COMPARATIVE PATHOLOGY AND
THERAPEUTICS. Vol. XXIX.-N o.
MARCH 3I, I9I6.
By A. E. METTAM, B.SC., M.R.CV.S., with Clinical Notes by J. T. CRAIG, M.A., M.R.CV.S., Royal Veterinary College of Ireland, Dublin. THE term diabetes mellitus is used to indicate an excess of urine which contains sugar, usually glucose. In the condition diabetes insipidus there is an excess of water, but the other usual constituents of the urine are not necessarily in excess; indeed, the secretion is usually considered as merely greatly diluted by an excess of water. The cause of diabetes insipidus is unknown, though it has been described as associated with certain lesions of the higher nerve centres, such as neoplasms of the cerebellum, hysteria, and concussion. It also occurs not infrequently in t):1e horse in chronic bacterial infections, such as tuberculosis and glanders, and here it may be considered as toxic in origin, which may also be the explanation of the ephemeral diabetes insipidus described as following the consumption of damaged food-oats and hay. It is not, however, with diabetes insipidus that we have to deal now, but with the far more serious condition diabetes mellitus. The cause or causes of diabetes mellitus are quite unknown. It is not a pathological entity, but probably depends upon several factors, more or less unknown, acting at the same time. There are no constant pathological findings, and, even if we agree that certain changes are commonly observed, it is not capable of direct proof that they are the actual causes. The terms glycosuria and mellituria are also used to signify the presence of sugar in the urine, but these terms are usually employed to signify a temporary condition-one amenable to treatment, or which is evanescent, whilst diabetes mellitus is applied to the more A
serious permanent condition. It may be mentioned at the outset that sugars others than glucose or dextrose may appear in the urine owing to the tissues not being able to utilise the quantities of sugar absorbed from the alimentary canal, the nature of the sugar depending upon the carbohydrates of the diet.
Digestion of Carbohydrates. The carbohydrates of the diet are acted upon by the juices of the alimentary tract and converted into soluble sugars, which are absorbed from the bowel into the blood and conveyed to the liver. It is possible that sugars as such in the diet may be absorbed from the stomach, but the starches, etc., still unchanged are digested in the bowel, where they are acted upon by certain enzymes secreted into the intestine. When the acid chyme is poured into the duodenum the mucous membrane of the bowel is stimulated to produce a chemical messenger or hormone, which, absorbed into the blood, eventually reaches the pancreas, and in turn stimulates that organ to produce its secretion. The external secretion of the pancreas is poured into the bowel, and the enzymes in the secretion are activated by another enzyme-enterokinase-which is formed by the mucous membrane of the bowel. The activated secretion is now able to attack among other things the starches of the chyme, which are rapidly changed into sugars. Starch is wholly converted into glucose (dextrose), lactose (milk sugar) is split into glucose and galactose. If any lactose as such is absorbed it will appear in the urine and give rise to lactosuria. Galactose is converted into glucose in the liver, and any which escapes the conversion will be excreted by the kidneys in the urine. Cane sugar is split by the intestinal invertase (another enzyme) into dextrose and lzevulose, and any escaping this change will be excreted as cane sugar in the urine. Lzevulose is converted into dextrose by the liver, and any which is not so changed will be excreted unaltered. Excessive amounts of cane sugar in the diet may be followed by the presence of dextrose and lzevulose in the urine.! The sugar taken up by the portal blood from the intestine is conveyed to the liver, where it is temporarily stored as glycogen. The sugar is converted into glycogen through the action of a ferment present in the liver, and it is probable that the same ferment is also responsible for the conversion of the glycogen into sugar. It is a peculiarity of enzymes that their action is reversible. They are able to build up and break down. The sugar produced by the glycolysisthe breaking down of the glycogen-may be used up in the ordinary way in health, eventually decomposing into carbonic acid and water, or it may be again converted into glycogen and stored in the muscles for instance. The amount of glycogen present in the muscles varies in different species. It is relatively abundant in the flesh of horses, less in swine, and still less in ox beef, though exceptionally it may be present in certain muscles of cattle in large amount. I t occurs also in the flesh of the fcetus. If for some reason the sugar is not or cannot be made further use of by the body it is excreted in the urine, and the condition of glycosuria or diabetes mellitus is established. 1
"General Pathology;" "Diabetes," by Beddard: Arnold, London.
The question now arises, What is it which prevents the further breaking down of the sugar and its disappearance as CO 2 and H .O? ·Sugar occurs normally in the blood, and the amount present is ,remarkably constant for different animals. In the dog, for instance, the amount is about 1'0 to 1'5 parts per thousand. If the amount present in the blood exceeds this proportion a hyperglyc~mia occurs and sugar is excreted with the urine. N ow, what is it which 'prevents the reduction of the sugar and leads to its appearance in the blood in large quantities, and to the eventual loss and excretion through the kidneys? I t is this very question which as yet has not been satifactorily answered, and when it has the causation of diabetes will be solved. In discussing diabetes it is usual, and we may follow the ordinary ·custom, to describe it, at any rate from the experimental standpoint, as (I) alimentary, (2) toxic, (3) puncture, and (4) pancreatic. What follows will give the reasons for this subdivision or classification ·of possible causes. By a process of elimination we may arrive at a working hypothesis which for a time must do the work of the actual proven fact.
( I) A li11le1Ztary Glycosuria. This form of glycosuria is usually temporary. It follows the ·consumption of large quantities of carbohydrates and it disappears when they are withheld. Possibly the sugar may also be derived from proteins in the diet, when these are in excess. It may occur in ·dogs which are actually fasting, the sugar being obtained from the fats and protein of the animal. The cause of alimentary glycosuria 'may therefore be an excess of carbohydrates in the diet and inability on the part of the liver to fix the sugar by conversion into glycogen, or a too rapid reconversion of the glycogen into sugar which is not acted upon by a glycolytic ferment but accumulates in the blood to be excreted by the kidneys. In this connection, however, it may be as well to direct attention to the description by Mutch in the Practitioner of May 1915 (annotated in the British Medical Journal of 9th October 1915) of the duodenum in diabetes mellitus. The reference is all the more 'interesting when the condition of our own subject is taken into .account,! He describes changes as occurring in the duodenum, which becomes elongated and increased in calibre, and on examina·tion after death the walls are seen to be thick and fleshy and somewhat milky in appearance. The same structural alterations can be traced also in the upper jejunum. This enlargement is almost invariably associated with ileal or colonic stasis, and in some 'patients the delay is extreme and the consequent modification of the lower bowel so advanced that there can be no doubt but that alimentary stasis preceded the onset of diabetes. Mutch claims the increase in size of the duodenum as the predominant change, and from his observations he concludes that diabetes mellitus is 1 'Witbin tbe last few days Professor ·w. A. Thompson of Trinity College, Dublin, has described a similar condition as being found in a laboratory cat which developed diabetes ' mellitus.
