in conjunction with his wife, both in western and eastern tropical Africa, and his modest volumes on Angola and the River Congo, dedicated to his partner in the pleasures and dangers of life in a tropical climate, and his zealous aid in the collecting of objects of natural history. He was one of the three almost contemporaneous discoverers of that very remarkable plant the Welwitschia mirabilis the others being Welwitsch and Baines; and he sent some of the finest specimens of it in existence to this

country.

After the loss of her husband, Mrs. Monteiro returned to Delagoa Bay, and spent five years in solitude, in the cottage built for her under happier circumstances, devoting her time to collecting insects, birds, and other natural objects, and studying the life history of insects and their relations to plants. The present book is an unpretentious narrative of her life and labours during that period, and a record of her observations and her experiments in breeding insects, illustrated with some of her own discoveries in the animal and vegetable kingdoms.

A Hand-book of Industrial Organic Chemistry. By Samuel P. Sadtler, Ph.D. (Philadelphia: J. B. Lippincott Company, 1891.)

In this book Prof. Sadtler has attempted to compress into about 500 octavo pages an account of those manufactures which depend upon the applications of organic chemistry. For what particular class of readers such a book is intended is rather difficult to determine. The scientific man is hardly likely to consult it in preference to the numerous special manuals to which he has access; and to the manufacturer the book is practically useless, owing to the comparative absence of all working detail. Considering the volume of literature which is required to give an approximately adequate representation of one industry aloneviz. the tar-colour manufacture-it would seem hopeless to expect anything of value from a chapter on the artificial colouring-matters, which, in well-leaded "roman spaced," attempts to give in 45 pages an account of the production and chemical nature of the numerous artificial and natural organic colouring-matters used in the arts, including their identification, chemical analysis, and detection on dyed fabrics. Certain of the other subjects are. it must be stated in fairness, treated with greater detail; and, as we should expect from Dr. Sadtler's connectionas an expert with the mineral oil industry, his description of the manufacture of petroleum and its associated products is reasonably complete. So also is the account of the cane-sugar industry. But, with the exception of the bibliographical and statistical information which occupies a relatively large share of the space devoted to each article, we see little else to commend. The book, however, is well got up; the paper and printing are all that can be desired, and the illustrations are, as a rule, much better executed than is usual in works of this class.

THE examples contained in this book are of the most elementary nature, and are intended for the use of those who have got no further than quadratic equations. In this series the exercises only deal with arithmetic and algebra, and are arranged in sets of papers which gradually become more difficult. The examples in arithmetic commence by dealing with the first four rules, simple and compound, and fractions; while those in algebra consist mostly of numerical values, addition and subtraction. Cube root and compound interest in arithmetic, and quadratic equations in algebra, form the highest limit to which these subjects are carried in this series. Throughout the work the author seems to have paid great care to insure accuracy in the answers; and

though we have worked out many problems, picked out at random, we failed to find any errors.

We may mention that, in working through the papers, the beginner will occasionally come across examples which appear to be far above the average standard; but these, on trial, will always be found very simple, and are placed there with the intention of encouraging boys to look up methods they have not reached, and so to find that "a little research enables them to do a new sort of question."

Teachers and taught alike should find this book a useful adjunct to the text-book they have in use.

LETTERS TO THE EDITOR.

W.

[The Editor does not hold himself responsible for opinions expressed by his correspondents. Neither can he undertake to return, or to correspond with the writers of, rejected manuscripts intended for this or any other part of NATURE. No notice is taken of anonymous communications.]

The Implications of Science.

It would be a great misfortune if such views as were expressed

by Dr. St. George Mivart in a lecture delivered under the ægis of the Royal Institution, and reported at length in your columns (pp. 60 and 82), were allowed to pass unchallenged. In case no abler challenger appears, will you allow me to say a few words about "the implications of science"?

