with the water. About 15 c.c. of this mixture was drawn into the injecting cannula. The cannula consisted of a piece of glass tubing drawn to a sharp point, with a lateral opening a short distance behind the point. This opening was made of a size just sufficient to allow the dust to pass easily through. No force was needed in the injection, the negative pressure due to the inspiratory act being sufficient to suck the moisture into the cavity.

The effects due to the simple dust were observed in several animals of both kinds. These consisted in hyperemia of the lung with exudation of fluid and red cells into the alveolar spaces. There was also exudation of fluid into the pleural cavity in several instances, and it contained usually a considerable number of red cells also. To some extent the epithelium covering, the pleural surface of the lung, was thrown off into the fluid.

The dust was rapidly absorbed into the lymphatic vessels, and could be seen, even after twelve hours, as small round yellowish patches in the pleural membrane. These when examined microscopically were found to contain large numbers of glass particles.

The injection of the bacillus pyocyaneus followed after a time, varying from one to twenty-six days.

In the following table we have a summary of the observations.

Guinea-pigs.-Six animals, with controls:

Two recovered completely.

Three survived the control animals for a period.

One was a contradictory instance, for this died sooner than the control

animal.

Rabbits-Seven animals, with controls:

Two recovered, the controls dying on the second day after infection. Three survived the controls in two cases one day, and in one case eight days.

Two cases in which the bacillus was injected into the opposite pleural cavity. One died with the control, the other earlier.

The conclusion which is derived from these experiments is that the early stages of inflammation confer a certain amount of immunity on the pleural cavity in which the inflammation has been set up. It does not seem to vary in its protective power during the first month or so after the inflammation has been set up. In those instances where there was not complete recovery it was usually found that bacilli occurred to a much greater extent in the pleural exudation of the control animals than in that of those infected with glass dust.

The death of the last two rabbits points to the fact that probably the protection conferred is strictly localised in the inflamed tissues.

5. On the Changes in Nerve Cells due to Functional Activity.
By GUSTAV MANN, M.D.

6. On the Effect of Gravity on the Circulation. By Dr. L. HILL.

7. Experimental Inquiry upon the Different Tracts of the Central Nervous System. By F. W. NUTT, M.D.

The origin and termination of the fibres of the fillet were investigated by noting the degeneration resulting from unilateral separation of the nuclei of Goll and Burdach from the arciform fibres issuing therefrom. The degenerated arciform fibres were traced by Marchi's method into the opposite interolivary layer, and thence into the fillet. The degenerated fibres of the fillet were followed most

distinctly into the optic thalamus, passing through the striæ medullares. They could not be traced beyond this to the cortex.

The author also gave a preliminary account of some experiments relating to the antero-lateral tract of Gowers. He considers his researches so far show that it is a crossed tract, consisting in the main of fibres proceeding to the middle lobe of the cerebellum, but also containing a few scattered fibres proceeding from the posterior roots, and apparently ending in the corpora quadrigemina.

MONDAY, AUGUST 13.

The following Papers were read :—

1. On the Mechanical Theory of Lymph Formation. By Dr. STarling.

2. On Lymph Formation. By WALTER S. LAZARUS-BARLOW, M.D.

During the course of experiments performed in the Pathological Laboratory, University of Cambridge, for the investigation of the pathology of the cedema which accompanies passive congestion the author was led to examine certain of the conditions that modify the flow of lymph in the normal animal.

He introduced a cannula into one of the lymphatics of the hind limb of a dog, and, as far as possible, ligatured all remaining lymphatics. The limb was then emptied as completely as possible of lymph by rapid and firm squeezing from the paw upwards, and an observation was immediately commenced upon the amount of lymph formed in the limb during one hour. The animal was under A.C.E. mixture and was kept absolutely at rest, excepting that coagulation of the lymph in the cannula was prevented by occasional gentle squeezing of the limb. Immediately before the expiry of the hour the limb was again emptied of lymph as completely as possibly by firm squeezing. The lymph was collected in carefully graduated tubes, and the amount thus obtained was regarded as the normal for the individual.

An elastic ligature previously arranged round the limb was then tightened to such an extent as, it was known from other experiments, would raise the pressure in the femoral vein from the normal 4-5 mm. of mercury to 25-35 mm. Exactly the same processes were carried out during the collection of lymph under these modified conditions as were described for the normal conditions. The duration of increased pressure was one hour. Lastly, the elastic ligature was removed, and the amount of lymph formed in an hour under normal conditions was again estimated.

The limb was therefore as free as possible from lymph immediately before beginning and immediately before ending each of the three portions of the investigation.

The author found that the amounts collected in the three periods were either absolutely identical, or that the amount collected during the period of high venous pressure was less than that collected when the pressure was normal. On no occasion did he find that an increase of venous pressure was accompanied by an increase in the amount of lymph-flow.