associated with a great enlargement of the duodenum which cannot be wholly accounted for by coincident intestinal stasis. Ileal stasis is usually present in diabetes mellitus, and in proportion to· its severity affects the prognosis adversely; and the urine of diabetic patients usually contains one or more products of the action of bacilli coli on tryptophane and tyrosine. The condition of the small intestine of the dog described by Mutch is remarkable, and there are other alterations to be noticed when we describe the histological examination of the bowel. There was also in our case an excess of mucus present in the bowel, and the condition of the wall of the gall bladder indicated much formation of mucus therein. Now mucin is a gluco - proteid, a combination of proteid with a carbohydrate. Amido-carbohydrates -sugars which are intermediate between the amines and true carbohydrates-take part in the formation of these proteins. The carbohydrate radicle in most cases is glucosamine, or a polysacharide composed of glucosamine units, such as chitin or chondroitin. Mucin is digested by pepsin and trypsin. By hydrolysis the carbohydrate may be reduced to sugar. Some portion therefore of the sugar present in diabetes, even in fasting animals, may be derived from an excess of mucus formed in the alimentary tract. The amount formed, however, may not account for the extraordinary amount present in the blood. The alimentary glycosuria has also been associated with a lesion in the liver, such as cirrhosis, but in these cases it is likewise claimed that there is an associated affection of the pancreas, and that the liver affection is due to the same cause as that causing the pancreatic lesions. True, there must be some other cause or causes operating than cirrhosis of the liver, which is so common among the domesticated animals that if it had anything to do with the etiology of the condition we should expect to meet with cases of diabetes every day of the week. It is claimed, of course, as is too often the case, that the cause of the cirrhosis of the liver is alcoholism, and therefore that we may assume that alcohol is the exciting cause of the diabetes; but this is altogether unwarranted, as cirrhosis is a common lesion among the lower animals and alcoholism is not the cause of it. If a diabetic has cirrhosis of the liver with a like condition of the pancreas the lesions in both glands may be due to the same cause, and not necessarily alcohol, but the lesions are merely an accident coincidental' and have no relation with the diabetes.
Plzloridzin and Toxic Diabetes. Phloridzin is a glucoside found in the bark of apple, pear, plum, and cherry trees, and is composed of phloretin and the sugar phlorose, which will reduce copper solutions. Phloretin is the active element in the phloridzin. If phloridzin is given hypodermically it sets up glycosuria. The amount of urine secreted is increased and it contains sugar, but when the poison is withheld thediabetes ceases. It does not cause a permanent diabetes. If the phloridzin be given for a long time the animal loses condition and death may occur. The diabetes is quite independent of the diet, and it occurs in fasting dogs when all glycogen stored in the liver has been used up; the sugar therefore comes from the tissues of the (2)
dog. It is held by most investigators that the sugar which is normally present in the blood is not increased in amount, that there is no hyperglyc~mia, that the elimination of the sugar is not due to excess of sugar in the blood, but rather that the kidneys become more permeable to sugar, that sugar escapes through the kidneys, and that there are demands more frequent and greater upon the tissues to make good the loss. Later investigations appear to show that the loss of sugar by the urine is really due to some lesion of the kidney, for if the phloridzin be injected directly into the renal artery, the secretion of the kidney fed by the injected artery is the first to excrete sugar; receiving the poison first, it is the first to respond to its effects. . I t has also been shown that geese which have had their livers removed, and which have therefore lost the main storehouse of glycogen or one of them, also excrete sugar in the urine after injection of phloridzin. But other toxic agents besides phloridzin may cause glycosuria, such as strychnine (in which case it is believed to be due to the disturbance of liver function), phosphorus, arsenic, uranium salts, corrosive sublimate, carbonic monoxide, amyl nitrite, curari, chloral, nitrobenzine, chloroform, acetone, ether, opium, morphia. prussic acid, mineral acids, etc. It should be noted that some of these agents are extremely toxic, and that they have specific actions upon certain epithelial cells-of the renal tubules, for instance, or upon the liver cells. Uranium salts, corrosive sublimate, chrome salts, and arsenic act upon the epithelium of the convoluted tubules of the kidney, for which they may be considered specific poisons, and not less harmful to the liver cells may be chloroform, phosphorus, and other of the poisons mentioned. I t is difficult to give a satisfactory explanation of phloridzin diabetes, but a lesion is apparently established in the kidney, and the pathological kidney may excrete or pass sugar which it is unable to retain. Against this simple explanation must be placed the observed fact that in phloridzin diabetes the diabetic condition ceases when the phloridzin is withheld, from which it may be deduced that any lesion established is of only momentary duration, and that the kidney rapidly becomes normally resistant to the passage of sugar. It is maintained that there is no hyperglyccemia, no increase in the sugar content of the blood, but only a leakage of sugar from the blood through the permeable kidney. The amount of sugar present in the urine cannot be accounted for by the :'splitting of the phloridzin into its poisonous principle and the sugar phlorose which is excreted. Glycosuria due to phloridzin is diminished or prevented if a lesion of the kidney be produced by the administration of aloin or potassium chromate, and the glycosuria may be checked or diminished if cantharidin be given. It has been noticed further in human practice that glycosuria diminishes and may cease in patients cachectic from tuberculosis. Large doses of adrenalin will produce glycosuria, and it is said that painting the pancreas with solutions of adrenalin has the same effect. Diabetes has been observed associated with tumour formation in the adrenals. It is said also that glycosuria will follow the therapeutic use of thyroid extract, and similarly the use of pituitary extract injected subcutaneously. The extract of the
pituitary body (posterior portion) when injected acts as a diuretic, and if repeated doses be given the diuresis continues for some time. According- to Biedl a definite influence of pituitary substance upon the metabolism of the carbohydrates has not been observed. The sugar content of the blood is not augmented, but the exhibition of adrenalin to dogs which have been previously treated with pituitrin is followed by a remarkable increase in the excretion of sugar. In certain cases of acromegaly (a form of gigantism associated with enlargement of the pituitary body) complicated with diabetes it was found that the patients had a chronic interstitial pancreatitis.
