



75. 

 2.8 



0.88 

 0.89 

 0.87 

 0.80 

 0.79 

 0.79 

 0.83 

 1.05 

 0.96 

 0.95 

 0.90 

 0.91 

 0.90 

 0.96 

 0.90 

 0.91 

 0.79 

 1.29 

 1.00 

 1.98 



Im^D^ 



D' ~ k ' 



a quantity in which the density, which varies 

 from liquid to liquid in a way which can not 

 be predicted, no longer explicitly enters. This 

 quantity iTT/D^ is probably very important 

 in molecular mechanics; if h were the same 

 for all liquids it would be most important, as 

 it would give us m, which we may call the 

 mean molecular moment of the liquid. h. 



however, probably varies from liquid to liquid. 

 So far we have made no assumption other 

 than those contained in Laplace's equations. 

 Now we shall go farther. If molecular forces 

 are electrical in their origin, as Professor 

 Sutherland and others think, then we are 

 almost justified in putting k equal to the 

 specific inductive capacity of the liquid; and 

 if on replacing h by this quantity we arrive 

 at values for the molecular moTnent, m, all 

 of the same order of magnitude, we can say 

 that the assumption is at least not an improb- 

 able one. The following table contains the 

 various quantities involved for twenty sub- 

 stances (all of those for which I have the 

 necessary data at hand), and we see that 



78 

 92 

 134 

 32 

 46 

 60 



102 



116 

 74 

 88 



102 

 74 

 76 

 18 



192 



27.5 



27.9 



27.9 



23.8 



23.1 



24.1 



23.8 



29.0 



27.2 



25.8 



25.1 



26.0 



25.0 



25.3 



25.1 



26.2 



18.2 



32.3 



74. 



44.6 



J-r- 



216 X 10» 

 299 

 660 



38 



78 

 139 

 267 



95 

 228 

 156 

 239 

 328 

 416 

 150 

 239 

 330 

 160 

 150 



24 

 430 



molecular moments 



0.71 X 10» 



0.82 



1.22 



1.14 



1.41 



1.79 



2.06 



0.96 



0.82 



1.18 



1.26 



1.41 



1.48 



1.08 



1.26 



1.45 



0.84 



0.52 



1.34 



1.13 





//■ m) 



is of the same order in every case, the greatest 

 (2.0) being four times the smallest (0.5), 

 while T varies from 18 to 74, fc from 2 to 75, 

 and M from 18 to 256. This appears to me 

 to be as satisfactory an indication of the cor- 

 rectness of this method of viewing the matter 

 as we should expect. 



In dealing with similar compounds, very 

 evident relations exist between the different 

 values of WT /B"- and also of [fci¥'r/Z)']%, 

 but the consideration of these and of other 



