34 Recent Chemical Researches . [January, 
than heretofore, apply this generalisation in the determina- 
tion of disputed atomic weights. A modification must, 
however, be made in the statement of the law, thus : — ■ 
“ The specific heats of the solid elements vary with the tem- 
perature : for every element, however, there is a point (T 0 ) 
from which the variation in the specific heat, with increasing 
temperature, is entirely insignificant. The product of the 
atomic weight into the value of the specific heat (estimated 
at temperatures so that T>T 0 ) is, for all the elements, a 
nearly constant number varying from 5*5 to 6*5.” 
The bearing of Prof. Weber’s researches upon the specific 
heat of allotropic modifications of the same element, and 
also on the question of varying chemical value of the same 
element in its compounds, will be discussed hereafter. 
Changes have been proposed by Mendelejeff in the num- 
bers which express the atomic weights of yttrium and 
erbium, of cerium, lanthanum, and didymium : these pro- 
posed changes are based upon considerations of the relations 
which appear to exist between the atomic weights and the 
properties of the elements. These relations exhibit the 
form of a periodic function. If the elements be arranged 
in order of their atomic weights, we may divide them into 
groups the properties of the members of each of which 
exhibit a regular variation as the atomic weights increase. 
Now, if the atomic weights generally accepted for yttrium, 
erbium, cerium, lanthanum, and didymium be the true 
atomic weights of these elements, we find that these bodies 
do not occupy the positions in the groups which, judging 
from their properties, we should expedt they would do. But 
as we are certain, so far as investigation has gone, of 
many of the properties of those elements, while we are 
not at all certain of their atomic weights, Mendelejeff pro- 
poses to alter the latter, so that the elements may fit into 
those places which, according to his theory, they ought to 
occupy. 
In order that the atomic weights of these elements should 
be more definitely settled it is necessary that the specific 
heats be determined. Mendelejeff states that he has deter- 
mined the specific heat of cerium to be 0*05 : this number 
multiplied into 138 (the number proposed for the atomic 
weight of this body) gives 6*9, which is in keeping with the 
average atomic heat of the elements. Nilsson is at present 
engaged in the preparation of considerable quantities of 
these metals, with the view of determining their specific 
heats. 