(3) Puncture Diabetes. Diabetes has been observed in patients showing affections of the higher nerve centres, such as tumours involving the cerebellum or medulla, after certain injuries to the brain, in nervous affections, etc., and Claude Bernard showed experimentally that diabetes could be produced by puncture of the floor of the fourth ventricle. The medulla is punctured at a point bounded above by the origins of the auditory nerve and below by a line joining the nuclei of origin of the vagi nerves. A trochar is pushed through the skull, the occipital protuberance in the mid line being the point entered, and then through the cerebellum and the posterior and median columns of the medulla. It has also been shown that injury of the vermiform lobe of the cerebellum may be followed by diabetes without any injury of the medulla. Two hours later (in Claude Bernard's experiment) sugar appears in the urine, which has increased in quantity. In dogs glycosuria lasts longer than in rabbits. The sugar in the urine folIows an increase of sugar in the blood-a hyperglyczemia-and the amount in the blood may be '3 per cent. as compared with the normal 'I or ·I5. Success of this experiment appears to depend mainly upon the condition of the animal operated upon, which must be good; if the animal is in poor condition or has been starved, then little or no glycogen may be present in the urine. After the amount of sugar present in the urine has reached its maximum it gradually decreases in quantity until the urine becomes normal again. A good deal of dispute has been waged round the question as to the proper explanation of the result of the puncture. In the first place, the lesion of the medulla is quite near to the vaso-motor , centre, and some have attributed the results to stimulus of this centre. It seems, however, to be generally held that puncture diabetes is of nervous origin, and that it can be explained anatomically. The nerves to the liver come from the hepatic plexus, and they reach the viscus along with the branches of the hepatic artery and portal vein, which they accompany in their distribution. The hepatic plexus is derived from the solar plexus, which receives branches from the vagi and the splanchnics. The vagi impulses travel centripetally, whilst those of the splanchnics flow out from the central axis and are centrifugal. If the vagi be cut and stimulated at their central ends, then impulses pass to the medulla, are switched into the cord, down which they travel, pass out by the thoracic nerves into the splanchnics, and so on to the liver. If the splanchnics be severed stimulation of the cut vagi has no effect. So long as the splanchnics remain intact it is possible to influence the glycogen-
holding capacity of the liver and the conversion of glycogen into sugar. Apparently the glycogen is rapidly converted into sugar by increased activity of the liver cells stimulated to increased effort by the nervous impulses brought by the splanchnics. The increased activity of the liver cells manifests itself possibly by the formation of an enzyme or, it may be, an activating agent, the end result being the conversion of the stored glycogen into sugar. An enormous load of sugar is thereby turned into the blood, and this in turn is got rid of through the kidneys with the urine. In dogs the glycosuria may last for a week, but this apparently is the extreme. The amount of sugar present in the urine reaches 2 or 3 per cent. The sugar comes apparently entirely from the liver, and the examination of the organ from an experimental animal reveals the total absence of glycogen. Further, the liver has apparently lost the power of fixing glycogen, for if a solution of glucose be injected into the radicle of the portal vein a pronounced glycosuria ensues, where:i.s if a similar experiment be made on a normal dog no excess of sugar is found in the blood of the hepatic veins.
(4) Pancreatic Diabetes. Probably at the present time most clinicians believe that diabetes is due to some disease of the pancreas, whether any pathological change can be discerned or not. If no anatomical lesion can be found it is usual to fall back upon disturbance of function, and to assert that disturbance of function is possible without there being necessarily any change from the. normal structure. It will be as well to examine the experimental evidence offered by destruction of the gland or its total or partial removal. It is customary to describe the pancreas as providing two secretions, an external and an internal. The external is poured into the duodenum, in reply, as we have seen, to the stimulus brought to the gland by the secretion of the duodenal mucous membrane. This external secretion is activated by the enterokinase procluced by the mucous membrane of the bowel, and it is able to act upon the various constituents of the diet by its ferments, which attack the proteins, fats, and carbohydrates. The external secretion of the pancreas is of great value, but it is not indispensable; its secretion may be prevented from entering the bowel, and the subject will survive. The so-called internal secretion has never been isolated, but it is believed to be turned directly into the blood, and it is thought that it either assists in the breaking down of the sugar or prevents sugar formation. This internal secretion, it is presumed, is formed by special structures in the pancreas-the islets of Langerhans, small collections of epithelial cells independent of the gland tubules proper. It is said that these cell groups or islets have an independent blood supply, and that they escape destruction when the pancreas as a whole atrophies from disease. On the other hand, it is said that they are connected with the ordinary tubules of the gland, that they are of the nature of reserves, and that under extraordinary conditions they may actually reproduce gland tubes. The question of the relation of the islets to the gland as a whole and their function cannot yet be said to be definitely settled.
For many years-for more than a century, in fact-diabetes mellitus and changes in the pancreas were associated, and this opinion received a decided impetus by the experiments done on the gland by Minkowski, Von Mering, and de D0minicis in 1889. Briefly, the results of the German and Italian experiments are as under. In dogs, cats, and pigs removal of the pancreas was followed by a most severe form of diabetes mellitus. In rabbits, owing to the position of the gland, total removal was difficult, and the results were not so certain. In dogs, after removal of the pancreas, the day following the operation the urine contained about I per cent. of sugar, but on the second day the sugar content rose to 4 to 6 per cent., and on the third day it might be as high as 8 to IO per cent. The presence of the sugar in the urine was permanent -a persistent diabetes was established, but it was possible to influence the amount secreted. If a dog from which the pancreas has been removed be fasted the sugar gradually diminishes in amount in the urine, but it does not disappear, even though the animal be starved for seven days. If the animal was given ordinary diet the amount of sugar passed varied according to the diet, being increased if much carbohydrate was given. The sugar in the urine was glucose, ann in course of time as the animals became emaciated the amount fell away, and, as is sometimes observed in man, occasionally disappeared completely prior to death. The experimental animals developed symptoms interesting to clinicians; the appetite became voracious, there was intense thirst, and if the animal was allowed plenty of water polyuria occurred. Despite the amount of food consumed the animals lost condition and became very emaciated and weak. If the pancreas was only partially extirpated the course of events was very different. If a portion of the gland was left in situ undisturbed no diabetes ensued. The amount required to be left is variously estimated as t inch (Von Meringand Minkowski), -[6 inch (Vaughan Harley), to -l-5 inch (Bayliss, Principles of General Physiology, 1915), but, as has been pointed out, it is difficult to estimate what is the proportion of the gland left behind. In the case of experimental removal of portion of the gland, even though the greater part of the gland be removed no diabetes follows the operation. Here it is as well to remark that this result disproves the contention that the diabetes is the result of damage done to the solar plexus at the time of the extirpation of the complete pancreas, for it is not possible for a portion of the gland to be ablated and no damage to be done to the plexus. Moreover, removal of the plexus alone is not followed by diabetes. If a portion of the pancreas be resected and grafted beneath the skin of the abdominal wall, the vessels of supply being left intact, and if after the graft has taken and made its new connections with the structures of the abdominal wall the remainder of the gland be removed, no diabetes follows. But if now the graft beneath the skin of the abdominal wall be removed, or if it becomes necrotic, then diabetes follows, and persists till death. These experiments therefore show that total loss of gland is followed by diabetes mellitus. but that the whole gland is not essential to health, since if a sufficient portion
