the tubules and the fluid inside by which a certain amount of the water passes again into the blood, and so leaves the urine in a more concentrated state than it was when it first passed from the glomerulus into the dilated end of the urine tubes (Ludwig). It is believed, however, that the epithelium, which lines the convoluted tubes, performs certain functions in connection with the secretion of the solid constituents.

The rapidity of the secretion of urine may be said to depend upon the following factors: (1) The relationship which exists between the pressure of the blood in the glomerulus of vessels and the urine in the capsule of the Malpighian body and in the tubuli uriniferi; (2) the state of the blood pressure in the venous system of the kidney; (3) the pressure upon the lymphatics; (4) the quality of the blood in the artery of the Malpighian tuft; (5) the state of the walls of the artery constituting the Malpighian tuft, and of the capsule itself, these being regarded as the filter through which the fluids and soluble constituents of the blood have to pass. The influence of vaso-motor nerves upon secretion must not, however, be forgotten; not only do they exert an influence upon the quantity and quality of the secretion by dilating or contracting the arterioles, but their influence upon the chemical processes, by reason of their communications with the secreting cells (Pflüger), must be remembered; (6) activity of tubular epithelium.

In the experiments I have endeavoured to imitate the conditions found in the Malpighian body of the kidney. I have not been able, however, to represent the lymphatic arrangement.

The apparatus, a drawing of which, kindly executed for me by my friend Dr Robert Moffatt, is shown in the following woodcut. It consists of a piece of rabbit's bowel A2 enclosed in a glass tube B2. To each end of the bowel a small glass T-tube is attached. One of those tubes is connected with a pressure-bottle A, and a manometer A1. The pressure exercised upon the fluid inside the bowel by the pressure-bottle and indicated by the manometer A1 will be designated the afferent pressure. The tube attached to the other side of the bowel conveys the fluid that passes along the bowel to the vessel B, and the pressure exercised upon the fluid which it contains, called the efferent resistance, is indicated by the manometer B1. Through a cork at the right hand side of the large tube, another

T-tube communicates, on the one hand, with the inside of the large glass tube containing the bowel, and, on the other, the vessel E and manometer C1. The vessel E will therefore contain the fluid that filters through the membrane A2.

A2 B

B'

The index in the manometer A1 will therefore represent the afferent pressure, and correspond to the arterial tension of the renal artery; B1 will indicate the efferent resistance (when applied to the kidney, the venous resistance), while the manometer C1 will show the resistance offered by the fluid in the tube B2 to the transudation of the fluid inside the bowel, and therefore correspond to the tension of the urine in the capsule of the Malpighian body.

In the first series of experiments water was passed into the bowel under an afferent pressure of from 10 to 50 mm. of mercury, the efferent resistance being the same in each experiment as the afferent pressure, so that no water passed along the bowel to B. The amount of fluid which transuded through the bowel was found to increase in accordance with the pressure used. It was shown that for every 10 mm. increase in the pressure, there was 533 c.c. more water filtered through the bowel per minute. Thus under a pressure of 10 mm. 2.133 c.c. transuded in a minute, and when a pressure of 50 mm. was applied 4.266 c.c. passed in the same time. From these results we would therefore conclude that the amount of fluid which transudes through an animal membrane is increased according to the tension of the fluid inside. In relation to this experiment, take the following:-Instead of an animal membrane the kidney of a recently killed horse was employed; in this case a three-quarters per cent. salt solution was used instead of water, as it

was found that when water is passed into the vessels it almost immediately passes through the walls and causes oedema of the tissue, and the onward flow of the fluid is prevented. The salt solution seems, however, not to pass through the walls of the vessels into the lymphatic spaces so readily if the kidney is quite fresh, but still it passes from the glomeruli into the dilated end of the tubuli uriniferi. I found it very difficult to get this experiment to work satisfactorily, as the kidney requires to be used immediately after the death of the animal, and a number of precautions need to be taken which it is not necessary to mention here.

