against the principle as the zero is never followed by appretiable transmissions.
The same principle holds in respect to all the liquids that I have been able to submit to experiment. It will be recollected that, in my mode of operating, the rays of heat, before they reach the liquid layer, must pass through a plate of glass. Now this substance becomes more and more interceptive in proportion as the sources employed are of a less elevated temperature, and consequently acts upon the calorific rays with an effect the same as that which a screen of variable transparency would produce in respect to light. The process therefore which I pursued in my first Memoir could not enable me to determine the exact ratios of the calorific transmissions through the same liquids when the source is changed; but it was possible to make it available for the purpose of establishing, in the greatest number of cases, the general law of decrease which we have just determined in respect to solid bodies.
Let us suppose that a thick plate of glass being submitted to the successive action of an equal quantity of heat, emanating from our four sources, gives these transmissions:
| 30, 18, 2, 0. |
Let us suppose a parallelopiped, with sides parallel to the faces of the plate, to be cut out of the glass, and the cavity thus made to be filled with a given liquid: let us then suppose that the transmissions of the system become all respectively inferior to the preceding, and are reduced, for instance, to
| 20, 8, 1, 0, |
it will be immediately concluded that the liquid acts on the calorific rays from different sources in the same manner in which its glass case does; that is, that it exhibits an order of decrease similar to that exhibited by the glass and by solid bodies in general. Now this is precisely the result furnished by the liquids contained in my glass vessels[1].
- ↑ In many instances I was unable to obtain any transmission, even by employing a very powerful radiation. It is thus that water, which transmits six or seven hundredths of the rays from a Locatelli lamp, completely intercepts the heat of the last three sources. Calculating the limit of error for the case least favourable to interception I found it 1616: the source was then brought very close to the liquid and an equal layer of oil employed, which caused in the index of the galvanometer a deviation of several degrees. Now if the water allows a passage to the radiation from bodies heated even to incandescence or brought to lower temperatures, the part transmitted must be less than 1616 of the incident quantity. I here speak of a layer of 3mm or 4mm in thickness: for it is possible and even very probable that layers much thinner than these may be in some slight degree permeable to rays of this kind. Thus we have seen glass of 0mm·07 in thickness transmit 12100 of the rays emanating from boiling water, while a plate of 1mm intercepted them totally. But as, in order to compare different transparencies, we must operate on a certain thickness of each medium (for the