be left behind diabetes mellitus does not occur. It was found that the animals which had lost the whole of their pancreas died after a few weeks, not from the loss of the gland but from the wounds becoming -contaminated and not healing despite every effort to keep the parts aseptic. Pfluger found the cause to be increased amount of sugar in the tissues, but it may be that another explanation is nearer the truth, and that is that the phagocytes are not so actively at work; these are loaded with glycogen and the granular leucocytes are poor in their specific granules. An interesting variant in the experimental work upon the gland is that recorded by Forschbach on parabiosis. Two dogs were grafted together after the manner of Siamese twins, and from one the pancreas was removed. So long as the two dogs were left united no alteration was observed in the animal from which the gland had been removed, but when it was separated from its companion diabetes rapidly supervened. The experiment is held to prove the presence of an internal secretion which was available for both dogs whilst they were united; this, the internal secretion, influencing the production or destruction of the sugar. It has also been shown that the external secretion of the gland is not essential to life and has little or no influence on glyccemia. The gland ducts have been ligatured, or blocked by the injection of paraffin wax, so that no secretion could reach the intestine, and in these cases diabetes mellitus did not occur. The gland as an external secretory gland was thereby destroyed, but as a producer of an internal secretion it was still active. According to Ssobolew, the islets of Langerhans of the dog persist when a portion of the pancreas is transplanted into the abdominal wall. Fifty days after the operation the lobules of the gland are atrophied, but the islets of Langerhans are well preserved. In the case of the gland remaining in position and undisturbed but the ducts of which are tied, it is found that an interstitial pancreatitis occurs, but the islets of Langerhans remain almost wholly unchanged, though surrounded by newly formed connective tissue which has replaced the tubular acini of the gland. It is said the islets persists even at the end of 400 days, the main pancreatic duct alone representing the gland. No diabetes is said to occur in these cases. The view that diabetes mellitus is associated with disease of the islets of Langerhans, and therefore the consequence of a deficiency, a 'Change in the nature of, or abolition of, the intestinal secretion, is due to Opie- (Diseases oftlte Pancreas): "I (Opie, loco cit.) have shown (1900) that these bodies, the islets of Langerhans, are diseased in diabetes mellitus." But then it is admitted also that there may be lesions of the gland proper, such as atrophy, chronic interstitial pancreatitis, pancreatic calculi, necrosis, carcinoma, lipomatosis of the gland, abscess, etc. (It is interesting to note in passing that one of the -e arliest associations of diabetes mellitus and pancreatic diseases was published in 1788. The case was one of pancreatic calculi and -sclerosis). In ninety cases of diabetes examined by Opie, seventynine, or 88 per cent., showed lesions of the islets of Langerhans, in forty-nine there was sclerosis of the islets, in twenty-four cases hyaline degeneration often combined with sclerosis, in one case acute 'inflammation, in two cases hcemochromatosis with deposit of iron-
containing pigment. If the islets are respected serious changes may occur in the pancreas without diabetes occurring, but it occurs if the number of islets be below n ormal. It is, however, not easy to give an idea of the number of islets present in what may be claimed as normal pancreas. According to Opie, the islets are more numerous in the tail than in the head or body of the g land. In the . dog receiving an inj ection of pilocarpine th e islets are more numerous in the splenic end of the pancreas (56'6) than in the part in contact with the duodenum (37'6) or in the descending portion in the mesentery of the duodenum (6-6). Ina control animal the numbers found in corresponding situations were 55, 67"6, and 9'6,. which appears to show that the numbers present are very variable .. From the experiments we may conclude that although the external secretion of the pancreas influences profoundly the changes occurring in the food in the bowe l, and that when it is absent many substances may escape digestion altogether and be lost, its abse nce from the bowel is not the influ encin g cause of diabetes. According to the majority of observers it is to the internal secretion we must look in seeking for the cause. The internal secretion has not b een isolated and its exact function is not known. Does it keep in check the sugar-forming powers of the liver--inhibiting the breaking down of the glycogen-or does it split the sugar molecule and cause its reduction? There appears to be no definite evidence as to its real action, and there is no· consensus of opinion among physiologists for our g uidance. In t eleos tean fishes the isl ets of Langerhans are large, and preparations can be made from them for use in experimental injections and for examination of reactions in vitro. Diamare and Kuliabko declare that the ordinary cells of the gland produce an amylolytic ferment (amylase), but the cells of the islets possess the power of destroying dextrose, that is, they produce a glycolytic enzyme. This is the only direct ev idence we have of the nature of the internal secretion, and as ferments are reversible in . action it is possible that the ferment is not only able to break down the sugar molecule but also to build it up, and that it exercises thi s function in the liver or in concert with liver enzymes. It is probable also that the ferment may be present in the blood, for after standi ng it has· been found that the sugar content in the blood has fallen, possibly owing to the action of a glycolytic ferment. It has also been held tha t the diabetes is due to some undetermined liver lesion, and various lesions have been met with in patients dying from diabetes. But what is the determinin g factor is not clear. The liver may still be able to store up g lycoge n by converting the sugar brou gh t by the portal blood, but if there be a lesion of the splanchnic nerves or interference wit.h cond uctivity of impulses along these nerves to the liver sugar m ay escape in large quantities through the liver and a hyperglyc;emia be established. The fact that interference with the sympathetic nerves (splanchnics} which are distributed to the liver may cause diabetes is interesting when we remember the relation of the sympathetic system to the suprarenal bodies and their internal secretion. Adrenalin is provided by the medulla of the suprarenal bodies, and that is developed from the sympathetic system. The action of adrenalin is the action:
