that part where the calcified cartilage was thickest (viz., at e in fig. 2), that the osseous lining of the calcified cartilage was found almost universally present; and we have seen that this condition implies an arrest in the process of conversion of calcified cartilage into cancellous structure. The languid condition of the ossific process at this part, was probably also the cause of the piece of cartilage c in fig. 2, being left uncalcified below the general level of calcification.

The general conclusion to which the examination of this exostosis leads, is that it grew at the surface as cartilage, which became converted into cancellated bone by an ordinary process of ossification, in which, however, the stage of calcification of cartilage occupied an unusually conspicuous position.

ARTICLE III.-Contributions to Toxicology. By DOUGLAS MACLAGAN, M.D., F.R.S.E, etc. (In continuation from Vol. XVI. p. 22.)

Oil of Bitter Almonds. Experimental investigation of the alleged poisonous quality of the Oil when freed from Hydrocyanic Acid.

(Read before the Pharmaceutical Society, November 10, 1853.)

IN the month of September 1853, a servant girl in a family in Edinburgh was poisoned, from having swallowed a small quantity of a bottle of flavouring matter sold under the name of quintessence of ratafia, but which, instead of being a solution of essential oil of bitter almonds in spirit, proved to be the undiluted oil itself. The case was sent to me for investigation, but as I was from home at the time, it was forwarded to Dr George Wilson, who ascertained these facts. Dr Wilson estimated the quantity of the essential oil which the girl had swallowed at a little under half a fluid drachm.

The ordinary commercial essential oil of bitter almonds, as is universally known, consists of a peculiar oil, (hydruret of benzule,) to which its peculiar aroma and pungency are due, associated with anhydrous prussic acid; the proportion of the latter varying from three to fourteen per cent. This hydrocyanated oil is therefore always more or less dangerous as a poison, generally highly so. It has also all along been matter of general belief that the potency of the oil as a poison was greater, the larger the amount of prussic aicd which it contained, but it has been disputed whether its poisonous action was due solely to the prussic acid, and whether the hydruret of benzule itself did not contribute to its noxious qualities. The hydrocyanic acid, as is well known to chemists, can be separated from the hydruret of benzule, and thus that constituent of the crude oil, which is admitted to be the chief source of its poisonous properties, can be removed without impairing the peculiar flavour for which the oil is used in cookery. Dr Wilson brought this fact under the notice of the authorities here, who called public attention

to it by a newspaper paragraph. Several of our chemists in Edinburgh, who had been accustomed to deal in flavouring essences made with the hydrocyanated oil, forthwith acting upon this intimation, set themselves to prepare oil free from prussic acid, and essences containing only pure hydruret of benzule, are now to be found in all the best establishments in this city.

The separation of the hydrocyanic acid from the hydruret of benzule is theoretically simple enough, but to accomplish this completely is not so easy, as the statements in chemical works would lead us to suppose. Distillation from some base, such as lime, potassa, or baryta, has been adopted, and the process most commonly given in books is to agitate the crude oil with a large excess of lime or caustic potash, to add to this a considerable quantity of a salt of iron and to distil; the prussic acid remaining behind in the form of prussian blue. The proceeding which seems to have met with most favour at the hands of our pharmaceutists in Edinburgh, is to agitate the oil with a large bulk of a mixture of milk of lime and caustic potash, and to distil without adding any salt of iron. As the use of common sulphate of iron, however, does not materially add to the expense, and as the compound, formed with the iron, is more stable, and less likely to yield up again any of its hydrocyanic acid, than the cyanides of calcium and potassium which are formed when no iron is present, I should be disposed to prefer the addition of the sulphate. The object of the present communication, however, being toxicological and not pharmaceutical, I do not enter into any details on this point, nor should I have expressed any opinion regarding it, had it not been that, operating on the small scale of two or three ounces of oil, I have succeeded easily and completely in removing all traces of prussic acid by the iron process. The most economical methods of working will doubtless soon be ascertained, because some of the wholesale establishments in London which deal extensively in this article, are now prepairing the pure hydruret of benzule on the large scale. So far as our present experience goes, it appears to be essential to success, whatever process is followed, that the agents destined to retain the prussic acid should be brought by continued agitation into accurate and prolonged contact with the oil. The prussic acid adheres to the oil with great tenacity, and notwithstanding all precautions, a second distillation has generally been required before the oil can be pronounced entirely free from prussic acid.'

