effect itself must be admitted to be less energetic when the temperature is higher. With certain liquids, such as alcohol especially, the direct influence of the temperature is almost insensible, which is probably owing to the great dilatability of that liquid rendering preponderant the influence of the decrease of density.

§ 4. Determination of the Magneto-rotatory Power of the Mixture of Two Liquids.

I began by mixing equal volumes of distilled water and rectified alcohol of sp. gr. 0.804. The magneto-rotatory power of the mixture was obtained by means of two series of experiments, made at two different periods, and which gave the following numbers: :

The magneto-rotatory power calculated as the mean of those of water and alcohol is 0.938, that of water being 1.000, and that of alcohol 0.876. The actual density of the mixture was found to be 0.935; the density calculated as the mean of those of alcohol and water would be 0.902. Now the ratio of the real rotatory power to the calculated is 1.0362; and the ratio between the actual and the calculated densities is 1.0365. Thus, in a mixture of equal volumes of water and alcohol,, which is accompanied by a sensible contraction (proved by the increase of density), the magneto-rotatory power augments in exactly the same proportion as the density-which proves that the molecular magneto-rotatory power does not change. We have seen, in the preceding Section, that it is the same with alcohol with respect to the changes of density proceeding from variations of temperature. In taking a mixture of equal volumes of water and alcohol, I chose that in which the change in volume is the greatest; it is evident, then, that with mixtures in other proportions the same result would be found.

The mixtures of water and sulphuric acid in different proportions gave very interesting results, which I think deserving of detailed exposition. This kind of experiments present some difficulties, especially with concentrated solutions, on account of the rapidity with which they attract humidity from the air, which changes their identity; nevertheless this inconvenience may be avoided by taking precautions. In fact we have seen that monohydrated sulphuric acid having a rotatory power of 0.750 acquires one of 0.757 after one or two removals from one vessel to another in the air, and one equal to0·768 after a greater number. My first experiments were made with the Paris sulphuric acid, of which the rotatory power is 0-800, and the density (taken

with great accuracy) 1.842. With this acid and distilled water I made mixtures in the following proportions :

9 volumes of water, and 1 of acid:

I ascertained the magneto-rotatory power of each of these solutions; and I compared this real power with that calculated by taking 0-800 for the magneto-rotatory power of the pure acid and 1.000 for that of water, taking account of the proportions of water and acid in the inixture, and supposing that no change of volume takes place by the act of mixing. I likewise determined the ratio between the actual density of each solution and its density calculated by taking 1·8421 for the density of the acid and 1.000 for that of water, and supposing that no change of volume is effected by mixing.

pure

The following Table, drawn up from the means of a great number of experiments, gives, for each of the solutions, on the one hand the actual and the calculated rotatory powers and the ratio between them, and on the other the actual and the calculated densities and the ratio between these.

*The result obtained, both as to the ratio of the actual to the calculated rotation and the ratio of the actual to the calculated density, with the solu tion containing 7 parts by volume of acid to 3 of water presents an evident anomaly, of which I have not been able to discover the cause; it is possibly the result of an experimental error, though I am not disposed to think so.

This Table shows that, in solutions containing only a relatively small proportion of acid or of water, the magneto-rotatory power increases in the same ratio as the density. Thus the ratio of the actual to the calculated power is 1.032 for the solution which contains, in 10 parts by volume, 1 of sulphuric acid and 9 of water; for the same solution the ratio of the actual to the calculated density is also 1032. Similarly the ratio between the actual and the calculated rotatory powers is 1.040 for the solution which contains, in 10 parts by volume, 8 of acid and 2 of water; and the ratio of the actual to the calculated density of the same solution is 1·045, or very nearly the same, only a little higher. But if solutions be taken which contain nearly equal quantities of water and acid (such as those comprised between 3 tenth parts of water and 7 of acid, on the one hand, and 3 tenth parts of acid and 7 of water, on the other), then the ratios between the actual and the calculated rotatory powers become much lower than those between the actual and the calculated densities; and this difference attains its maximum in the solution which contains exactly equal volumes of acid and water.

As to the rotatory powers themselves, the ratio between the actual and the calculated power does not differ much from one solution to another; yet it increases slightly, but regularly, with the concentration of the solution.