of the sympathetic system of nerves, which are vaso-constrictors, and its effect upo n vessels is precisely that of stimulation of the sympathetic fibres. If adren a lin be injected into the peritoneal cavity diabetes may be established, and if it be painted over the intact pancreas diabetes occurs. If into a mesenteric vein of a starved dog whose liver is free from glycogen a solution of sugar be injected only a minute qu an tity of sugar eve ntually reaches the urine, but if into a dog with puncture diabetes a similar injection be made a marked glycosuria is spee dily established. Further, if a I per cent. solution of common salt be injected into the blood stream glyc os uria occurs, but not if the splanchnics be cut. The glycosuria is due to stimulation of the centre in the medulla, just as is the case in puncture diabetes. . According to Biedl (Internal Secretions): "From the data . .. set forth we are justified in concludin g that the pancreas, by means of its internal secretion, inhibits the formation of sugar in th e liver, and that this inhibition is effected throug h the agency of certain nervous appara tuses. Suppress ion of pancreatic activity abolishes the normal check upon the formation of sugar, in consequence of which the glycogen present in the organ is released, that is to say, th e g lycogen derived from the carbohydrates in the food or, in default of these, from other material, is converted into glucose, and as a con sequence hyperglycce mia and glycosuria follow. The pancreatic hormone appears to affect the sam e peripheral nervous apparatus as adrenalin,. for, as Kauffman showed, resection of the splanchnic nerve is followed by pancreatic diabetes in exactly the same manner as adrenalin glycosuria." The hormone mentioned, the internal secretion, has n ot been isolated, but it has been claimed that a juice obtained by expression from the pancreas mixed with a si milar juice obtained from the voluntary muscle can reduce sugar, and the conclusion drawn naturally is that the action of the pancreas is C'to facilitate the con sumption of glucose by the muscles." It has recently been shown that the reducing powers of glucose solution after the acti on of the combined juices could be restored to its original height by boiling with I per cent. hydrochloric acid. The apparent disappearance only occurred in concentrated solution s of glucose, for if "the product of the action of the combined ex tracts was diluted ten times and allowed to stand the original reducing pow er returned." "It was evid ent, therefore, that the effect was due to the activation of some enzym e system, which acts, as usual, in a syn thetic manner on glucose, in a hydrolytic manner on the disaccharide formed in concentrated solutions of glucose" (Bayliss). Naturally, when it was shown that in many cases of diabetes the pancreas was at fault organo-therapy with the gland was attempted. It has proved quite unsatisfactory and useless. Raw pan creas and pancreatic extracts have been shown to be valueless in the treatment of diabetes. Before leaving the subject it is necessary to refer to the experimental work of Hedon ( 1913), mentioned by Bayliss in his Principles of General Pltysiology, 1915. Hedon's experiments show that glycosuria is due to the absence of the pancreatic hormone. An anast o mosis was made between the pancreatic vein of a n ormal dog
and the jugular vein of a dog from which the pancreas had been removed. As a result the glycosuria of the latter almost disappeared, and there was a diminution in the amount of sugar in the blood. The liver also has an important part to play in the process, and this is exemplified by the following. A portion "of the normal pancreas of another dog was intercalated by vascular anastomosis in the circulation of a diabetic dog. This had a similar effect to the previous form of experiment, but only when the venous blood of the pancreas, presumably containing the hormone, was allowed to pass through the liver, by anastomosis with the splenic vein of the diabetic dog " (Bayliss). From his experi ments Redon concludes that" the liver plays an important part, and that there is some influence exerted by the pancreatic hormone on the excretion of sugar by the kidneys." "The latter may be decreased permeability of the kidney to sugar, or there may be some effect upon the reabsorption of the sug ar from the glomerular filtrate, or there may be some change in the state in which the sugar exists in the blood." The presence of sugar in the urine-glycosuria-is an essential or specific symptom of diabetes. The amount of sugar present may be very great; more than 2 Ibs. weight may be secreted in a day by the human patient, and, according to PflUger," a depancreatised dog fed upon protein free from g lucosamine excreted in two months 2'5 kilogrammes (about S ~ Ibs. ) of dextrose , more than could possibly have been obtained from the carbohydrate stored in the body" (Beddard ). Other sugars besides g lucose may be present in the urine. Frequently the urin e has a peculiar fruity odour, and this with the sweetness of the urine had been recognised very early by the physicians. The odour is due to the presence of aceto-acetic acid. In addition to these there may be acetone and beta-hydroxy-butyric acid . The acid by oxidation is converted into aceto-acetic acid, and by the loss of CO 2 this last is converted into acetone. Acetone in the urin e (acetonuria) is not peculiar to diabetes; it occurs also in febrile conditions, and both acetone and aceto-acetic acid may be present in the urine during long continued inanition. The rapid formation of acetone bodies in diabetes sets up coma diabeticum, and this, which is really an acidosis or acid poisonin g, is a common cause of death. The subject we have had under observation was an Irish terrier bitch about nine years old. Hz'story.- The owner stated that up to within three months of the animal being brought into the clinic it had been healthy. Then she began to appear less lively and more easily tir ed at exercise and to pass large quantities of urine. Although the appetite was good, condition was gradually lost. About six weeks prior to the clinical examination her sight became impaired ; she frequently knocked up against objects, which she made no attempt to avoid. Symptoms.- The animal was admitted to the clinic on 3rd March 19I5. She was thin and scraggy over the shoulders and neck, but over the loins the condition was fair. The animal was bright, and ,quite normal in her movements. Pulse 120 and irregular, temperature 101'4° F., conjunctiva= injected, buccal mucous membrane slightly paler than normal. The appetite was normal; no marked
thirst. There was no objection by the patient to manipulation over the loins. Both eyes were affected with cataract, which involved nearly the whole of the lens that may be seen through the pupil. The pupil of the left eye was narrower than that of the right. No changes in the size of the pupils occurred when the animal was put into a dark room and then exposed to a bright light. With the katoptric test applied to each eye the first image only was clearly seen, the second was a blur, and the third was absent. The lens of each eye was pearly white in appearance, with stria': radiating from a central point towards the periphery. The lens appeared like one fixed in formalin. A sample of the urine was light coloured and had a sweetish odour. It was slightly acid, and the specific gravity was I050. Fehling's solution was reduced, a profuse red precipitate of cuprous oxide occurring. A quantity of urine was run into a fermentation tube, a piece of ordinary yeast the size of a pea was added, and the tube was then placed in the incubator at 33° C. Fermentation rapidly occurred. It was found by actual estimate that the urine contained.4·8 per cent. of glucose. There was no albumen present in the urine. Diagnosis: Diabetes mellitus.-Since the animal was blind and no amelioration of the condition was likely the owner asked for the animal to be destroyed. She was, however, kept under observation for some twelve days. No attempt at treatment was made, and the animal was fed with flesh and bread and milk. No marked thirst was noticed and the appetite remained good. The condition did not apparently alter. The quantity of urine passed daily varied from 20 to 30 fluid ounces. The specific gravity was taken on four subsequent occasions, and at each time was found to be higher than that observed at the first examination; it was I060. The amount of glucose present was again estimated, and found to be IO per cent. The temperature varied between IOI o and I02'5° F. The pulse rate remained constant. On the 13th March a blood count was made; the red blood corpuscles numbered 6,400,000 per cm. The blood serum examined by the fermentation test showed the presence of sugar. Post-mortem Examination.-The animal was destroyed by chloroform, and a post - mortem examination made immediately. The carcase was fairly well nourished. The liver was large and fatty, the central portions of the lobules being lighter in colour than the periphery. The liver substance was soft and friable. The gall bladder was somewhat distended, and on opening into it the mucous membrane was observed to have a peculiar granular appearance, due to presence of numerous minute cysts in the mucous membrane. The cysts varied from the size of a vetch seed downwards. Some of the cysts were dark brown in colour and appeared like gall stones, which at first sight they were thought to be. The pancreas was somewhat enlarged and of a peculiar dead white colour, but otherwise there was, to the naked eye, no structural abnormality. The small intestine appeared thickened, and the increased thickness was apparently due to changes in the mucous membrane, which was somewhat congested and covered with mucus. The ca':cum was abnormal, being a nearly straight blind tube about 2! inches long. There was
