70 M. MELLONI ON THE IMMEDIATE TRANSMISSION 
submitted to the rays emerging from any one of our five screens, we 
shall always find the same ratios between the different terms of each 
series. 
As to the black and the green glasses, their changes of transmission 
occur sometimes similar, sometimes contrary to those of the other plates. 
We should not however be surprised at these irregularities, as the green 
and black colours alter the natural diathermancy of the glass and give it 
an aptitude to transmit quantities of heat which will be more or less con- 
siderable in proportion as the rays issuing from the different screens 
possess themselves a diathermancy more or less analogous to that intro- 
duced into the vitreous substance by these two colouring materials*. 
* Ina note to the preceding Memoir (page 8)I have said that, for the study of ca- 
lorific radiations the thermomultiplier is preferable to every former thermoscopic 
apparatus. The great number of experiments that I have since performed by 
means of that instrument have produced in my mind a thorough conviction of 
the truth of that opinion. As there are still many experimental researches to be 
made not only in that class of phenomena, of the history of which we have 
scarcely given an outline, but in every branch of the study of radiant heat, it is to 
be wished, for the interests of science, that every investigator would furnish him- 
self with a thermomultiplier. ‘This apparatus, in the state of perfection neces- 
sary to ensure good observations, is unfortunately one of those which a person 
cannot construct for himself until he has made several attempts which are at- 
tended with a great loss of time, and which cannot succeed in many places for 
want of the requisite means. For these reasons I have thought it advisable to 
put some one in Paris in the way of supplying them to the public. There are 
excellent ones to be had at M. F. Gourjon’s, rwe des Nonandiéres, N°. 2. The 
description of the ingenious means employed by this able mechanic to give 
to the instrument every improvement which I was desirous of having in- 
troduced into it would occupy too much time. I shall therefore confine my- 
self to the mention of the principal defects found in the first instruments of this 
kind presented to the Academy of Sciences by M. Nobili and myself (at the 
sitting of the 5th of September 1831), but now laid aside for the improved 
thermomultipliers constructed by M. Gourjon. 
In the first place the volume of the thermoelectric pile was too bulky, (being 
from 36 to 40 centimetres square in section,) a circumstance which rendered it 
impossible to operate on small pencils of calorific rays: in the next place the 
galvanometer did not mark fractions lower than half a degree, and the magnetic 
needles, instead of standing at the zero of the scale, settled sometimes to the 
right and sometimes to the left at a particular distance for each galvanometer, 
amounting in some instances to10 degrees. In fine, the mountings being almost 
all of wood the pieces became warped by the hygrometrical variations in the 
atmosphere, and the instrument was rendered unserviceable. 
The thermomultipliers of M. Gourjon have thermoelectric piles the actin 
surfaces of which are not larger than the section of a common thermometer 3 
centimetres square). As to the galvanometers they are mounted entirely in 
copper with the exception of the small pieces necessary for the purpose of iso- 
lation : the minuteness of their indications extends to a fourth and even a sixth 
part of a degree, and the needles, when at rest, stand exactly at the zero of the 
seale. Itis almost needless to add that with these improvements the instrument 
has lost nothing in sensihility. 