Whatever method of purification be adopted, it is necessary that the distilled product be accurately tested for prussic acid. The ordinary iron test is the simplest and best, and if performed with a little

It was suggested by a chemical friend that perhaps a single distillation might be made to yield at once oil free from prussic acid, by putting the lime or potassa into the still along with the marc of bitter almonds and water. This a priori did not appear to be likely, from the known facility with which the amygdalin and emulsin, the two constituents of the almond which yield the hydrocyanated oil, are attacked by alkalies. To satisfy my friend, I made the trial, and found as I expected, that no oil at all was obtained, the distillate reacting and smelling strongly of ammonia.

care, is capable of revealing exceedingly minute traces of prussic acid. For this purpose twenty or thirty minims of the sample under trial are to be agitated for several minutes in a test tube with twice their bulk of aqua potassæ of the ordinary pharmacopoeia strength, and the mixture thrown upon a wetted filter, by which the oil is all retained, and a clear fluid is got, to which a few drops of a solution of sulphate of iron are to be added. This solution answers best if it has been made for some time, and has by the action of the air become partially peroxidated. Hydrochloric acid is now added in sufficient quantity to re-dissolve the excess of oxide of iron, which is thrown down by the potash, and if there be any prussic acid present, the prussian blue in the fluid becomes at once apparent. Agitation of the oil with water is quite insufficient for bringing the prussic acid within the agency of the iron test. The oil seems to prevent the water from dissolving out the prussic acid. It t appears frequently to have happened that in a sample of oil which has, when newly distilled, been tested and pronounced free from prussic acid, a subsequent trial by the same operator has shown it to be present in notable quantity. This may of course sometimes be due to the agitation having been more carefully attended to in a second trial. But perhaps it may be the case that by rest the prussic acid has some tendency to separate spontaneously from the oil, and that thus the tests detect it more readily. It has been suggested, that another explanation of this may be, that in process of time, prussic acid, which has actually been at first completely removed by the rectification, is re-developed in the oil by keeping; this, if true, would form a serious objection to the use, as a flavouring condiment, of the oil, however carefully rectified. It is not the case however. First, Because pure hydruret of benzule contains no nitrogen, as may be proved by applying to it Lassaigne's elegant process, of heating it with a fragment of potassium, and subsequently testing it for the cyanide in the usual way by the iron and hydrochloric acid. Secondly, In a sample of oil rectified by myself and found free from prussic acid in February 1852, the application of the iron test to half a drachm in November 1853, did not reveal the least trace of prussic acid. Lassaigne's test also showed it to contain no azote, and of course in the absence of this element, no hydrocyanic acid can be developed. There is, therefore, if due care be taken, no difficulty, either in freeing the oil from prussic acid, or on determining that this has been successfully accomplished.

But the important question in relation to public welfare is,-supposing the oil to be quite free from prussic acid, does it or does it not possess poisonous properties? As very conflicting statements have been put forth on this point, it will be proper briefly to notice what has been said regarding it.

Vogel of Munich, in 1821, appears to have been the first to make direct experiments on the subject. He used oil which had been rectified by distilling it twice from a concentrated solution of caustic potash. He inferred it to be pure, because he could find no hydro

cyanic acid in the residue left in the retort on the second distillation. One drop of this oil placed on the tongue of a small bird proved instantly fatal with strong convulsions; but four drops on a puppy six weeks old produced only temporary exhaustion. From these experiments, very unsatisfactory both as regards the physiological effects and the purity of the oil, he concluded that the oil was an active poison. Orfila is stated by Dr Christison to have confirmed these results by his experiments. I have not been able to find the original account of any experiments by Orfila with pure oil. In his Médicine Légale, vol. iii. p. 744, he does not cite any experiments of his own, but says in reference to the hydruret of benzule, "It is extremely poisonous and acts nearly like the empyreumatic oil of tobacco."

Liebig (Chemie Organique, 1843) states simply that the hydruret of benzule, when swallowed, acts as a poison; and Dr Gregory (Hand-book of Organic Chemistry, 1852) says, "It ought to be borne in mind that the commercial oil is highly poisonous, not only because it contains hydrocyanic acid, but because the hydruret of benzule is poisonous."