no attempt at the usual spiral formation. In the bowel several .specimens of the dipylidium caninum were found. The kidneys were large, light in colour when split along the boundary layer, and more friable than normal. The heart was flabby and dilated, the mitral valve slightly thickened and incompetent. The lungs were greyish-pink in colour and normal. Traces of the thymus gland were observed in the anterior mediastinum. No macroscopic lesions were observed in the central nervous system. Both eyeballs were examined. In each case the capsule of the lens \vas quite clear; the substance only was involved, and hardly any portion had ·escaped. The cataract'> were firm, had the appearance of motherof-pearl, and were irregular on the surface. The right was not quite so opaque as the left. 140 cc. of blood was collected during the .post-mortem examination from heart and lung vessels, and centrifugalised at 4000 revolutions per minute. The proportion 'of plasma to blood was as 2 is to S. The plasma was pinkish and contained sugar. Tested in the fermentation tube with yeast after two hours' :incubation, 2 inches of the tube was occupied by gas which had displaced the plasma. The plasma was found to contain '39 per cent. of glucose. The red marrow was examined in films made and stained. There was a marked deficiency of granular leucocytes and granular myelocytes. Indeed it was difficult to find any granule-containing cells at all. For histological examination portions of different organs and tissues were fixed in absolute alcohol, Zenker's solution, or FormolMUller, and afterwards sectioned and stained in a variety of ways. The Kidneys.-A few of the glomeruli showed fibrosis conversion -of the capillary tufts into more or less dense connective tissue, but most of them showed other changes, which appear to me eM.) characteristic. The capillaries of the glomeru lar tuft are widely ·dilated and relaxed. The vessels forming the tuft are very evident, and the endothelial cells with their nuclei are clearly seen. It has been denied that the capillaries of the glomerulus show nuclei in their walls, but 'the sections from these kidneys show the .capillary walls to be formed as usual by endothelial cells. Many are without blood, but in others the corpuscles are to be found, though they may appear only as ghosts. The loss in hcemoglobin may be ante-mortem, since the fixation of the kidney elements was otherwise good. The epithelial covering of the capillaries appears to be normal. Bowman's capsule appears to be normal; there is no thickening or cellular activity of the lining epithelium, nor is there any infiltration around the capsule. The convoluted tubules are normal save that they contain a granular, evidently precipitated, material, possibly the result of the fixative. There was no albumen in the urine which had left the kidney, and if this precipitate indicates albumen then in the lower parts of the tubule this albumen must be removed by absorption. I n Henle's tubes there is a great amount of fat in the epithelial cells, but apparently no glycogen, since staining for glycogen in the several ways recommended failed to reveal any of the carbohydrate. The amount of fat in Henle's tubes is remarkable; the epithelial cells in the paraffin sections are large, clear, and glistening, and the fat present does not seem to be ordinary infiltration fat. Fat is prac-
tically absent from the cells of the convoluted tubules and from the glomeruli, and none could be discovered in the structure of the blood vessels. A certain amount of venous congestion is present, many thin-walled veins being distended with blood. In certain parts of the kidney the descending tubule of Henle's loop showed the ,e pithelial cells to be heavily pigmented, the pigment being of a yellowish brown colour. It did not give the iron reaction, however, and its nature was not ascertained. The most remarkable feature ·of the kidney is undoubtedly the peculiar dilated relaxed condition of the capillaries of the glomeruli. Tlze Li'ver.- There is some fat in the liver cells about the central vein of the lobule. The fat is in minute globules in the liver cell, and the globules are numerous; only exceptionally is the appearance that of infiltration fat. The epithelium of the bile ducts contains
FIG. l. Section of the dog's kidney. It shows a glomerulus with relaxed c apillaries, and the walls of the same formed by nucleated endothelial cells. Another smaller glomerulus is fibrous and imperm eable. The change probably occurred before the diabet es b ecame established. The granular debris in the renal tubules is also seen.
numerous fat globules of minute size staining with osmic acid. In the liver cells generally, but more pronounced near to the central vein, there is an indefinable change readily recognised but not easily described. The liver cell does not react as usual to eosin; it is somewhat swollen, more translucent than normal, and the protoplasm appears to be reticulated or vacuolated. The nucleus of the cell, especially towards the centre of the lobule, does not stain so well as usual, and is somewhat shrunken and deformed. These cells appear to be in a condition of necrobiosis; they are dying though not dead, they have undergone some abnormal change which as yet has not affected the cells towards the periphery of the lobule.
In certain parts of the lobule some cells show the presence of infiltrated fat, and many of the liver cells when stained with Best's carmine method apparently contain glycogen. I am not altogether satisfied, however, that the carmine method gives us a true picture or a reliable indication of the glycogen content. It is very remarkable that the staining reaction indicative of the glycogen present should occur only on one side of the cell in all the cells of the section. It is to be expected that the glycogen would be diffused throughout the cell protoplasm. A moderate amount of cirrhosis occurs, especially about the bile ducts, and there is some chronic venous congestion, but insufficient as yet to damage the liver cells.
FIG. 2. Section of the dog's kidney. It shows a Bowman's capsule with its contained glomerulus. Leading, away from the capsule is the renal tubule with a dilated neck continuous with the proximal convoluted tubule, the commencement of which is indicated. The darkly stained tubules on the left of the photograph are descending tubules of Henle's loop, the epithelium being filled with pigment granules.
The Gall Bladder.- The mucous membrane of the healthy gall bladder of the dog shows an arrangement which recalls that of the reticulum of the ox. There are shallow compartments separated by delicate walls, recalling the honeycomb appearance of the ox's stomach. But there, of course, the likeness ends. The epithelium of the mucous membrane of the gall bladder is simple, tall, and
columnar, the nucleus being somewhat compressed and situated towards the fixed extremity of the cell. Above the nucleus the protoplasm is clear and transparent, or it may contain granules. A striated basilar border or a corresponding structure to tha t observed in the epithelial cells of the intestine occurs. The epithelial cells are supported by a basement membrane, nucleated and composed of dense connective tissue, below which lies a layer of non-striped muscle mixed with a not inconsiderable amount of elastic tissue. The gall bladder of this dog when opened contained a certain amount of bile, and the mucous membrane had a peculiar appearance. It appeared to be studded over with a fine deposit of dark brown calculi or gravel, but on further examination these were found not to be gall stones but minute cysts in the mucous membrane,
FIG. 3. Section of the dog's kidney. The two tubules in the centre of the photograph lying side by side and lined by It cubical or polygonal epithelium are the ascending tubes of Henle's loop. The cells are loaded with fat. The fat is not stained.
