M. E. H. Amagat on the Isothermals of Fluids. 651 



find in „ -= additional terms of the form — -== or ™^-. 

 Jb ctjii a. oil* -Hi 



These terms may be very large when E is small compared 

 with f2, so that /'(E) may be appreciable for values of E for 

 which F is vanishingly small. In other words, the variation 

 of 7 with temperature ought, in general, to be perceptible at, 

 temperatures below the point of incandescence. 



k 



LXI. JSote on the Isothermals of Fluids, and in particular of - 

 Hydrogen. By M. E. H. Amagat *. 



IN an interesting paper which Mr. Rose-Innes has recently 

 published (Phil. Mag. July 1901) relative to the scale 

 of absolute temperatures this author remarks that the iso- 

 thermals of hydrogen, which within the limits of temperature 

 and pressure of my experiments appeared sensibly straight 

 and parallel, should in reality only follow this law ap- 

 proximately. 



To avoid any misconception I desire to make a few remarks 

 on this subject. The results which Mr. Rose-Innes cites are 

 those of my memoir of 1881 ; the isothermals in this research 

 reached only 400 atmospheres; in these limits and at the 

 temperatures at which I worked the curvature of these lines 

 could not be detected with certainty, but in the much more 

 extended research which I published in 1893, and in which 

 the superior limit of pressure was raised to 3000 atmospheres, 

 T have returned to this point and have insisted at length on 

 this fact, viz. : the isothermals for all bodies, including 

 hydrogen, show a perfectly distinct curvature, but one which 

 can only be determined with certainty by prolonging these 

 lines much farther than was done in my previous research. 

 The following are the words of the passage referring to this 

 point (Ann<des de Chimie et de Physique)'. — 



" La forme des isothermes au dela de l'ordonnee minima est 

 un des points que je m'etais surtout propose d'etudier ; ces 

 courbes a partir d'une distance d'autant plus petite de 

 1'ordonnee en question que la temperature est plus basse, 

 m'avaient paru se transformer en lignes sensiblement droites; 

 j'avais remarque les indices d'une legere concavite, mais a 

 peine indiquee et j'avais pense que le reseau suffisemment pro- 

 longe se transformerait pen a p< j u en un faisceau de droites 

 paralleles. Cette hypothese etait d'autant pius seduisante. 

 que le coefficient angulaire de ce faisceau donnait de suite le 



* Communicated bv the Author. 



