gases in war and the massacre of women and children. We would rather suggest that musical art possesses an element of permanence which is

not to be found in the other arts. The collection of masterpieces which the present generation has inherited from the classical composers is so large that a modern composer, working on the same lines, would find it next to impossible to assert his influence. Consequently, musicians have had to seek fresh fields and pastures new by tearing up the "scraps of paper” which their predecessors regarded as binding.

OUR BOOKSHELF.

Surface Tension and Surface Energy and Their Influence on Chemical Phenomena. By Dr. R. S. Willows and E. Hatschek. Pp. viii + 80. (London: J. and A. Churchill, 1915.) Price 2s. 6d. net.

THE appearance of this book, following close on that of a similar work by Michaelis, is a welcome sign of the increasing interest now taken in this subject. The distribution of matter and energy at any interface, although of the first importance, has only come into prominence since the development of colloid chemistry.

The scope and character of the book may be indicated by a reference to the chief subjects discussed. They are:-the fundamental ideas of surface tension; intrinsic pressure; Gibbs's surface excess formula; recent experimental work on interfacial concentrations; electrical phenomena at interfaces, with special attention to the Lippmann electrometer and the dropping electrode; and, in conclusion, such matters as condensation on gas ions, effect of electrification on the vapour pressure of drops, and "waterfall electricity."

It is scarcely surprising that the authors have found it no easy task to co-ordinate these varied subjects; and to this difficulty is doubtless due a want of clearness in a few places. There is one notable omission; one would have expected to find the development of the surface excess equation, either as given by Gibbs himself, or by Thomson or Milner. Curiously, no mention is made of Milner's experimental work, the first attempt to test the formula.

In spite of these defects, the work must prove helpful to advanced students and research workers, biological and technical, who have a practical interest in adsorption and allied phenomena. There is a useful index, but it seems a pity that no references to original papers are given. W. W. T.

Fire Tests with Glass. "Red Books," Nos. 196 and 197. (London: British Fire Prevention Committee.) Price 2s. 6d. each. THESE two small books embody the British Fire Prevention Committee's Report on Fire Tests made respectively with skylight openings and windows filled in with "wired glass" manu

factured in our own country. The skylights were five in number, each 2 ft. square, and arranged The glazing horizontally in a single straight row. was subjected to fire for an hour, followed by water from a steam fire-engine applied at close range for two minutes on the fire side. No fire passed through the glazing, but more or less water found its way through three of the five. Details of the tests are given, with illustrations showing the effects of the fire. The three vertical windows

were subjected to a precisely similar test with much the same results. In the case of two of the windows neither fire nor water passed through the glazing, but in the third, though no fire passed through, the application of water caused perforation and some water got through. temperatures reached before the application of water were not less than 1500° F. (or 8155° C.). The maximum size of the vertical glazing tested was four feet by one foot.

The

The results of the tests clearly indicate that British wired glazing, when suitably fixed, can effectually check the spread of fire in a manner comparable with fire-resisting partitions and doors of much greater thickness and weight. The subject is well worth the serious attention of those interested in the limitation of damage by fire, especially in cases where the admission of light is desirable. J. A. A.

LETTERS TO THE EDITOR. [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 Okapi.

IN my letter published in NATURE of May 27 (vol. xcv., p. 342) dealing mainly with the "Supposed Horn-Sheaths of an Ökapi," I stated that “it is only when extremely young that the backward slope of the back is very noticeable." It is perhaps the only statement I have made regarding the okapi which was not based on my own observations, and it appears to be erroneous. The impression was derived from a photograph reproduced in M. Fraipont's "Monograph on the Okapi," of a very young one captured by natives and brought into one of the Uele stations. I have since seen a photograph of the same animal from another source which shows that there was very little backward slope. At maturity the height of the okapi at the shoulders is only 2 to 3 in. at the most more than that above the hindquarters. The following measurements taken from three animals lying as they fell, one by Mr. A. E. H. Reid, and two by myself, bring this out quite clearly :

Mr. Reid has given me the following additional measurements for his specimen :-Girth of neck in front of shoulders 2 ft. 11 in., and girth of chest behind shoulders 5 ft. 7 in.