66

The great objection I take to Dr. Mivart's view is, that he does not appear to recognize any distinction between a real and a verbal truth. He apparently puts our knowledge of "the law of contradiction" into precisely the same category as our knowledge of our own continuous existence," and draws but a slight distinction between these items of knowledge and such an item as the law of gravitation. Whereas, in fact, the soonly expresses a verbal convention. called "law of contradiction" is not a necessary truth at all, it It is not a law, but is of the nature of a definition. On the other hand, our knowledge of our own existence, in the present, comes to us by direct apprehension, and really is a necessary truth" to each of us individually; though, since our knowledge of our existence in the past depends on the accuracy of our memories, this latter may easily be erroneous. That the memory exists is of course indisputable, but it may well be that the fact it professes to recall either took place differently, or even did not take place at all. Our confidence in our memories depends upon induction -ultimately on inductio per enumerationem simplicem—in just the same way as our belief in the law of gravitation does, and neither of these items of knowledge can therefore be necessary truths, though we may well hold them with so strong a conviction that the distinction may for practical purposes be ignored.

The "implications of science which Dr. Mivart insists on are nearly all truisms (that is, purely verbal assertions)—all those to which he ascribes universal validity in any regions of time or space are such. I may repeat here what I have said elsewhere: "The supposed peculiar certainty of mathematical conclusions is solely due to the fact that they are truisms."

For example, the assertion "Two straight lines cannot inclose a space" is certainly not a "necessary truth." Either its terms those definitions, or else its terms are defined by denotation, as are defined by connotation, so that its truth depends solely on representing real things in space, and the truth of the assertion can only be proved by induction from actual experience with those things. In the first case the truth is arbitrary, not neces sary; and in the second case it might conceivably be false, as was shown by Helmholtz. It is of course true that the imaginary dwellers on a sphere might still conceive what we call "straight lines," but if they chose to reserve that term for geodesics of their space they would be within their rights in doing so. This is practically what Euclid does, and this is why he requires "axioms" which are not necessary truths; even though, in fact, they are true as far as we can test them.

So also there is no useful sense in saying that twice two must be equal to four under any conditions of time or space. Doubtless, if the inhabitants of the Dog Star defined "twice," "two," and " 'four" as we do, then "twice two" would to them be "four"; but to say that it was so could only give verbal in

formation.

And if the people in the Dog Star chose to define four as I + I + I, the so-called " necessary truth" would not even be true! Again, we do not "recognize that what we know 'is' cannot at the same time 'not be,' we define it to be so. To know that anything is," is indeed to possess real knowledge; but in order to conclude that therefore it cannot "not be," we require no further knowledge, except as to the meanings of the words employed in the argument. The "law of contradiction" never tells us whether anything "is" or "is not.” It only tells us that the terms "is" and "is not" are not applicable to the same thing. This is part of the definition of the terms. If anyone chooses to say a thing both "is" and "is not," there is no law against his doing so, only if he does so he is not talking the Queen's English. Dr. Mivart is wrong in speaking of the "objective absolute validity of the law of contradiction." Its validity is not only not objective at all, but even subjectively it is not absolute, but depends on the arbitrary meanings assigned to its terms. It is exactly on a par with the assertion that at chess one king cannot give check to another. EDWARD T. DIXON.

Trinity College, Cambridge, November 29.

The Koh-i-Nur.

ABSENCE from home and pressing business since my return have delayed my sending a reply to Prof. Maskelyne's second article upon the above subject (NATURE, November 5, p. 5). So far as I can discern Prof. Maskelyne's primary object in writing these articles, it is to endeavour to maintain the hypothesis put forward by him many years ago; and with this object in view he has made a number of statements, from which I have culled not a few that may be ranged under either of two heads -firstly, those which I believe can be shown to be distinctly contrary to the evidence; and secondly, those which, if not directly contradicted by the evidence, are quite unsupported by it. In my first reply I gave samples of these statements which afforded perfectly clear issues, and as these have been unanswered, it is useless to refer to others in detail at present.

Some readers of what has already been written have expressed to me their regret that finality has not been attained by this discussion. For my own part I have a feeling of sincere regret at any additional confusion being introduced into the subject. Some of the statements referred to may, unless a warning be given, be quoted in the future, as others have been in the past, by writers who may not have the means or may not be willing to take the trouble to refer to the original authors.

