standard chloride paper is also true for bromide papers; viz., that the same tint is gained by exposing the papers for a time 1 to an intensity 10, as is gained on exposure for a time 10 to an intensity 1, or for a time 2 to an intensity 5; and so on.

Hence, if a pendulum strip be made with bromide paper, and the same paper be exposed to varying intensities of light, and the tints thus produced be read off on the bromide pendulum strip, it should be found that the numbers from the strip calibration-table are to one another in the same proportion as the numbers indicating the intensities of light. On trying the experiment, however, it was found impossible to read off on the bromide strip, owing to the slight difference in tint between the two ends.

On exposure of brom-iodide paper in the same ways, the following numbers were obtained as the mean of six experiments by the aid of the perforated plate (direct sunlight), and of four by means of the cylinders (diffused light).

Brom-iodide paper, with direct sunlight :

As with the bromide paper, these numbers are found, on graphical representation, to yield a mean curve identical with that derived from the previous experiments with the disc-photometer, &c., (that is Bunsen and Roscoe's law is likewise true for this paper).

On endeavouring to read off brom-iodide papers exposed to varying intensities of light on a brom-iodide pendulum strip, it was found impossible to get any trustworthy results, owing to the slight difference in tint between the two ends of the strip.

The following numbers were derived from the mean of four experiments with chloro-bromide paper with diffused light from the zenith (cylinders) :

These numbers, also, on graphical representation, give a curve identical with that derived from the disc-photometer experiments.

On reading off papers exposed to varying intensities of light (in the cylinders) on a chloro-bromide pendulum strip, the following numbers were obtained as the mean of four experiments. (The numbers in the column P are the relative times in which chlorobromide paper would, on exposure to light of uniform intensity, gain the tints assumed by the papers experimented on: the columns A, R and Diff. have the same meanings as hitherto.)

Considering the great difficulty of reading off correctly on a chloro-bromide pendulum strip, which exhibits much less gradation of tint than a chloride one, the differences are sufficiently small to show that Bunsen and Roscoe's law is obeyed by this paper.

On exposing chloriodide papers to diffused light of varying intensities (cylinders), the following numbers were obtained as the mean of four experiments:

-

These numbers also, on graphical representation, are found to yield a mean curve identical with that derived from the disc-photometer experiments.

Chlor-iodide papers exposed to varying intensities of light (cylinders), and read off on a chlor-iodide pendulum strip, yielded the following numbers as the mean of three experiments. (The column Q indicates the relative times in which chlor-iodide papers would assume the tints gained by those experimented on, on exposure to a constant source of light.)

As the chloriodide pendulum strip shows much less gradation in tint than the chloride strip, these numbers show that Bunsen and Roscoe's law is obeyed by this paper.

By exposing chloro-brom-iodide paper to diffused light of varying intensities, the following numbers were obtained as the mean of four experiments :

On graphical representation, these numbers are found to give a mean curve identical with that derived from the disc-photometer experiments. On endeavouring to read off chloro-brom-iodide papers exposed to varying intensities of light, on a chloro-brom-iodide pendulum strip, no good results were obtainable, from the slight differences in tint between the two ends of the strip.

Since all the above papers obey Bunsen and Roscoe's law, the table given on p. 45 also denotes the relative intensities of light to which pieces of any papers prepared with the above halogens must be exposed for the same time, in order to gain the same series of tints.

The results obtained by the foregoing experiments, together with those previously obtained by Mr. McDougall, may be briefly summed up thus:

(1.) Papers prepared with solutions of different alkaline chlorides darken to the same extents in the same relative times, no matter with what alkaline metal the chlorine be combined, and no matter what be the constant percentage of chlorine in the salting solution. (2.) The same holds true for papers prepared with bromide solutions (and hence, probably, for all the others).

(3.) Photo-chemical induction does not exist to any appreciable extent with chloride, bromide, and brom-iodide papers (and hence, probably, with none of the others).

(4.) The relative times for which pieces of any one of the above papers must be exposed to light of a constant intensity in order to gain a definite series of tints vary with each paper.

(5.) The relative intensities of light to which pieces of any one of the above papers must be exposed for a constant time, in order to gain a definite series of tints, vary with each paper.

(6.) The relative times for which pieces of any one of the above papers must be exposed to light of a constant intensity in order to gain a definite series of tints, are in the same ratio to one another as are the relative intensities of light to which the same pieces must be exposed for a constant time in order to gain the same series of tints.

(7.) The relative times for which pieces of any one of the above papers must be exposed to a constant source of light in order to gain a given series of tints, or the relative intensities of light to which they must be exposed for a constant time to gain the same series of tints, are given by the table on p. 45.

(8.) By the aid of this table, the relative sensitivenesses of any of the above papers with respect to any particular tint can be calculated, when their relative sensitivenesses with respect to some one given tint are known.

In conclusion, I beg to tender sincere thanks to Professor Roscoe for the valuable advice and assistance which he has given me in carrying out the above investigation.

VII.-Prognosis of new Alcohols and Aldehydes.

By Professor H. KOLBE, in Leipzig.

In my several publications on Organic Chemistry, I have repeatedly directed attention to the probable existence of a class of compounds, related to the ordinary alcohols and aldehydes in the same manner as diamines and triamines are related to monamines.

In a former memoir on the Secondary Alcohols,* I placed ammonia and carbin,† ethylamine and ethylcarbin, in parallels, thus:

and endeavoured to demonstrate, from the known existence of secondary and tertiary ammonias, the possibility of the existence of secondary and tertiary alcohols, several of which are now actually known.

A similar parallel, drawn in a different direction, opens out a new field of chemical research. For just as diamines are produced by the substitution of diatomic radicals for hydrogen in ammonia, so may it be regarded as probable, or, I should rather say, predicted with certainty, that dicarbinols will be produced under similar circumstances from carbinols, as indicated by the following constitutional formulæ :

For reasons developed in the memoir above referred to, I have proposed to designate methyl as “carbin,” methylic alcohol as "carbinol," ethylic alcohol as "methyl-carbinol," which names are used in the present communication.