In the experiments alluded to the salt solution was passed into the artery under various pressures, the venous resistance being equal to 20 mm. in all except the first, in which case no resistance was offered to the exit of the fluid by the vein. In the first experiment the solution seemed simply to pass from the arterial into the venous system, very little being pressed into the urine tubules. When, however, the efferent resistance is raised to 20 mm., and at the same time the afferent pressure advanced to 40 mm., the increase in the amount of fluid pressed into the ureter is obvious. In the other experiments upon the kidneys of animals, the results of which I will not give in detail, a somewhat similar plan was adopted. The following are the results:-(1.) When the fluid contained in the ureter is subjected to pressure, the quantity of fluid that passes from the vein is diminished in relation to the pressure employed, and so also is the amount of fluid that transudes from the glomerulus into the capsule of the Malpighian body. (2.) The quantity of fluid that passes from the vein depends upon the amount of afferent pressure; the greatest increase takes place between 40 and 50 mm. (3.) The temperature of the fluid affects the rapidity of the flow through the vessels and the quantity that transudes into the tubuli uriniferi. The higher the temperature the greater is the amount of fluid passed from the ureter, and the more rapid the circulation through the vessels. (4.) When the fluid is pressed into the artery, it finds its way readily into the vein, but when injected into the vein, it does not escape by the artery. There must, therefore, be some arrangement in the kidney, probably in the Malpighian body, by which regurgitation of the fluid is prevented.

The results of the experiments with the bowel show (1.) that

the amount of fluid that transudes is in accordance with the pressure upon the fluid inside. (2.) For every 10 mm. increase in the afferent pressure 275 c.c. more fluid transudes per minute, and the flow along the bowel is increased; whereas, when the efferent resistance is increased, the amount of fluid that transudes is augmented by 31 c.c. in the same time, and the flow along the bowel is diminished. Therefore the afferent pressure may be said to be expended in two ways-increasing the amount of fluid that transudes, and the quantity that passes along the bowel-but the efferent resistance exerts its whole force in pressing the fluid through the membrane, therefore, 10 mm. increase in the efferent resistance has more effect than the same increase in the afferent pressure, and for the same reason we would suppose that a given increase in the venous resistance would conduce more to rapid secretion of urine than the same increase in the arterial pressure, unless when the venous resistance is extreme when other factors come into play. (3.) The addition of urea slightly retards the transudation of water through the membrane. The filtrate contains the same percentage, whatever pressure may be employed, as the original solution. (4.) Albumen also retards the transudation of water, but it differs from urea in this respect, that the percentage of albumen in the filtrate is in relation to the pressure. The higher the pressure the larger the quantity of albumen in a given amount of the filtrate. (5.) The presence of urea in a solution of albumen assists the filtration of the albumen at the expense of the urea. The following table shows the results:

(6.) The higher the temperature of the solution the more rapid the transudation of the fluid. Thus, when water was passed into the bowel at a temperature of 15.9° C., and under a pressure of 45 mm., 1425 c.c. filtered through in thirty minutes; whilst, when

4 R

VOL. IX.

the temperature alone was raised to 34-2° C., 197 c.c. transuded in the same time.

At the beginning of this paper I referred to Ludwig's theory regarding the secretion of urine, namely, that the blood is subjected to a high pressure inside the glomeruli, a free filtration into the dilated end of the tubuli uriniferi takes place, and this filtrate, which is at first very dilute, gradually parts with a portion of the water that holds it in solution. This is believed to take place by a process of diffusion between the fluid in the tubuli uriniferi and the blood in the veins surrounding them on all sides. Now, if it were not that albumen retards to a certain extent the passage of crystalloids (salts and urea) through an animal membrane, then the fluid in the inside of the urine tubules would be of the same concentration (in crystalloids) as the blood. But it has been shown that when a solution of albumen and urea are filtered through an animal membrane under pressure, the filtrate is less concentrated than the original solution, particularly as regards the amount of albumen, but also to a slight extent the urea. This is more especially the case when the pressure is not great. If the blood contained nothing but crystalloids (urea and salts) then the fluid inside the tubuli uriniferi would be the same as the fluid circulating in the vessels, and no diffusion would take place during the passage of the urine from the glomerulus to the pelvis of the kidney. This is, however, not the case; the blood circulating in the vessels contains a large quantity of albumen, and, if the theory above stated be correct, more urea than the fluid in the tubules, so that, putting aside any special function the epithelium may have, diffusion must result, and a portion of the water in the tubules pass back again into the blood. This diffusion will take place as the urine passes along the tubuli uriniferi either till it becomes of the same concentration as the blood outside, or makes its escape into the common ducts that convey it to the pelvis of the kidney. Therefore, the longer the urine remains in the tubuli uriniferi, other things being equal, the more concentrated will it be.

4. Note of a Method of Studying the Binocular Vision of Colour. By John G. M'Kendrick, M.D.

There are several well-known methods of mixing colours, such as the superposition of two spectra or of different parts of the same