able to support them and their progeny without great difficulty. Now a little consideration will show that the longer the life and the slower the reproduction of the trees, the greater will be the contrast. If the plant infested by the borers had been an annual herb, it might have contrived to perfect its seeds, and the death of the old stem would be but a natural and inevitable process, and fresh plants might have been produced in sufficient numbers to continue the species in spite of all insectattacks. But in the case of trees-oak-trees especially, the rate of growth and reproduction is such that, unless the insect-borers can live in galls, they will destroy the plants entirely, and themselves in consequence. Indeed, I have no doubt, that if all the gall-makers now existing could suddenly be transformed into stem-borers, the genera Quercus, Rosa, and Salix, now so dominant, would shortly disappear from off the face of the earth. The other hypothesis-here assuming that the production of galls is due more to the tree than the insect-is this. Suppose an oak tree with four branches, all attacked by internal-feeding insects. Two of the branches produce swellings in which the insects live, while the other two produce none, and the insects have to devour the vital parts. Now the two branches which produced no swellings would quickly be killed by the insects, but those which produced galls would live, and the more perfect the galls, the greater the insect-population they would be able to support. Hence the tree would finally, by the survival of its gall-producing branches, become purely gall-producing, and we may assume that its progeny would inherit the peculiarity.

I am aware that the above arguments will sound a little like those of the Irishman, who said he ought not to be hanged, because, in the first place, he did not kill the man; in the second place, he killed him by accident; and thirdly, he killed him in self-defence,"--but I do not represent either of the above hypotheses as the precise truth of the matter, and I think they sufficiently illustrate the principles involved.

T. D. A. COCKERELL.

West Cliff, Custer Co., Colorado, March 16.

On the Use of the Edison Phonograph in the Preservation of the Languages of the American Indians. THE present state of perfection of the Edison phonograph led me to attempt some experiments with it on our New England Indians, as a means of preserving languages which are rapidly becoming extinct. I accordingly made a visit to Calais, Maine, and was able, through the kindness of Mrs. W. Wallace Brown, to take upon the phonograph a collection of records illustrating the language, folk-lore, songs, and counting-out rhymes of the Passamaquoddy Indians. My experiments met with complete success, and I was able not only to take the records, but also to take them so well that the Indians themselves recognized the voices of other members of the tribe who had spoken the day before.

One of the most interesting records which was made was the song of the snake dance, sung by Noel Josephs, who is recognized by the Passamaquoddies as the best acquainted of all with this song "of old time." He is always the leader in the dance, and sang it in the same way as at its last celebration.

I also took upon the same wax cylinder on which the impressions are made his account of the dance, including the invitation which precedes the ceremony.

66

In some

In addition to the song of the snake dance I obtained on the phonograph an interesting "trade song," and a Mohawk war song" which is very old. Several other songs were recorded. Many very interesting old folk-tales were also taken. of these there occur ancient songs with archaic words, imitation of the voices of animals, old and young. An ordinary conversation between two Indians, and a counting-out rhyme, are among the records made.

I found the schedules of the United States Bureau of Ethnology of great value in my work, and adopted the method of giving Passamaquoddy and English words consecutively on the cylinders.

The records were all numbered, and the announcement of the subject made on each in English. Some of the stories filled several cylinders, but there was little difficulty in making the changes necessary to pass from one to the other, and the Indians, after some practice, were able to "make good records" in the instrument. Thirty-six cylinders were taken in all. One apiece is sufficient for most of the songs and for many of the short stories. The longest story taken was a folk-tale, which occupies

nine cylinders, about "Podump" and "Pook-jin Squiss," the "Black Cat and the Toad Woman," which has never been published. In a detailed report of my work with the phonograph in preserving the Passamaquoddy language, I hope to give a translation of this interesting story. Boston, U.S. A., March 20. J. WALTER FEWKES.

Solar Halos and Parhelia.

A MAGNIFICENT display of solar halos and parhelia was witnessed here this afternoon, exceeding in beauty and brilliancy that observed on January 29, 1890, and described in NATURE, February 6, p. 330.

The phenomenon was similar to the one of January 29, except that the mock suns were distinctly outside the first circle or halo, at a distance of 5 or 6°, and were when first seen at 3 p.m. above the level of the true sun; a handkerchief stretched at arm's length from one to the other gave the blurred image of the sun several degrees lower.

At 3.49 the patch of white light appeared about 90° from the right mock sun and connected to it with a curved band of white light, concave side upwards. The right mock sua must ther have been below the level of the sun, as the band appeared to pass upwards through it to the sun. This band only remained a few minutes; the right sun and zenith arc at the time were most intensely brilliant, with the colours exceptionally clear and vivid. The zenith arc, and the patch of white light, were the last to disappear at 4.22.