The okapi has no "hump" at the withers, as might be supposed by the length of the spinous processes of the dorsal vertebræ and as represented in many mounted specimens. The back is as straight as that of any horse or antelope. The neck is deep at the base and tapers to comparatively small dimensions immediately behind the head without any curve. The body is rotund, and the limbs remarkably clean and sleek. The animal as I have seen it has distinctly graceful proportions, and does not give one the impression of being all angles as in the giraffe. When standing on the alert it holds the neck fairly high, slightly above the line of the back, with the head poised at an angle, and the ears well forward.

The animal is probably a surviving primitive form of a family closely allied to the ancestors of the giraffes, which, as the forest once covering all tropical and probably most of subtropical Africa gradually disappeared by the agency of man, have become modified, bulkier animals under the influence of sunlight and freedom in the open scrub country, where to browse on the young shoots of the thorny acacia trees, grown tall and table-topped owing to annually recurring grass fires, has necessitated the development of a longer neck.

The giraffe, now restricted to Africa, at one time, during the latter portion of the Tertiary period, I think, roamed far and wide over southern Europe and throughout a large extent of Asia. Similarly the okapi's area of distribution before the destruction of the forest was probably much wider than at present. There is, in fact, evidence that it inhabited the once forest-covered regions of the Upper Nile Valley. It was pointed out in 1902 that among the twelfthdynasty paintings from Beni-Hasan, Egypt, there exists a picture, long known to archæologists, which portrays a creature termed "Sche," said to have a resemblance to the present-day okapi, except that the upper lip is somewhat protruded and proboscis-like, and there are no zebra-like markings, the body and limbs being of a uniformly reddish colour. It is possible that this picture represents a form of the okapi known to the ancient Egyptians. C. CHRISTY.

July 5.

Testing Respirators.

THE chief difficulty in the determination of the relative efficiency of these safeguards which are so freely offered to the public, and also utilised for the purposes | of war, is the want of a standard method of testing the same.

As the result of a few preliminary experiments, I would suggest in the testing of different fabrics impregnated with chemical substances that the conditions of working must be standardised, and the following conditions met. A suitable container for the mixture of air and the poisonous gas is required, and this must lead directly to the surface of the fabric which is exposed to the action of the mixed gases as they pass through its substance.

The rate of flow of the gas through the cloth must slightly exceed that of the inflow of air under actual breathing conditions, and this must be standardised. The reduction in the amount of the added gas is observed by actual analysis, and the test only made after the air has been passing for a stated period, which must not be less than five minutes. The composition of the air being tested must correspond with

Bird Migration.

A SMALL item which may be of interest to those watching bird migration was noted on a recent voyage to America.

On s.s. St. Louis, of the American Line, about 6.45 p.m. (ship time) on May 3 a swallow came on board, evidently very tired, white breast feathers rather dirty, and, settling down, was caught by one of the passengers. It took some water but died during the night. There was no identification band on either leg.

The point of interest is that the ship was then about 49° N. lat. and 23° W. long., which would put her some 560 miles west of Cape Clear-about W. by S.and some 680 miles N.W. from Cape Finisterre. The wind had been fairly steady from E.S.E. for the previous thirty hours, blowing 20-30 miles an hour.

The first swallows had been noticed at the place where I now am on Saturday, May 1, and no doubt the bird which reached the St. Louis had got separated from the general migration.

The bird seemed in fairly good condition, its brown throat and indigo head were sleek and glossy; the soiled breast feathers may have been due to the steamer's smoke as the bird came in from the lee side. ED. WILDING.

Dunedin, Jordanstown, Co. Antrim, June 23.

Mercury Ripples showing Interference. THE accompanying print of a photograph of mercury ripples showing interference, made in this laboratory, exhibits singularly well the circular waves from the two sources as well as the interference pattern pro!uced. The two points of disturbance are maintained

by a forked pointer attached to the prong of a fork of frequency 50. With daylight illumination from a window, and a rotating sector to render the effect stroboscopic, a good natural picture of the surface is obtained. S. G. STARLING. Physical Laboratory, Municipal Technical Institute, West Ham, E.