the contents of which had become stained with bile. On examining sections of the mucous membrane of the gall bladder it was found that the mucous membrane was somewhat hypertrophied, and that the cysts were formed of pent - up secretion owing to adhesion between the folds of the mucous membrane. The pent-up secretion was mucus, and apparently the epithelial cells lining the gall bladder are capable of producing abundance of mucus. The irregularities upon the surface of the mucous membrane of the gall bladder had become greatly enlarged, and by fusion had converted the shallow .compartments of the normal mucous membrane into cysts. Sections of the mucous membrane taken in different planes could all be .explained from this point of view. The gall bladder did not contain B
any calculi, and the common bile duct was quite patent to its orifice in the duodenum. The Pancreas.- The gland is a compound tubular gland of loose structure. The secreting lobules are held together by a lax arrangement of connective tissue. The lobules are formed of tubules lined by a single layer of epithelium of the low columnar type. The ceHs are of the serous variety, with a vesicular nucleus and a protoplasm containing' granules which are the prezymogen granules, the antecedents or mother granules of the ferment of the secretion. In addition to the ordinary cells of the gland tubules are the so-called centroacinous cells, which apparently are the commencements or originating cells of the intercalary ducts of the gland. They may be observed towards the extremities of the tubules lying upon the
FIG. 4. Section of the adrenal cortex. Zona fasciculata. The black masses are pigment. Large cells with several nuclei may be seen near to the pigment. l\Iany of the cells of the zone show fat (clear spaces).
chief cells of the gland. The cells of the tubules form the external secretion, which under stimulus is poured into the duodenum. In addition to the tubules are collections of epithelial cells, first discovered by Langerhans and known as the islets or islands of Langerhans. As has already been seen, it is concerning the function of these cells that much of the controversy has arisen, but later researches have indicated that the cells forming the islets of Langerhans produce the hormone which influences and controls the formation and dispersal of sugar in the body. In the work of Opie especially, the relation of the islets of Langerhans to diabetes mellitus has been particularly emphasised, but cases occur where the islets are unaffected in diabetes. The most typical specimen of the islets which I possess is from the pancreas of a man who died
from diabetes. This, however, is only of interest from the negative point of view. Diabetes may be due to a variety of causes, and there does not appear to be any constant dominating lesion in it. In Opie's cases the majority showed changes occurring in the pancreas, with lesions of the islets. In the pancreas of this dog a great number of sections were cut and examined from all parts of the gland, but although they were stained in different ways and by various combinations no islets were discovered. Sections made from control animals, dogs and cats, and from man, showed islets as usual. Otherwise the organ was normal; there was no sclerosis, no inflammation of either one variety or another. The ducts of the gland arid the blood vessels were alsc normal. I t has been shown that injections of paraffin wax or ligature of the main duct is followed by atrophy and disappearance of the gland
FIG. 5. Section of the pars anterior of the pituitary gland (hypophysis cerebri). The general appearance is well shown.
tubules and the epithelial lining, but that the islets are respected and so long as the islets remain diabetes does not occur. Prolonged stimulation of the gland by secretion does not lead to any change in the islets, and if only a small portion of the gland is left in the body there is no formation of new islets from the tubular structures. On the other hand, from Dale's experiments we learn that ligaturing "the duct destroys existing islets and acini equally, and that the islets found in the sclerosed pancreas are a new formation from acinous tissue. A similar conversion is brought about by severe inanition and by exhausting the gland by injections of secretin" (Beddard). It has also been suggested that the islets are reserve structures and may be converted into ordinary tubular glandular
structures, and, if so, we can understand why the islets have disappeared from the pancreas of this dog, which may have been in a condition or state of chronic stimulation. From the condition of the cells lining the tubules the gland appeared to be exhausted, the cells being practically free from granules. The Thyroids and Lymphatic Glands.- The thyroids were normal and so also were the lymphatic glands examined, save that a small gland near to the pancreas contained a very large amount of some iron-containing substance which took on the Prussian blue colour after ferrocyanide of potassium and hydrochloric acid. Tlte Spleen.- The spleen was contracted, the capsule puckered, and the endothelial cells of the serous covering were cubical in shape. The non-striped muscle fibres of the trabecula:: (and of the
FIG.6. Pars anterior and portion of pars intermedia of the pituitary gland. The appearance of vesicles or follicles is shown, and the secretion is stained blue (black in photograph).
capsule) were unusually prominent; indeed, it would be correct to say they were hypertrophied, and the increased size was most evident on cross section. The spleen pulp contained a good deal of ironcontaining pigment which gave the characteristic reaction with the ferrocyanide test, and some few giant cells, but otherwise was normal. There was a certain degree of hyaline degeneration of the arteries of the Malpighian corpuscles, but the lymphoid tissue of these structures was unaltered. The Suprarenal Glands.-Examinations of sections of the suprarenals proved of interest. Those treated by the Marchi method showed the cells of the cortex to contain fat (lipoids) in minute droplets which took on the black stain. The amount of fat was much less in the zona glomerulosa than in the other zones. The
cells of the medulla gave no reaction for fat save in a cell or two, in which the derivation of the fat was not quite clear. In the cortex there were signs of hcemorrhages into the zona reticularis, and in the zona fasciculata were masses of pigment which gave the iron reaction. The cells of the zona near to the pigment were vacuolated and apparently had once contained fat. Some of the cells contain large overgrown nuclei, others are multinucleated. These cells are much larger than those of the normal or ordinary cells of the zona, and it is not quite clear what is their origin or nature. I am inclined to think they must be considered as related to the foreign body giant cells, since they are only to be found near to the accumulations of pigment. The medulla of the gland is not without interest. The chromaffin cells are very pronounced in the fragments fixed with the chrome
FIG. 7. This is a more highly magnified field from fig. 6, showing the epithelium surrounding spaces filled with secretion. The appearance suggests the thyroid gland.
salts. Many of the cells contain clear translucent spaces, like vacuoles, occupied by a substance which has not stained nor is it attacked by osmic acid. Sections made from material treated by the Marchi method show complete absence of osmic acid reducing substance in the medulla, though the fat is stained in the cortex. The globules stain, however, a ruby red colour with magenta followed by indigo carmine, as in Podwyssozki's stain. It is probable that the droplets are some form of colloid, a degeneration of the cells of the medulla perhaps. In sections of the suprarenal portions of ganglia were met with upon the capsule of the organ, but in all cases the nerve cells appeared to be normal. Sections were made also from the ccrliacomesenteric plexus, but so far as the nervous clements are concerned
the plexus was quite normal. Here, however, as elsewhere, the vessels in the connective tissue binding together ganglia and nerves were unusually dilated and relaxed, the endothelial cells were very prominent, and the walls of the capillaries appear beaded with nuclei about a dilated lumen. Small Intestine.-We have previously referred to the opinion expressed by Mutch as to the condition of the duodenum in his patients and also to Professor Thompson's cat. In this dog the whole length of the small intestine showed the walls to be unusually thick. Apparently all the coats share in the increase, but not equally so, the mucous membrane being mainly affected. In vertical section the crypts of LieberkUhn are found to be very long and the lining epithelium shows many mitotic figures (four occurred in one
FIG. 8. Section of the mucous membrane of the gall bladder. The epithelium is well shown, and a collection of mucus in a false cyst may be seen. The mucous membrane is hypertrophied.
