3 § 
Recent Chemical Researches . 
[January, 
the total energy of nitrogen is comparatively small, it seems 
difficult to imagine that the hydrochloric acid molecule will 
be split up into its constituent atoms, and that these will 
then each oscillate about the nitrogen atom. But is it not 
possible to avoid this difficulty by supposing that there is 
such a thing as molecular as well as atomic valency ? — that 
although in the hydrochloric acid molecule the lines of force 
of its atom are mutually satisfied, there may be directions 
in which the molecule, as such, may exert a certain amount 
of chemical energy ? On this view we may picture to our- 
selves the hydrochloric acid molecule oscillating about these 
two points, at which the nitrogen atom is still free to exer- 
cise chemical force. But the force exercised at these points 
is, by supposition, less than that exercised at the other three 
points ; therefore when the compound is heated the hydro- 
chloric acid molecule is easily driven beyond the sphere of 
aCtion of the nitrogen atom, and we have the phenomenon 
of dissociation. 
Whatever theory may be adopted of actions such as this, 
it is evident that, in the determination of the valency of an 
element, attention should be paid to the whole of its com- 
pounds, but that reliable conclusions are only to be drawn 
from those which can exist in the gaseous form. 
Researches bearing upon the valency of the individual 
elements have been carried out of late by various observers. 
Many disputes have arisen as to the valency of the elements 
which constitute the nitrogen group, so that an especial 
amount of interest attaches itself to any new researches 
into the constitution of the compounds of the elements. 
Michaelis concludes, from general observation of a large 
number of its compounds, that phosphorus is a pentavalent 
element. Wurtz has determined anew the vapour density 
of phosphorus pentachloride, by diffusing the vapour of this 
compound into a space filled with the vapour of phosphorus 
trichloride. By this means the dissociation of the penta- 
chloride was prevented. The number 7*226 represents the 
average vapour density calculated from the results of twelve 
experiments ; the theoretical vapour density for PC 1 5 (two 
volumes) is 7*217. We have in this determination an almost 
certain proof of the pentavalency of phosphorus. Another 
proof that phosphorus is really a pentavalent element has 
been supplied by Thorpe, who has succeeded in preparing a 
gaseous pentafluoride of phosphorus, PF 5 , by the adtion of 
arsenic trifluoride upon phosphorus pentachloride : the new 
compound adts upon glass, and is soluble in water, but may 
be colledted over mercury. It seems, then, very probable 