Man's True Thermal Environment.

I FULLY agree with Mr. Grabham (NATURE, June 24, p. 451) as to the unsuitability of the constanttemperature (37° C.) psuchrainometer for many parts of the earth's surface. It is for this very reason, combined with the advantage of its much greater simplicity, that I am experimenting with the constant-energy form of instrument mentioned at the end of my former letter (NATURE, May 6, p. 260). The effect of moisture can be brought into play with any type of psuchrainometer by providing its

exposed surface with a water-wetted muslin cover, and no doubt in this condition the apparatus approximates more closely to the human body.

My interest in the matter, however, is physical rather than physiological. My immediate aim is to study the extent to which "atmospheric cooling" can be predicted from the readings of the existing meteorological instruments. It seems best, therefore, to begin with the simplest case of cooling, namely, that which is free from the thermal complications accompanying evaporation. JAMES ROBERT MILNE. Physical Laboratory, University of Edinburgh, July 1.

HIGH EXPLOSIVES.

MR. LLOYD GEORGE'S recent speech in

the House of Commons, as Minister of Munitions, emphasised the very important part played by high explosives in the present war. It is essential at the outset to distinguish clearly between a propellent charge, which forces the projectile or shell through the bore of the gun, and the high explosive charge filling the shell itself and causing it to burst, through the intervention of a time or percussion fuse. Modern military propellants consist of gelatinised guncotton (nitro-cellulose), either alone or mixed with varying proportions of nitro-glycerine, pressed into any required shape. The finished explosive is of a colloidal, horny nature, and a piece of it held in the fingers, whilst burning, can be blown out quite easily. A charge lit in the enclosed space of the chamber of a gun can discharge a projectile with a velocity of about 1000 yards per second, developing in the chamber a pressure of, perhaps, twenty tons on the square inch. If the same quantity of the ungelatinised material were ignited in the gun-chamber it would detonate and blow the gun to pieces.

There is thus a wide difference between the effects produced by the burning of a propellant in the open, and in the chamber of the gun. In the latter case, before the projectile begins to move, the gases evolved produce pressure in the chamber, thus greatly accelerating the velocity of the explosive reaction.

The forces at work in the gun, however, are insignificant in comparison with those brought into play when a high explosive detonates. Even in the open, without any containing envelope other than a thin cylinder of paper, the writer has obtained with a high explosive a velocity of detonation of the explosion wave of some seven miles per second. When the high explosive is in an enclosed space, such as a shell, the velocity of the detonation wave is greatly accelerated, and in an almost infinitesimal period of time the explosive is converted into gases. The volume of gases produced varies according to the nature of the explosive, but, generally, for those used in shells, it may be taken that, at the temperature of explosion, the volume of gas evolved occupies from 15,000 to 20,000 times the volume of the original explosive. This is the reason for the enormous destructive and shattering effect of a high explosive.

The nitro-glycerine high explosives used in mining are unsuitable for shell filling, owing to the sensitiveness of nitro-glycerine to shock, which would cause premature detonation in the bore of the gun. The ammonium nitrate group of high explosives, also used in mining, which contain nitro-hydrocarbons, and in some instances aluminium, have been advantageously adapted for shells. Although the hygroscopic nature of ammonium nitrate is detrimental, this may be successfully overcome.

Abel, in 1865, first proposed the use in mines of compressed wet gun-cotton fired by means of a dry gun-cotton primer; this was later used for filling torpedoes, but it has the disadvantage of low charge density. Gun-cotton cannot be compressed to a greater density than 1'25. In other words, a torpedo head which would hold 125 lb. of compressed gun-cotton could hold from 160 to 180 lb. of the denser trinitrotoluene or picric acid, with a corresponding increase of destructive power.

Sprengel, in 1875, first showed that picric acid could be detonated, and in 1881, Turpin, in France, demonstrated the practical possibility of using it for filling shells. The idea was rapidly taken up by other countries.