field of a single crypt viewed with a No. 6 objective and No. 4 eyepiece). Numerous leucocytes traverse the epithelium, passing into the lumen of the crypt. Round about the crypts the membrane contains numerous leucocytes, and towards the muscularis mucos;e the leucocytes are mainly of the eosinophile variety. The villi are unusually thick and stunted, crowded together, and approximately of the same length. The epithelium covering the villi is thrown into waves, ;J.S if the villi had shortened or contracted. Goblet cells occur in excess. The muscular fibres in the villi are well developed, and supporting the blood vessels are numerous leucocytes, many of which are migrating through the epithelium of the villi. Some of the villi are greatly congested, and in places there are evident h;emorrhages. Some of the villi have lost in part their epithelial cap, and this loss
must have occurred before death, since the bowel was fixed within a short period of death. A considerable amount of mucus covers the mucous membrane. Pituitary Gland (Hypophysis Cerebri).-l n view of the close interrelationship of the ductless glands, the pituitary gland of the dog was fixed and examined, and though the literature is now getting very extensive, and can only be touched upon here, a short account of the structure and function of the organ is included. The pituitary is composed of three parts-the anterior, intermediate, and posterior. The anterior is glandular, the posterior is composed of nerve matter, and so also is the intermediate part, which is connected with the infundibulum and links up the pituitary with the brain. The infundibulum may be hollow, and the cavity is in direct communication with the third ventricle of the brain.
FIG . 9. A section through the mucous membrane of the small intestine,showing the submucosa and portion of the muscular coat.
The anterior part is composed of irregular columns of cells deposited in a delicate connective. tissue very rich in dilated capillaries. Many of the cells are apparently epithelial cells and appear to be of two kinds, one with abundance of acidophilous protoplasm around the nucleus, the other with a minimum amount of protoplasm about the nucleus, which not infrequently is collapsed and irregular in shape. I t is not easy to determine if the cells are all of one type, the difference in appearance being due to activity and exhaustion, the acidophilous cells, very prominent in sections, being in the resting or active condition, the others being exhausted. Many, even of the eosinophilous type, show peculiar, dense, contracted, irregular pyknotic nuclei. The relationship of the eosinophilous cells to the blood vessels is quite interesting; they are close
to the wall and appear actually to rest upon the wall itself. There is little doubt but that the connection between the blood stream and the cell elements is a very intimate one indeed. In places there is the appearance of an attempt at follicular or alveolar formation. It is not rare to find the above-mentioned cells arranged about a space which contains a structureless substance suggesting a secretion formed by the cells about the space-a structure recalling in some respects the vesicles of the thyroid gland. With ha:matoxylin and eosin this substance stains of a dull slate colour, with Mallory's connective tissue stain it takes on a sky-blue colour. With the latter reagent the eosinophilous cells stain a dark ruby red, very similar to that of the corpuscles in the vessels. In some follicles the presence of an odd cell staining red, and others about the lumen with a vacuolated protoplasm not taking the red, suggests exhaustion of cells in producing the secretion. Schafer describes a basophile variety of cell which he says is much less common than the acidophil. In places there are accumulations of a third variety of cell which apparently is a lymphocyte, and therefore in the anterior part of the pituitary body is included fragments of lymphoid tissue. It is in the portion of the anterior part, near to the pars intermedia, that we find the alveolar arrangement best marked, and sometimes the contents of the alveoli or follicles appear to be escaping into the cavity of the pars intermedia. It is considered as likely that some secretion does pass from the pituitary body into the third ventricle (eventually), and afterwards mingles with the cerebrospinal fluid. In this pituitary there is plenty of evidence of activity; there are numerous follicles occupied with secretion and some of it passing, its actual state suggests flowing, into the cleft in the pars intermedia. The condition of the vascular supply and the state of the cells all suggest great activity of this portion of the pituitary. The relation of the eosinophilous cells to the blood vessels seems to warrant the assumption that the cells form the internal secretion which is turned into the blood. "Under certain circumstances," says Schafer,!" in man or mammals, and especially in cases of thyroid absence or insufficiency, a product of the secretion of the anterior lobe cells which has the appearance of colloid appears to accumulate between them, and in part the cells come to be arranged round the colloid in the form of vesicles which are not very unlike those of the thyroid gland. Normally this appearance is not observed in the pars anterior (Herring), although it is characteristically seen in the pars intermedia." Extracts of the pituitary gland have a remarkable influence upon the vascular system, causing a great rise of blood pressure, and apparently it is from the pars intermedia and pars posterior that the extracts are obtained. Whatever their nature may be, apparently the active principles of the extract are produced in the pars intermedia and pass into the pars posterior. The effect of the extract upon the blood vessels is a direct one, and not due to a stimulating action upon the sympathetic endings, as with adrenalin. The extract also acts upon the unstriIJed muscle of the uterus, upon 1 "Lane Medical Lectures," 1913. An Introduction to the Study of the Endocrine Glands and Internal Secretions. Leland Stanford, Judor University Puhlications.
the kidney, and upon the mammary gland. It is also said that the extracts affect carbohydrate storage, causing glycogen to disappear from the liver, and favouring the production of alimentary glycosuria by lowering the limit of sugar assimilation. The gland has been removed from the living animal and the effects of loss of the gland noted. After partial removal a temporary glycosuria ensues, but it soon passes off, and there follows increased tolerance for sugar; "animals which have acquired this tolerance wiII suffer loss of the pancreas without becoming diabetic." Injury to the posterior lobe followed, it is believed, by liberation of stored up secretion, gives rise also to marked glycosuria and polyuria; the glycosuria may pass off but the polyuria remains. The polyuria (diabetes insipidus) may alone occur. The relation of the thyroid to the pituitary is of more than usual interest, because certain animals may live for months after the thyroid has been removed. After thyriodectomy the gland enlarges, the sella turcica deepens, the colloid is in greater amount, and the formation of false vesicles is more common. This I have verified in a goat which mentally deteriorated. It became stupid and listless, remained indefinitely in one position, and made no effort to clean itself. The skin became thickened, wrinkled, and very scurfy, and hay seeds began to sprout from the tangles-encrusted hair. Upon the kidney the pituitary extract has a remarkable action. Injection of the extract causes constriction of the vessels generally, but not those of the kidney, which indeed dilate; the kidney swells, and there is a remarkable flow of urine. Even if there be suppression of the urine during an