On the other hand, it has been contended upon experimental evidence that the oil if quite free from prussic acid is not poisonous.

Stange (Büchner's Repertorium, xiv. and xvi.) rectified the oil by caustic baryta, and found after this treatment that five drops of the oil killed a kitten in four minutes. He afterwards found, however, that the oil which he had regarded as pure still contained prussic acid. He rectified it again, and then found that in doses of from four to twelve drops it did not injure a rabbit, nor did six drops placed in the cellular tissue act as a poison. He found that of the unrectified oil two drops placed in the cellular tissue produced fatal effects on a rabbit.

Schroeder and Hertwig, at Berlin, 1826, used a salt of iron as a means of removing the hydrocyanic acid. They found that neither dogs, cats, nor pigeons were poisoned by the purified oil. The pigeons and cats got from three to eight drops, the dogs twenty or thirty drops, without injury. Hertwig himself swallowed ten drops without experiencing any other effects than those due to its pungency in the mouth. Göppert of Breslau found that four drops of the purified oil had no effect on small birds, and twenty-one drops none on a rabbit, except what was due to the pungency. Robiquet and Boutron Charlard also came to the conclusion that if quite free from hydrocyanic acid it was not poisonous.

Wöhler and Frieriehs, 1848, in the course of their experiments on the elimination of foreign matters by the kidneys, gave to a small dog two grammes, equal to thirty-two grains, with no poisonous effect. They showed that the oil in passing into the urine was converted into hippuric acid, but that if the dose was large, part of the oil passed through the kidneys unchanged. This observation, which has been abundantly confirmed by other experimentalists, I have also been able to verify.

Of the more recent British authorities almost all countenance the idea that the purified oil is not poisonous. Dr Christison, as a deduction from the various experiments quoted by him, states that it is not more poisonous than other volatile oils, and that "if the purified oil retains active poisonous properties, it must be owing to the hydrocyanic acid not having been entirely removed."

Dr Taylor (On Poisons) states shortly that "the oil deprived of prussic acid and perfectly pure does not act injuriously;" and in reference to this he quotes a paper by Mr Grindley.—(Pharm. Journ., vol. vii.)

Mr Grindley used red oxide of mercury, lime, and chloride of iron for purifying the oil, as he states that he could not succeed with the lime and iron salt alone. Mr Grindley's purified oil, subjected to Lassaigne's test, showed that it contained no nitrogen, and was therefore free from all traces of prussic acid. Of this purified oil five, and afterwards ten, drops given to a rabbit produced no bad effect. Drs Royle and Neligan, in their respective works on Materia Medica, both admit that the pure oil is not poisonous.

An experiment of Dr Pereira's, as narrated in his own words in the posthumous volume of his Elements of Materia Medica, p. 1776, shows how readily mistakes may be made as to the purity of the oil. "In some earlier experiments which I made on this subject, I found it to be highly poisonous, though I could not detect an atom of hydrocyanic acid in it. After the sample had been kept a few months, however, I readily detected the acid in it by the potash and iron test. By a second and third rectification, I completely deprived it of all traces of the acid, and I then found that four drops of it giyen to a small rabbit had no more effect than the same quantity of any other volatile oil; that is, the animal appeared dull for a few minutes, and the respiration was quickened." But to the article on Bitter Almond, in which this statement of Pereira's is made, there is appended the following remark by his editors, Drs Taylor and Rees, "Mitscherlich has experimented on the effects of this oil on dogs and rabbits; he found it poisonous even when quite free from hydrocyanic acid." This statement, made without any qualification, and backed by the authority of so eminent a toxicologist as my friend Dr Taylor, was somewhat alarming to those, who believed, that in selling essences made of the purified oil, they were vending harmless articles. I may remark in passing that there is a trifling error in this statement. Dr C. G. Mitscherlich does not appear to have made any experiments on dogs. I have not seen his original paper, which seems to have been printed in the Medizinische Zeitung, 1848, but in a very full abstract of it in Büchner's Repertorium for 1850, I find no mention of any experiments save on rabbits. Dr Mitscherlich's observations are known in England chiefly from a very brief abstract in the Pharmaceutical Journal, vol. x., translated from the Pharmaceutisches Central Blatt.

The more important points in Mitscherlich's experiments are the following. Two drachms of the purified oil were injected into the