The amount of lymph-flow, however, being the product of two factors-viz. the amount of fluid poured out by the blood-vessels and any modifications in that amount introduced during the sojourn of the lymph in the tissues-it was necessary before concluding that an increase of lymph-formation does not accompany an increase in venous pressure to determine whether an excessive amount of fluid had accumulated in the tissues in the form of oedema, since it was possible that the increased venous pressure might have caused an increased outflow of fluid from the capillaries and venules, and that that increased outflow might not have shown

itself in an increased flow from the lymphatics, because it was in part stored up in the tissues as oedema fluid. To determine this point observations of the specific gravities of arterial blood and blood-plasma, venous blood and blood-plasma, of muscle and of skin were taken before and after the pressure in the femoral vein was raised. It was found that these underwent no changes whatever, either in the affected limb or in other parts of the body (with the exception of a rise in the specific gravity of the venous blood and blood-plasma in the affected limb) during venous obstruction, which caused a rise of pressure in the femoral vein, varying in individual cases from 48-75 mm. of mercury. The modification of specific gravity of the venous blood and blood-plasma in the affected limb manifestly depends upon its lengthened sojourn in the limb.

The author concludes, therefore, that increase of venous pressure in a limb for one hour not only does not cause an increase of lymph-flow from the lymphatics, it does not even cause an increased passage of fluid through the blood-vessel walls. Independently, however, of the venous pressure, the amount of lymph-flow is increased when the tissues have been starved for some time or are overstocked with their own katabolic products.

The following conditions were investigated :

1. Before and after prolonged complete anæmia (three hours), produced by an Esmarch's bandage.

2. Before and after hæmostasis, or complete cutting off of the limb, with whatever blood and lymph it might contain, from the rest of the body by means of a tight elastic bandage for one hour.

3. Before and after stimulation of the sciatic nerve, while the limb was completely anæmic and persistence in situ of the katabolic products, the whole lasting one hour.

All these conditions are followed by arterial dilatation in the part; and, inasmuch as arterial dilatation subsequent to section of the sciatic nerve was found by the author to be unaccompanied by any modification in the amount of lymphflow, he concludes that the increase occurring under the conditions given above is immediately conditioned by the excessive needs of the tissues.

Under extreme conditions, such as those just given, the effect of an increase of venous pressure is markedly different from what it is when such conditions have not been introduced, for now an increase of venous pressure is accompanied by an increase in the lymph-flow, while the lymph-flow diminishes when the venous pressure is again allowed to return to normal. In other words, the lymph-flow now varies directly-the author cannot say whether it be proportionately-with the venous pressure. It now, therefore, bears some resemblance to mechanical filtration. Inasmuch, however, as an obstruction to the outflow of blood from the venous side itself intensifies the necessity of the tissues by damming up in them the waste products they are so anxious to get rid of, this resemblance to mechanical filtration may, after all, be only apparent.

The author concludes, therefore, that lymph-formation does not depend upon purely mechanical conditions of the circulation, but he regards it as dependent upon the needs of the tissues: those needs are, in some as yet unrecognised way, made known to the circulatory apparatus, and lead to variations in the amount of blood-flow through the part, in extreme cases active arterial dilatation of the most marked kind being induced. In such extreme cases, further, the lymph-flow varies directly with the venous pressure, and there is a resemblance to mechanical filtration; but there are reasons for supposing that this resemblance is an apparent and not a real one.

3. On the Innervation of the Portal Vein.
By W. M. BAYLISS and Dr. Starling.

4. On some Vaso-dilator Reflexes. By W. M. BAYLISS.

5. On the Production of Heat in Hibernating Animals. By RAPHAEL DUBOIS, Professor of Physiology in the University of Lyons.

For several years the author has investigated the production of heat in the marmot in order to find out what part of the nervous system is essential for the rapid production of heat which takes place when the animal wakes up from its winter sleep.

Section of the spinal cord at the level of the fourth cervical vertebra prevents the animal from raising its temperature. Destruction of the grey substance of the brain produces a similar effect. If the section of the spinal cord be made at the level of the seventh cervical vertebra the animal grows warm slowly and incompletely; but if the operation be performed between the fourth and fifth dorsal vertebræ, then the curve of the rise of temperature presents the normal form. There is therefore a limited portion of the cord, between the fourth cervical and the first dorsal vertebræ, through which pass the centripetal or centrifugal, or both, impulses, placing the cortex of the cerebral hemispheres in communication with the rest of the organism. The pathway is through the grey substance of the spinal cord, for section of the antero-lateral or of the posterior columns does not prevent the animal from producing heat, whereas destruction of the grey matter in this limited portion of the cord produces the same effect as total section. This operation produces immediate loss of the power of movement and conscious sensibility in the greater part of the body; but the absence of the capacity to produce heat must not be attributed to this sensory and motor paralysis, for section a little lower, at the fourth dorsal vertebra, does not change the normal curve of temperature.

