501 



Dr. S. Young on the Determination 



(except in the case of the alcohols) with those given by 

 M. Mathias. 



The law of MM. Cailletet and Mathias may be described 

 very briefly as follows : — The means of the densities of any 

 substance in the state of liquid and of saturated vapour, when 

 mapped against the temperature, fall on a straight line. 

 I quite agree with M. Mathias that in the great majority 

 of cases the verification of the law is perfect ; but with the 

 three alcohols, especially methyl alcohol, there appears to me 

 to be very decided curvature, too pronounced to be neglected, 

 and I think it must be admitted that these substances behave 

 exceptionally in this as in many other respects. 



M. Mathias admits that at low temperatures the alcohols 

 cease to follow the law ; but he is satisfied with the agree- 

 ment between the observed and calculated mean densities for 

 a range of 200° from the critical point. The constants 

 employed by M. Mathias for ethyl alcohol can be ascertained 

 from his paper, and I give below the mean densities cal- 

 culated from these constants, together with those calculated 

 from the observed molecular volumes given in my paper 

 (lor. cit.). 



The constants for the formula D = A + ut are: — 

 A = *4149, «=— -0005570', *= temperature (Centigrade). 



Means of Densities of Liquid and Saturated Vapour. 



Temperature, 

 Centigrade. 



Observed. 



Calculated. 



AX 1000. 



224 



•3937 



•4024 



4-87 



39-45 



•3865 



•3929 



+64 



6575 



•3749 



•3782 



4-33 



89-7 



•3636 



■3649 



4-10 



10555 



•35(H) 



•3560 







1281 



•3442 



•3435 



_ 7 



144(55 



•3350 



•3342 



- 8 



161*65 



•33D9 



•3303 



- 6 



163*8 



•3240 



•323<> 



- 4 



187 9 



•3104 



•3101 



- 3 



206-35 



•299S 



•2998 







22135 



•2899 



•2915 



+ 16 



229-2 



•2850 



•2871 



421 



234-3 



•281 1 



•2843 



4-32 



238-9 



•2781 



•2817 



43H 



2431 





•2793 



! 



From the above differences it will, I think, be concluded 

 that there is considerable curvature of the " diameter " in the 

 case of ethyl alcohol, and with methyl alcohol it would be still 

 more marked. 



1 now give a revised table of the critical densities 



and 



