ade at intervals of 2 hours, the results being given under No. 21.

See Table I.

der to increase the total change and lessen proportionally the experiment, it became necessary to use a stronger solution, to the length of column, and to reduce the interval elapsing first contact and first observation as far as possible. To hese conditions the following method was adopted :-About 3. of powdered lactin were rubbed in a mortar with about centims. of water for half an hour, the solution filtered, and observation taken one hour after first contact. The metal onging to the polariscope was also discarded, and a glass one ted from a piece of tubing 17 millims. wide, by sealing on a side r the introduction of a thermometer, and grinding the ends until it measured 242 millims., the greatest length admitted polarimeter. Two glass disks were cemented on the exs, and the tube covered from end to end by a helix of thin tin through which a current of water might be passed to keep the ture constant; to guard further from variations in temperature e was covered with cotton wadding. With these precautions periments were made (Table I, Nos. 1, 2, 3), the result being è total change was nearly doubled. In all the other experihe method was slightly varied, the lactin being placed in a ith a ground glass stopper, 60 cub. centims. water placed on the whole shaken vigorously at intervals for half an hour, and the first observation taken as before. Each experitended over six hours, and included ten observations. For servation three or four readings were made, and the average In Nos. 4, 5, 6, 7, 8, 9, 13, 14, 15, 16, 17, 18, 19, 20, of the nying table, varying weights of sodic and potassic chloride roduced. In every experiment the thermometer was read at e time with the rotation; and the average temperature, as its extreme variation, is given in the table. That the different ents might be compared, we have expressed them by the

コー

y=a+bx+cx2,

ay is the angle of rotation, a the time in half-hours, counting Le first contact of the lactin with water, and a, b, and c, are ts. The values of a, b, and c, were calculated by the method squares.

able II are given the equations, accompanied by the probable a single comparison of the calculated and experimental values The sum of the actual errors is in nearly all cases zero.

1

By the aid of these equations we can now calculate the initial specific rotation of lactin, or the rotation when x=0 calculated to unit of weight. When x=0, y=a; and the permanent rotation being known,

the initial specific rotation =

ax 59.17

permanent rotation

The following are

the values found, the chloride experiments being averaged by themselves.

·0177

·0401

....

*0455

....

....

·0088

*0255

91.37

95 .30

92.16

Average of the twenty experiments, 92°.63.

dy

On differentiating the equations, putting=0, and calculating the

da

values of x and y, we find that the values of y thus got do not correspond to the permanent rotation, but are always greater; showing that the change in rotatory power does not progress according to the

same law throughout, but that, at the point referred to, a new reaction begins. This value of y is proportional to the amount of lactin in solution, indicated by the permanent rotation; and the specific rotation calculated from it in the different experiments is practically constant, its average value (from twenty experiments) being 64°.8. The following are the values of x and У when

dy-0.

da

This break in the change seems to point to the dual nature of lactin mentioned by Fudakowski,* whose experiments show that lactin, like cannose, gives two glucoses--lacto glucose and galactose.

An increase of temperature evidently hastens the change; but the exact relation of temperature to the rate of change has not been discovered.

The presence of sodic or potassic chloride increases the amount of lactin in solution, but has no apparent effect on the rate of change.

IV. Action of Hydric Nitrate on Lactin.—We made an attempt to trace this action, but did not succeed in overcoming experimental difficulties. The first of these was the impossibility of completing the action with the quantity of acid required for the first change. If a

* "Deut. Chem. Ges. Ber.," ix, 42–44.

larger quantity of acid were used, the first changes were so rapid as to evade measurement; moreover, the oxalic acid formed, by crystallizing in the acid liquid, made accurate observation impossible. By adding the acid in small successive portions, we nevertheless succeeded in obtaining an outline of the reaction, of which the curve drawn below is an accurate general expression.

Dubrunfaut, who has also examined this action,* asserts that the rotatory power first rises to 1 of the original amount, then falls 용 gradually to zero, again rises to of the original rotation, and once more falls to zero: the highest rotation corresponding to galactose, the first point of inactivity to mucic acid; the second rise probably to dextro-tartaric acid; and the second fall to the formation of oxalic acid. Our experiments show the formation of a lavo-rotatory substance, perhaps lævo-tartaric acid. The general form of the curve constitutes it an interesting and novel addition to chemical curves.

V. Note on Solubility.-The mutual relations of water and lactin in solution undergo a change upon which the change of rotation most probably depends. Water shaken with a large quantity of very finely powdered lactin at a temperature of 17° C., takes up a quantity of lactin corresponding to a solubility of 1 part lactin in 10-64 parts water. With four hours' contact, the solubility increases to 1 part lactin in 7-49 parts water. The permanent solubility got by the analysis of the mother-liquor of lactin crystallized over oil of vitriol is 1 part lactin in 3-23 parts water. In the solution of the lactin a fall of temperature of 0°45 C. was observed. Pohlt also found a depression of temperature (0°-88 C.); while Dubrunfaut alleges that heat is evolved.

Conclusions.

I. The initial specific rotation of lactin is 92° 63.

II. The permanent specific rotation of lactin is 59°.17.

* " Compt. rend.," xlii, 228.

"Journ. Pr. Chem.," lxxxii, 154.

III. The change of rotation of a solution of lactin can be expressed by a mathematical equation.

IV. When the specific rotation 64°.8 is reached, the law of change must be expressed by a different equation.

V. The initial solubility of lactin is 1 part lactin in 10-64 parts

water.

VI. The permanent solubility is 1 part lactin in 3-23 parts water.

IV. "On the Microrheometer." By J. B. HANNAY, F.R.S.E., F.C.S., lately Assistant Lecturer on Chemistry in the Owens College, Manchester. Communicated by H. E. ROSCOE, LL.D., F.R.S., Professor of Chemistry in the Owens College, Manchester. Received December 11, 1878.

(Abstract.)

In this paper the author reviews the work done by chemists and physicists in determining the relation between the chemical composition of a liquid and its rate of flow through a capillary tube. Poiseuille* ascertained, in a very accurate manner, all the physical laws relating to the rate of flow, as regulated by temperature, pressure, and dimensions of the tube; but on examining saline solutions he could make nothing of the numbers presented, because he used percentage solutions instead of solutions proportional to the equivalent of the body dissolved. Graham,† noticing that Poiseuille had discovered a hydrate of alcohol by running various mixtures of alcohol and water through the tube, examined mixtures of the various acids with water, and found that the hydration proceeded by distinct steps of multiple proportions. Several others, notably Guerout, ‡ have since worked on the same subject, but as they have only worked on organic liquids, and have done all the rates at the same temperature, the results throw no light on the phenomena. Thus water runs about five times as quickly at 100° as at 0°; and in a series of alcohols, such as Guerout experimented upon, the differences between their boiling points were very great, so that, their vapour tensions or molecular mobilities being quite incomparable while at the same temperature, the experiments do not admit of any real interpretation. The author reserves the organic part of the investigation, which requires the determination of vapour tensions, till a future paper, and in the present deals with saline solutions.

The phenomenon of the flow of liquids through capillary tubes has

* "Ann. de Chim. et de Physique," [3], t. vii, 50.

+ "Phil. Trans.," 1861, p. 373.

"Comptes rendus," lxxix, p. 1201; lxxxi, p. 1025.