There are several references in Prof. Maskelyne's last article to authors with whose writings I have considered it to be my business and duty to make myself familiar. I possess their works, and of one of them I have recently published a detailed commentary, while of another I have a com nentary in course of preparation. Among these authors are Garcia de Orta and Chappuzeau, and Prof. Maskelyne's remarks lead me to conclude that he has not a very intimate acquaintance with their writings and with those of some of their contemporaries. From internal evidence it is practically certain that at the time Garcia wrote his book he had not visited the Mogul's Court, and could not, therefore, have seen his jewels, though, for the sake of argument, Prof. Maskelyne suggests he had. As for the discredited Chappuzeau, whose malicious statements are quoted without their refuta ion, I need only say that Prof. Joret's investigations have cleared Tavernier of the charges of plagiarism, &c., which were made against him, and they have further disclosed the fact that his own original manuscript documents, from which the "Travels" were prepared, are still extant (see preface to the second volume of my edition of the "Travels").

Now, as to the De Boot mistake, to which Prof. Maskelyne again refers as though it had an important bearing on the subject, it is the case that Mr. King, in a footnote, pointed out the error in De Boot's quoting as from Monardes. The footnote does not occur in Mr. King's account of the diamonds, but elsewhere. When I wrote, I had Prof. Maskelyne's quotation (Edinburgh Review), as from Mr. King, before me, and thus I was for the moment misled as to the extent of Mr. King's knowledge. Seeing, then, that it was Prof. Maskelyne's misquotation which misled me, his not having accepted my invitation to explain, coupled with his crowing over me for having been misled (by his own words), is one of the most extraordinary

features in this controversy. Two years ago I annotated my original paper with the remark that Mr. King had noticed the mistake of De Boot about Monardes, but it was then too late to correct the press.

The confusion which has most unfortunately been introduced into this subject by authors has now, it is to be fervently hoped, culminated in the publication by Prof. Maskelyne of a figure of a huge mounted jewel, which, going much further than his previous reference to it might have led one to expect, he labels "The Mogul." What the authority may be for this sketch, we are not clearly informed; all, apparently, that can be said for it is that "it speaks for itself." I cannot understand how Sir John Malcolm can be responsible for it, at least as it is labelled, because I know what he has published about the Shah's jewels, especially the Darya-i-Nur and its companion the Taj-e-mah. Kerr Porter, Eastwick, and others who have described the Shah's jewels, make no mention of the existence of any such stone as this figure represents.

"It speaks for itself"; and I must venture by two alternatives to hazard an interpretation of what it says. Firstly, the amorphous-looking mass may be intended to represent some uncut stone, possibly a ruby; but why should it be the Mogul's diamond, which is known to have been cut? Secondly, it seems to be more probable that the figure may have been taken from a native sketch which originally professed to represent, but greatly exaggerated the size, and omitted the facets, of the Koh-i-Nur. Prof. Maskelyne says it was accompanied by two other stones in the same mount: so was the Koh-i-nur (see the copies of the original model in the Tower and in several public museums). The character of the mount is somewhat similar to that in the Hon. Miss Eden's sketch of the Koh-i-Nur. This is all that, as it appears to me, can be legitimately deduced from this figure which has been left "to speak for itself."

As to Prof. Maskelyne's own sketch of the Koh-i-Nur, I thank him for it, because I think it may perhaps serve to aid readers who have not seen the original in accepting the hypothesis put forward by me, that it had been mutilated after cutting.

Through the kindness of Mr. L. Fletcher, F. R.S., Keeper of the Minerals in the British Museum, I have recently had an opportunity afforded me of seeing the original plaster model of the Koh-i-Nur, and of comparing it with a glass model similar to the one upon which my remarks as to the mutilation were based, and I find them to be identical in form and all essential details. V. BALL.

Dublin, November 13.

Pfaff's "Allgemeine Geologie als Exacte

Wissenschaft."

In this work (Leipzig, 1873) there is a speculation (on p. 162) that in early geological times the carbonic anhydride, while yet free on the surface of the earth, was sufficient in quantity to exert a pressure of 356 atmospheres. If this had been the condition of things at any time when the surface temperature was below the critical temperature (30°9 C.), it follows that abundant liquid carbonic anhydride flowed over the surface of the earth, or floated upon the seas; unless it be supposed, which is not probable, that this quantity could be held in solution in the water. Other very important and interesting effects are also involved. The statement of the 356 atmospheres has been quoted without question by so high an authority as Dr. Irving in his Metamorphism of Rocks.

Pfaff's result, however, is based on a statement of Bischof's (as quoted by Pfaff), that the calcium carbonate of all formations would suffice to cover the surface of the earth to a depth of 1000 füsse. Pfaff takes 44 per cent. of this to be CO2, and assumes the specific gravity of the rock to be 2'6.