The methods of manufacturing nitro-hydrocarbons suitable for shell filling are very similar to those in use for producing nitro-glycerine. A mixture of sulphuric acid and nitric acid is used, and large quantities, very frequently as much as a ton or even more, are made in one operation. To obtain the highest yields of pure products very great attention must be paid to the composition of the acids, to the efficiency of agitation, and to the temperature, which is regulated by internal heating or cooling coils.

Picric acid, discovered in 1771 by Woulfe, of London, when used for filling shells has a different name in each country. It is called Mélinite in France, Lyddite in England, Pertite in Italy, Shimose powder in Japan, Granatfüllung 88 in Germany, and Ecrasite in Austria. It is

toluene (obtained from coal-tar naphtha), and was first proposed for use in shells by Haüssermann, in 1891. The result of the first nitration of toluene is a mixture of mono-nitro-toluenes, which are then treated with a mixture of strong nitric acid and sulphuric acid, and raised to the third or trinitro degree of nitration in one stage. Trinitrotoluene, when pure, forms brownish-yellow crystals with a melting point of 81° C. It is very stable, not igniting below 300°, but when it explodes it does so with great violence. Its density when melted varies between 1'57 and 160. Neither picric acid nor trinitrotoluene can be detonated with certainty by fulminate of mercury, and a small quantity of an intermediate priming charge is employed. In the case of trinitrotoluene, the use of tetranitromethylaniline has been found suitable. Tetranitraniline is a very powerful explosive, and has a higher density than either trinitrotoluene or picric acid.

The nitro-hydrocarbon high explosives used for the shell bursting are, in the molten state, poured into the cavity of the shell, in which they solidify, sufficient room being left for the priming charge and the detonating fuse. All the above nitrocompounds can only be obtained in a state of purity by re-crystallisation from various solvents.

Fulminate of mercury has been mentioned several times as a detonator, or initiator of explosion. Lead azide has been used, in conjunction with fulminate of mercury and tetranitromethylaniline, as a detonator for high explosives. It will be noticed that the majority of the high explosives referred to are derived from coal-tar products, and it is therefore evident that Mr. Lloyd George's statement, "If there were a shortage in the coal supply for any reason, the consequences would be very calamitous," is one which must be taken very seriously.

GEORGE W. MACDONALD.

SCIENCE IN THE SERVICE OF THE STATE.

taste.

occasional addition of crésylite (trinitrocresol) or a salt of that substance, the object of which is to reduce the temperature of fusion. It is manufactured from phenol (carbolic acid, obtained from the distillation of coal-tar) by first dissolving in sulphuric acid and then treating the resulting phenol-sulphonic acid with nitric acid in excess. It forms yellow crystals with an intensely bitter It has a specific gravity of 1'777 and melts at 122.5° C. Picric acid, if heated gradually, takes fire without explosion, giving rise to dense black smoke, but the application of a red-hot rod will cause it to detonate, as will also the explosion of a capsule of fulminate of mercury. Owing to the readiness with which it forms certain unstable metallic salts, the use of picric acid is not free from danger, and it is largely on this account that it is being rapidly replaced by the somewhat less energetic but much safer trinitrotoluene.

Trinol, trotyl, trilite, tritolo, or T.N.T., as trinitrotoluene is variously called, is made from

military, engineering, chemical, medical, and economic. and economic. Its successful prosecution requires more than an extremely high efficiency on the part of the officers in their professional work. Everything that chemical, physical, and engineering science can suggest must be pressed into service. The scientific men of the country have been keenly aware of this necessity from the beginning of the war, and many of them have individually done a great deal of important work for the Government and the various Services. The Royal Society has formed a War Committee, to which the Government has confided the solution of many pressing scientific problems arising out of the war. The public thanks of the country have been given to the Royal Society by Mr. Asquith. We note also with pleasure the issue, by the Council of the Chemical Society, of the letter (see p. 523) announcing that the Council has constituted itself a consultative body to consider, organise, and utilise all suggestions and inventions which may be communicated to it.