The cirro-stratus cloud during and after the display was rapidly advancing from the north. Driffield, April 9.

Cambridge Anthropometry.

J. LOVELL

I HAVE read with much interest, in NATURE of March 13 (p. 450), Mr. Venn's very interesting article on anthropometry at Cambridge.

There is in his tables one rather peculiar feature, of which I find no notice taken in the text. It will be seen on reference to the tables that, while the other physical characteristics increase from A to B, and from B to C (weight and height being irregular, however), the breath is highest in A, less in B, and least in C. thus falling with the intellectual fall.

It is true that the difference in this as in most of the other characteristics is so slight as to be-as Mr. Venn saysfall instead of rising with the other physical characteristics practically negligible; but still the fact that this should steadily strikes me as peculiar. I should be glad therefore to hear if Mr. Venn has any comment to make on this phenomenon, or any explanation thereof to suggest. F. H. P. C. April 4.

A Remarkable Meteor.

ON Thursday, April 10, at 10.40 p.m., I observed a meteor of extraordinary brilliancy shoot from a point just east of B Leonis. It travelled over about 10° in a north-westerly directiva, and was visible for fully two seconds. Its apparent diameter, as nearly as I can judge, was about a quarter of that of the full moon; its colour, a very vivid pale green. J. Duxx. Much Marcle, Herefordshire, April 11.

vania.

Earthworms from Pennsylvania. NEARLY twenty years ago, a very aberrant earthworm was described by a French naturalist, who obtained it from Pennsyl I should be greatly indebted to any naturalists or travellers who may find themselves in that part of the United States, if they would collect some of these worms and send them to me The most convenient mode of transmission would be to pac the living worms in moist earth with moss or grass, in a tin bos perforated at one end: this should be inclosed in a wooden box Both small and large worms should be collected: some might be preserved in strong spirit, but living specimens would be the most useful. W. BLAXLAND BENHAM. University College, London, April 10.

Crystals of Lime.

SINCE the appearance of my letter on this subject (p. 515) f have found that similar crystals have been recently observed by Mr. J. Joly, and were described by him in the Proceedings of the Royal Dublin Society, vol. vi. p. 255. H. A. MIERS.

might a kilogramme be defined for a French butcher as the weight of a cubic decimetre of distilled water at 4° C., and the butcher's business be absolutely stopped because he did not possess any distilled water and because the temperature of his shop was 20° and not 4° C. In fact, Lectures II. and III., although containing a large amount of valuable information, are professorial rather than practical.

On p. 74 a Ruhmkorff induction coil is correctly described, but in Fig. 36 on the same page the primary coil, with the vibrating interrupter and four cells in its circuit, is shown as consisting of many convolutions of fine wire, and the secondary of a few turns of thick wire. On p. 83 one centimetre is given as equal to o0328087 of a foot-that is, correct to six significant figures-while even in the second edition, "the call" for which "has afforded the opportunity to erase several typographical errors and to remove some other blemishes which had escaped notice and correction in the first edition," the previous statement is immediately followed by the announcement that one inch equals 2.500 centimetres, an equation which is only correct to two significant figures, the number expressed correctly to six significant figures being 2:53995. But why not use 2'5400, the value commonly adopted, and which is correct to four places of decimals? As a further example of the want of precision which runs through this book, it may be mentioned that on p. 9 a falling body acquires per second a velocity of 981 centimetres per second. Throughout the whole of p. 85, where the number is frequently mentioned, the body, as if a little tired cannot get up a velocity of more than 980 centimetres a' second. Proceeding, however, to the next page, the body, like the reader, turns over a new leaf, and hurries up its speed, for it acquires per second a velocity of 981 centimetres per second all through this page. Further on, however, in the book, the poor falling body gets tired again, for on p. 97 it cannot do more than the 980. On p. 87 we find the statement, "Hence one foot-pound = 1356 joules, or one joule 7373 foot-pound," whereas a simple division shows that if the first part of the statement be correct, the second is not.

To say that "the work is numerically measured by the product of the displacement and the mean stress estimated in the direction of the displacement" is learned and academical, but might not the poor electrical artisan mix this up with the displacement of the factory hands that usually occurs when there is no stress of work?