On these data, and taking the fuss as = 0.3 metre (as stated elsewhere by Pfaff), the CO2 would exert a pressure, not of 356 atmospheres, but of 33'2, approximately. It appears, in fact, as if Pfaff's result was, through some oversight, calculated as just ten times too great.

Perhaps there is some other explanation of the discrepancy. But, lest it prove an error, I have thought well that attention should be drawn to it, the statement being made on such high authority. J. JOLY.

Physical Laboratory, Trinity College, Dublin.

SEISMOMETRY AND ENGINEERING IN RELATION TO THE RECENT EARTHQUAKE IN JAPAN.

AT 6.38 a.m. on October 28, I was awakened at my

house in Tokio by the long swinging motion of an earthquake. There was no noise of creaking timbers, and there were no shocks such as usually accompany earthquakes. It was an easy swing, which produced eness and nausea. As recorded by bracket seismographs this continued for ten or twelve minutes. During the interval there was ample time to study the movements of these instruments, and the conclusion that could not be avoided was that rather than acting as steady points these heavy masses were simply being swung from side to side-horizontal displacement was not being measured, but angles of tip were being recorded. That many of our seismographs are useless as recorders of horizontal motion whenever a vertical component of motion is recorded, is a view that I have held for many years, and therefore when these two have been recorded in conjunction I have been inclined to receive the records with caution.

Further, the measurement of vertical motion as recorded bs a horizontal lever arrangement can only be trusted if we can assure ourselves that the advance of the waves has been at right angles to the direction of the lever. If ths condition is not fulfilled, then the seismograph for vertical motion may also become a tip-recording instrurent. As another indication that during this particular earthquake earth tips occurred, I may mention that the witer in a tank with perpendicular sides which is about 25 feet deep, 60 feet long, and 30 feet broad, rose quickly, first on one side and then on the other, to a height of 3 or 4 feet-much in the same way that water would nse and fall in a basin that was being tipped from side

to s.de.

Assuming what is said to be correct, it must not be confided that modern seismographs are useless. For earthquakes where the motion is horizontal, they give reis which practically are absolutely correct. When vertical motion occurs, in many cases if not in all, the records must be interpreted in a new light. The so-called borizontal displacements may be employed in determining the maximum slope of a wave, and if from an instrument recording vertical motion we are assured that we have measured the vertical height of a wave, we can at least approximate to the length of the same. The period of he waves being recorded, it follows that the velocity of propagation may be calculated.

Although it seems possible to use our present bracket semographs as angle measurers, it is evident that there are other types of instruments, where swing due to inertia .m.nimized, which will act more satisfactorily. To tan a true measure of vertical displacement, the most event solution would be to use a number of lever arrangements in different azimuths. Other methods may, Bowever, suggest themselves.

stone-undoubtedly put up in the flimsiest manner-lie as heaps of ruin between Japanese buildings yet standing. Cotton mills have fallen in, whilst their tall brick chimneys have been whipped off at about half their height. Huge cast-iron columns, which, unlike chimneys, are uniform in section, acting as piers for railway bridges, have been cut in two near their base. In some instances these have been snapped into pieces much as we might snap a carrot, and the fragments thrown down upon the shingle beaches of the rivers. The greatest efforts appear to have been exerted where masonry piers carrying 200-feet girders over lengths of 1800 feet have been cut in two, and then danced and twisted over their solid foundations considerable distances from their true positions. These piers have a sectional area of 26 x 10 feet, and are from 30 to 50 feet in height. Embankments have been spread outwards or shot away, brick arches have fallen between their abutments, whilst the railway line self has been bent into a series of snake-like folds and I ummocked into waves. The greatest destruction has taken place on the Okazaki-Gifu plain, where we have all the phenomena-like the opening of crevasses, the spurting up of mud and water, the destruction of river banks, &c.- which usually accompany large earthquakes. At Ökazaki and Nagoya the castles have survived. The reason for this may be partly attributable to the better class of timber employed in their construction, but principally to their pyramidal form and to the fact that they are surrounded by moats. Here and there a temple has escaped destruction, partly, perhaps, on account of the quality of materials employed in its construction, but also in consequence of the multiplicity of joints which come between the roof and the supporting columns. At these joints there has been a basket-like yielding, and the interstice of the roof has not, therefore, acted with its whole force in tending to rupture its supports. Cn the hills which surround the plain, although the motion has been severe, the destruction is not so great. These hills are granites, paleozoic schists, and other rocks. There is nothing volcanic. In the small cuttings where the railroad passes from the hills out into the plain, no effects of disturbance are observable, the surface motion probably having been discharged at the faces of the inclosing embankments. The general appearance outside the cuttings, however, is as if some giant hand had taken rails and sleepers and rubbed them back and forth until the ballast lying between them was formed into huge bolster-like ridges. Crossing the hills and proceeding to other plains, it is noticeable that there has been more movement on the alluvium than on the rocks.