The section of the cord at the fourth cervical vertebra abolishes, on the one hand, the contractions of the thoracic muscles, the activity of which is very great during the rewarming and insignificant during the torpor; on the other hand, it cuts off the very important connections of the sympathetic system with the higher centres. If the cervical sympathetic be cut on both sides above the inferior cervical ganglion there is no very marked delay in the production of heat, whereas the rise of temperature is very slow and incomplete when the inferior cervical and the first thoracic ganglia are removed on both sides. These ganglia, however, are only connecting links, for the same result is obtained by section of the two splanchnic nerves or of the branches which pass directly from the abdominal sympathetics to the semilunar ganglia. Extirpation of the semilunar ganglia produces the same result as removal of the inferior cervical and first thoracic ganglia.

Experiments show that it is by acting upon the portal system that the sympathetic shares in the general process of heat production. It regulates the quantity and pressure of the blood which flows to the liver, and in this manner the produc tion of heat in the liver and the transformation of glycogen into sugar, to be utilised for combustion when the animal awakes. The thoracic muscles become exceedingly active when the animal awakes, and thus require more glycogen or other combustible material for their contraction.

A full account of this research will shortly be published in 'Les Annales de l'Université de Lyon.'

6. On 'Pigeons' Milk.' By E. WAYMOUTH REID, Professor of Physiology in University College, Dundee.

John Hunter (1786) discovered the fact that pigeons feed their young for some days after hatching upon a substance resembling the curd of milk, and formed in the lateral pouches of the crop of both cock and hen. This method of feeding is as yet only known in this tribe of birds.

Claude Bernard (1859) studied the nature of the substance, and found that

microscopically it consisted of masses of the epithelial scales of the crop mucosa. loaded with fat globules. An analysis made for Bernard by Leconte gave

'Casein' and salts
Fat
Water.

No sugar was present—a fact noted also by Hunter.

Hasse (1865) and more recently Max Teichmann (1889) have also written on the subject from the histological point of view.

The lateral pouches of the crop of the non-breeding pigeon are not glandular; the epithelium is stratified and free from fat, the submucosa provided with small vascular papillæ.

The change in the crop membrane necessary for the formation of the 'milk commences during the incubation of the eggs, and though not visible to the naked eye till two or three days before hatching, makes itself evident by the appearance of fat-droplets in the cells ten days before this event. The main change consists in a great thickening of the epithelium, accompanied by rugose folding with reticulation, while at the same time the structure becomes enormously vascular and capillaries penetrate the epithelial layers (Hasse and Teichmann).

Small pellets of curd-like matter form in the pits of the reticulated surface, and as soon as the young are hatched these are transferred by the parents to the crops of the 'squabs,' often to the extent of 40 per cent. of the weight of the bodies of the young.

In its histological features the process of formation of the 'milk' resembles more closely that of the formation of sebum than of milk, for whole masses of fat-holding cells are cast off from the walls of the pits in the membrane; yet, unlike the sebaceous process, the nuclei of the cells persist.

Interpapillary involutions, then, of the thickened stratified epithelium of the crop act as sebaceous glands during the period of formation of the milk.'

This period lasts for from seven to nine days after hatching, and the maximum

of activity is reached about the second day after hatching.

The young are fed almost exclusively on this substance for the first three days, though a few crushed grains are also supplied by the parents. The parents appear to crush the grains at first, though later they are supplied whole. This fact is accounted for by the condition of the gizzard membrane of the early 'squab,' for the horny secretion of the tubular glands of the mucosa takes some days to consolidate. No digestive ferments are supplied by the parents along with the 'milk,' and the proventriculus of the young'squab' pigeon, even at twelve hours, is rich in proteolytic ferment, its glycerine extract digesting fibrin with ease. The crops of neither adult (breeding or non-breeding) nor young birds form any amylolytic or proteolytic ferment; in both cases, however, multitudes of bacteria and cocci are present, and the acidity of the contents (reaction of Uffelmann, but no reaction with phloroglucin and vanillin) is probably due to lactic fermentation. The pancreas of the squab' is capable of digesting starch at the time of hatching.

In the squab the cell bodies of the 'milk' are dissolved off by the secretion of the proventriculus, and the fat set free in the gut is found in the cells of the villi, and also in the leucocytes of the blood. The fæces of the 'squab' are fat-free, though at an early stage they contain considerable proteid.

Though sugar is undoubtedly absent from the milk,' a young squab' pigeon. before it has received any food contains sugar. In one case a triple alcoholic extract of a minced and finally pulverised 'squab' yielded over 2 per cent. of its body weight of reducing sugar, while a subsequent triple aqueous extract gave 16 per cent. of the body weight of an amylose yielding sugar on boiling with dilute sulphuric acid. This amylose struck no colour with iodine, and attempts to demonstrate glycogen in the bodies of unfed squab' pigeons have failed, though the pectoral muscles of adult birds are very rich in this substance.

As regards the proteids of the milk,' extracts with normal saline solution" by trituration and digestion with thymol at 40° C. show absence of albumins, proteoses, and peptones; presence of globulin and of caseinogen (clot with rennet,