It seems to me that we may infer from these observations that the combination of the acid and water diminishes the molecular rotatory power, since the ratio between the actual rotatory power of a certain volume of the mixture and the rotatory power of the same mixture calculated on the supposition of there being no contraction is lower than the ratio between the actual and the calculated density, while the two ratios would have been equal if the combination of the water and the acid had not modified the rotatory power of each. When the proportion of water or of acid is very slight, the ingredients combining in very small quantity and mixing in that one of them which is in excess, the solution behaves like a simple mixture. A very curious fact is the almost complete equality of the ratios between the actual and the calculated magneto-rotatory powers of all the solutions, which would seem to indicate that the chemical action which brings to the magneto-rotatory power of the solution a modification which renders it different from what it would be if the liquid were only a simple mixture, acts sensibly in the same degree upon each; I say sensibly, because the ratios, though not differing much, show a tendency to increase with the increase of the proportion of acid relatively to that of water.

I submitted to the same experiments two other solutions of sulphuric acid in water, which were kindly supplied to me by

Professor Marignac. Both were prepared with monohydrated sulphuric acid (HO SO3) of sp. gr. 1.83211 at 20°, and with the specific magneto-rotatory power (as I have stated above) of 0.750. One of these solutions consisted of HO SO3 +5 aq; the other of HO SO+10 aq. The actual magneto-rotatory power of each of these solutions was compared with its magneto-rotatory power calculated as the mean of the rotatory powers of the water and of the acid (HO SO3) mixed in the same proportion in which they enter into the solution. Similarly the actual densities were compared with the densities calculated on the supposition that no contraction occurred in the mixture.

We see that, as in the preceding experiments, the ratio between the rotatory powers differs less from the ratio between the densities in just the same degree as the proportions of water and acid in the mixture differ more. What is remarkable is, that, as before, the ratio between the actual and the calculated rotatory powers is sensibly the same for each solution.

It must not be lost sight of that the specific rotatory powers are always taken with equal volumes, and that the calculated powers are the means of the respective powers of water and the acid, assuming the mixture to take place without change of volume. If in the calculation account is taken of the contraction, then we find, as might have been expected, that the calculated power is higher (always with an equal volume) than the actual power; in fact it is 0.965 instead of 0.943 for HO SO3 +5 aq, and 0.998 instead of 0·980 for HO SO3 + 10 aq. Thus the magneto-rotatory power of the compound is less than the mean of the respective powers of the water and the acid which form the combination. This was already evident from the fact that, without taking account of the contraction, the ratio between the actual and the calculated powers is less than the ratio between the actual and the calculated densities.

§ 5. Determination of the Magneto-rotatory Powers of some Isomeric Liquids.

M. Berthelot had kindly given me, in the spring of 1869, a certain quantity of two isomeric liquids; but not having prepared them himself, he took care to tell me that he did not guarantee their perfect purity. They were ethylvaleric ether and amylacetic ether, both having the general formula C14 H14 04. Submitted several times, and at different periods, to experiment, the second always exhibited a stronger magneto-rotatory power than the first. The most accurate experiments gave 0.877 for the specific magneto-rotatory power of the first, and 0.895 for that of the second. When I wished to compare their rotatory powers with their densities, I was much embarrassed on account of the difficulty of ascertaining these densities. Thus, according to M. Delffs, the density of amylacetic ether is 0.863 at 10°, and according to M. Kopp 0.8837 at 0°. The density of ethylvaleric ether is, according to M. Delffs, 0.870 at 13°-5, and according to M. Berthelot himself 0·869 at 14°. On the other hand, M. Adolphe Perret (who was so obliging as to determine for me the densities of most of the liquids I used in my researches) found for the density of both the ethers in question at 16° the number 0.870; so that the samples upon which I operated had, apparently, the same density; the very perceptible difference, therefore, presented by their rotatory powers cannot depend upon a difference of density. Besides, if either the density attributed to amylacetic ether by M. Delffs or that attributed by M. Kopp were taken, this would not be sufficient to explain the superiority of its magneto-rotatory power. There is probably, then, in the different molecular grouping presented by these two isomeric liquids a cause of the greater magneto-rotatory power of amylacetic than of ethylvaleric ether.

Subsequently having been obliged by M. Wurtz with specimens of some isomeric liquids prepared by himself, I was enabled to extend the field of my investigation. Besides the acetate of amyle and valerate of ethyle which I had already submitted to experiment under the names of amylacetic ether and ethylvaleric ether, M. Wurtz sent me a specimen of butyrate of isopropyle, which is isomeric with the two others. The general formula of these three compounds is, according to the notation adopted by M. Wurtz, who doubles the number for hydrogen, C7 H14 O2 instead of C14 H14 04. Following the system adopted by M. Wurtz, the formula becomes :

For acetate of amyle,
For valerate of ethyle,
For butyrate of isopropyle,

C2 H3 (C5 H11) 02:
C2 H5 (C5 H9) O2;
C3 H7 (C4 H7) O2.