Earthquakes yet continue, and in the Gifu plain each one is preceded by a boom as if a heavy gun had been fired in some subterranean chamber. Although the survivors, who may number, perhaps, two millions, are. for the most part, destitute, have witnessed the most terrible scenes, and are yet surrounded by the dead and the dying, yet there is no panic. They hear a "boomb," and run laughing to the middle of the street to escape the shock which the unaccountable noises herald. The Japanese have their feelings, but on occasions of this sort there is no helplessness in consequence of hysteria or mental prostration. As to what happens with Europeans under like circumstances, I must leave readers to consult history. Tokio, November 7.

JOHN MILNE.

In

NH.

THE discovery of this remarkable compound of hydrogen and nitrogen by Prof. Curtius, in the chemical laboratory of the University of Kiel, formed one of the

most interesting chemical events of last year. The extraordinary nature of the compound-manifested by its fearfully explosive properties, together with its acid character, by virtue of which it forms salts with metals containing only metal and nitrogen-mark out for it a place among the most attractive of hitherto discovered substances. It was first obtained by Prof. Curtius in the form of a gas, by treating with soda a compound containing the organic radicle benzoyl in the place of the hydrogen atom, and subsequently warming the sodium salt thus produced with dilute sulphuric acid. The gas was described as possessing a frightfully penetrating odour, and as being absorbed by water with extreme avidity, forming a solution of strongly acid properties, which liberates hydrogen in contact with metals. So great, indeed, is the affinity of azoimide for water, that in these earlier experiments it was not found possible to collect the gas in the anhydrous state. Shortly after the publication of his first communication (see NATURE, vol. xlii. p. 615), an improved method of preparing the solution in water was devised by Prof. Curtius. It consisted in distilling a soda solution of a derivative containing the radicle of hippuric acid with dilute sulphuric acid. He was thus enabled to obtain a tolerably large quantity of the aqueous acid. By successive fractionation of this solution in water, and finally distilling the last product of the fractionation over fused calcium chloride, pure azoimide itself was eventually isolated, and found to be a volatile liquid, boiling at 37°.

Owing to the terribly explosive nature of both the free acid and its salts, the work has been attended with considerable danger, and has, unfortunately, been delayed by a lamentable accident which befell Prof. Curtius's assistant, Dr. Radenhausen, who was seriously injured by the explosion of a quantity of the anhydrous acid. At length, however, Prof. Curtius is able to publish some further particulars concerning the acid and its salts, and an important communication from him will be found in the current number of the Berichte of the German Chemical Society. The following is a brief account of these further researches, together with a résumé of the present state of our knowledge of this interesting compound and its derivatives.

Sources of Azoimide and its Derivatives. Azoimide and its salts have been obtained from two distinct sources, both organic. One source, the first employed by Prof. Curtius, is benzoyl-glycollic acid, C&H CO-O-CH,COOH; the second is hippuric acid, CH&CO-NH-CH,COOH. During the investigation of the reactions of his previously discovered compound of hydrogen and nitrogen, hydrazine, NH, Prof Curtius found that both benzoyl-glycollic and hippuric acids reacted with hydrazine hydrate, forming hydrazine derivatives.

Benzoyl-glycollic acid reacts with two molecules of hydrazine hydrate, forming benzoyl hydrazine, CH,CONH-NH, and the hydrazine derivative of acetic acid, NH-NH-CH,COOH, with elimination of water. When benzoyl hydrazine is treated with nitrous acid, it is conNO

verted into a nitroso derivative, CH&CO-N

When the sodium salt of azoimide is distilled with dilute

sulphuric acid, azoimide escapes as a gas, which condenses along with water in the form of an aqueous solution.

Hippuric acid reacts with one molecule of hydrazine with formation of hippuryl hydrazine, C,H,CO-NHCH,CONH-NH. When this substance is treated with nitrous acid, a compound is obtained which was at first considered to be a nitroso compound, but is now discovered to be in reality a diazo compound possessing the constitution CH,CO-NH-CH,CONH-N N-OH. This substance may be isolated in quantity, and yields salts of azoimide directly upon treatment with alkalies. If soda is employed the sodium salt of azoimide is obtained, from which azoimide itself may, as before, be liberated by distilling with dilute sulphuric acid. It is more convenient, however, as will be described later, to employ it directly for the preparation of the ammonium salt of azoimide by saturating its alcoholic solution with ammonia gas; from the ammonia salt, if desired, azoimide itself may be obtained by converting it into the insoluble silver salt, and distilling the latter with sulphuric acid.

Preparation and Properties of the Sodium Salt of N

Azoimide, Na-N

N

The method of preparing the sodium salt of azoimide, now adopted as most convenient by Prof. Curtius, is somewhat different from the earlier one just described, although based upon the same lines. Instead of benzoylglycolic acid, ethyl benzoate, CH,COOCH, is employed. This substance is converted readily into benzoyl hydrazine by treatment with hydrazine hydrate : CH5COOH + N2H, H2O

= CH CONH–NH,+ C,H,OH + H,O.

The benzoyl hydrazine is next treated with sodium nitrite and glacial acetic acid, whereby it is quantitatively transformed into benzoyl azoimide, the benzoyl derivative

of the new acid:

C&H

3

CONH-NH2+ HNO2 = CH,CO-N2+ 2H,O. The benzoyl azoimide thus obtained is finally dissolved in an equal weight of absolute alcohol, and the equivalent

of an atom of sodium is also dissolved in a little absolute alcohol, and the two solutions mixed; the mixture is then digested for several hours upon a water-bath, when the sodium replaces the benzoyl radicle, and ethyl benzoate is regenerated :

CH,CO-N2+ C2H2ONa = C,H,COOC2H5 + Na-Ny.

Upon cooling, the solution deposits crystals of the sodium salt, and the remainder may be precipitated from the mother-liquor by means of ether. The ethyl benzoate is recovered by distillation with very little loss, and may quantity of the sodium salt of azoimide. be employed again for the preparation of a further

The sodium salt, NaN, obtained by this method is substantially pure. It is very soluble in water, but is, strangely enough, not hygroscopic. It is almost insoluble in ether and alcohol. It gives a feebly alkaline reaction, and possesses a briny taste. The crystals do not explode

The ammonium salt, which is by far the most convenient salt to start with for the preparation of the free acid and its metallic salts, is best prepared from the curious diazo compound of the amide of hippuric acid, CH,CONHCH CO-NH-N=N-OH, before mentioned. This substance is readily obtained in calculated quantity by first acting with hydrazine hydrate upon the ethyl ether of hippuric acid, and subsequently treating the hippuryl hydrazine thus produced with sodium nitrite and glacial acetic acid. Diazo-hippuramide appears to be most prolific in its reactions. Prof. Curtius states that it reacts with almost every class of organic and inorganic bodies with which he has brought it in contact, and generally without the application of external heat. Thus, when treated with water, alcohol, haloid ethers (alkyls), aldehydes, free halogens, or hydrazine derivatives of organic acids, it evolves free nitrogen gas, and forms compounds which are derived from hippuramide by replacement of a hydrogen atom in the NH, group by the radicle of the reacting substance. On the other hand, when acted upon by alkalies, ammonia or substituted ammonias (amines, or by diamide (hydrazine) and its derivatives, salts of azoimide are formed. Thus last reaction, when ammonia is employed, forms the most convenient mode of obtaining the ammonium salt of azoimide.

About a pound of diazo-hippuramide is placed in a flask of 2 litres capacity, and covered with 600 grams of 85 per cent. alcohol. The flask is then placed in a freezing mixture, and ammonia gas is led in until the liquid is saturated with it. The flask and contents are then allowed to stand twenty-four hours in order to complete the reaction, when the diazo compound is quantitatively converted into hippuramide and the ammonium salt of azoimide:

CH_CONHCH_CO-NH-N_N–OH + 2NH,

= C ̧H ̧ÑO—NHCH,CONH, + NH ̧ −Ñ, + H2O. The liquid is then boiled, the flask being fitted with an upright condenser, until no more ammonia escapes, when the heat is removed, and the solution allowed to cool. After standing another twelve hours, the clear alcoholic solution is decanted from the mass of hippuramide crystals, and treated with four times its volume of ether, when 70 per cent. of the total yield of the ammonium salt is precipitated in the form of a white powder. The remaining 30 per cent. of the azoimide may be recovered by recrystal lizing the hippuramide from water, adding the motherliquor to the ethereal-alcoholic solution after removal of the precipitated ammonium salt, and treating the whole of the liquid with solutions of lead, silver, or mercurous salts, when the azoimide is precipitated in the form of the difficultly soluble lead, silver, or mercurous salts. The hippuramide is readily converted, by boiling with hydrazine hydrate, into hippuryl hydrazine, which may thus be used again for the preparation of more of the diazo compound.

The precipitated ammonium salt is washed with ether and dried in the air. The snow-white crystalline powder thus obtained, consisting of fine anisotropic needles, may be recrystallized from boiling alcohol. It is only

sparingly soluble in absolute alcohol, but on boiling for some time in a flask fitted with inverted condenser, the whole passes into solution. Upon cooling, the salt separates out in large colourless crystals, tabular in form, and frequently aggregated in step or fan-shaped forms. These aggregates often resemble those of ammonium chloride very closely, but the crystals do not belong to the cubic system. The crystals are readily soluble in water, and, upon allowing the aqueous solution to evaporate in vacuo, large transparent prisms are obtained, which, however, soon become turbid in air.

The ammonium salt of azoimide reacts in a feebly alkaline manner. It is not hygroscopic, although so readily soluble in water. It dissolves easily in 80 per cent. alcohol, but, as above described, with difficulty" in absolute alcohol. It is insoluble in ether and benzene. It is distinguished by its great volatility. When the crystals are allowed to lie exposed to air, they gradually disappear, eventually passing away entirely in the form of vapour. Upon gently warming a small quantity of the salt in a test-tube to a temperature very slightly superior to 100, it sublimes like ammonium chloride, condensing again, however, in brilliant little prisms. This operation requires great care, for if the heating proceeds too rapidly the substance explodes with great violence.

As may be expected, great difficulties were met with in obtaining an analysis of a substance so explosive. Upon attempting to determine its composition by combustion with copper oxide in a stream of dry air, the apparatus was destroyed upon each occasion with a fearful detonation. Only one-tenth of a gram of the salt was employed, placed in a small platinum boat. At first the compound sublimed out of the boat into the cooler portion of the combustion tube; the little sublimed crystals then commenced to fuse into yellow drops, and immediately this occurred, in each experiment, the tube was shattered to fragments with a frightful report. The platinum boat was in each case torn to fine splinters. Eventually, however, Prof. Curtius succeeded in obtaining a satisfactory analysis by performing the combustion with copper oxide in a stream of carbon dioxide

The ammonium salt may be readily converted into the sodium salt by evaporation with caustic soda upon a water-bath.

An aqueous solution of azoimide may be prepared by distilling any of its salts, preferably the sodium or silver salts, with dilute sulphuric acid. It is more conveniently obtained, however, by dissolving the crystals of diazohippuramide in dilute caustic soda, warming the solution for a short time upon a water-bath, so as to insure the formation of the sodium salt, and subsequently distilling the liquid with dilute sulphuric acid. The latter is allowed to drop slowly from a dropping funnel upon the soda solution contained in a flask and maintained at the temperature of ebullition. The flask is connected with a condenser, and the azoimide, as it escapes, is carried along with the steam, and condenses in the receiver in the form of an aqueous solution. This solution may be concentrated by precipitating it with silver nitrate, collecting the insoluble silver salt, and distilling it with sulphuric acid diluted with eight times its volume of water. The aqueous solution of azoimide possesses a most intolerable odour.

Free azoimide itself may be obtained by the fractional distillation of the concentrated aqueous solution thus prepared. The first fraction is collected separately and refractioned. Upon repeating this process with four successive first fractions, an acid containing over 90 per cent. of azoimide is obtained, which distils at about 45. The last 10 per cent. of water may be completely removed