On p. 99 it is stated that the "E.M.F. of Clark's cell 1435 true volt," but, as no indication has been given in this book that there is more than one volt, we are left in ignorance of the reason why the volts used to measure the E.M.F. of a Clark's cell have to be so especially true, and why 10o C.G.S. units, which is the volt that has been previously used, is not good enough for this sort of measurement. On looking in the index for the definition of the "Ohm British Asssociation," we find ourselves referred to p. 136, and the reader is left to wonder what is a "B.A.U." of resistance used some forty pages previous to this. Similarly the "Legal Ohm" is spoken of and its value given in terms of a "B.A.U." thirty-seven pages before the reader is told what a 66 Legal Ohm" is. For this the arrangement of the book and not the index is, of course, to blame. And while on this subject we should like to point out that the indexes of scientific books appear to furnish a conclusive proof of the inherent modesty of scientific writers. Take up some large and important treatise, and turn to the index. There you are told that the book contains almost nothing. On the title-page the publisher may have indiscreetly added after the author's name line after line of small print enumerating the various scientific and unscientific societies to which the author belongs, but in the index all pretension to such a wide acquaintance with science is disclaimed. You may have a distinct recollection of reading in this very book many

pages on some special subject, but rack your brains as you will to discover under what heading in the index this subject may have been entered, not a reference to it can you find. Accumulators, storage cells, transformers, the volt, voltmeters, &c., seemed likely subjects to be treated on in "Short Lectures to Electrical Artisans," but the index says no; and it is only by carefully reading through the book that you discover that it contains much valuable information on these very points. We would suggest to the writers of scientific treatises, and also to those who communicate scientific papers to leamed societies, that the practical man of to-day cannot possibly afford the time to read through ninety-nine things tha he does not want to know about, before he can light on the one thing regarding which he is searching for information.

In speaking of Messrs. Crompton and Kapp's meter, on p. 115, Dr. Fleming says:

"The only difficulty which arises in connection with such an instrument as this, is the tendency of a long thin iron wire of this kind to retain strongly residual magnetism and fail to de-magnetize itself, but this effect would only prevent the return of the indicating needle to zero when the current was stopped, but would not prevent the instrument from giving a definite and fixed deflection corresponding to a definite and fixed current passing through the coils." It was no doubt a somewhat delicate task for Dr. Fleming when lecturing to Mr. Crompton's staff to fully criticize Mr. Crompton's meters, but since actual published experiments on some of these meters show that, for the low readings, the apparent value of a given current differs by as much as 10 per cent., depend. ing on whether the current is ascending or descending, we fail to see how the scientific knowledge of any artisans can be improved by their being told that no such error exists

Fig. 50, p. 122, showing the level of the columns of water in stand-pipes attached to a horizontal tube through which water is flowing, was never drawn from an actual apparatus. The author has forgotten that the water has not merely to flow through the horizontal tube Aa, but through the much longer vertical tube CA, and therefore, there is a much greater difference of level between the height of the water in the cistern and in the first stand-pipe, aa', than there is between the level in this stand-pipe and in the next, bb'. If Fig. 50 were correct, it would follow that when a battery of even large internal resistance was sending a considerable current the difference of potentials at its terminals was equal to the E.M.F. of the battery. Not merely, then, is this opportunity lost of explaining to the readers that the difference of potentials at the terminals of a battery mar be very much less than the E.M.F., but the information conveyed by the diagram is actually contrary to fact.

The statement that "Storage cells for lighting purposes cease to give a useful discharge when the electromotive force falls below two volts" is hardly consistent with the fact that, when storage cells are discharged at the current that is considered quite safe by the Electrical Storage Power Company, the E.M.F. for nine-tenths of the period of the discharge is slightly below two volts.

We have said enough to show that, although the book called "Short Lectures to Electrical Artisans" is written by one who, from his University and factory experience, has a large amount of valuable information at his command, the second edition reads far too much like an uncorrected proof of the first edition; and instead of the statements it contains possessing weight because they are made in the book, there is an uneasy feeling when reading its pages that any statement may be wrong, and requires to be checked. We trust, however, that the sale of this, the second edition, may be large and rapid, so that the author may have an opportunity of shortly bringing out as a third edition a book more worthy of his acknowledgol

power.

"Absolute Measurements in Electricity and Magnetism," by Prof. A. Gray, is a most interesting book to read. It opens with a detailed description of Gauss's methods for determining the horizontal intensity of the earth's magnetism, and with an account of the results of the measurement of the variation, produced by a unit field, on the magnetic moments of steel magnets of different sizes tempered to different degrees of hardness. If it be desired to determine the magnetic moment of a barmagnet as well as the horizontal intensity of the earth's magnetism, which is of course necessary when variations of the magnetic moment of a bar are in question, Gauss's methods are admirable. But if the value of H is all that is needed, then the simpler method of employing an earth inductor with a ballistic galvanometer, which is described on pp. 317-21, might well be employed. It. would, therefore, have been well to give a reference to this method in the first two chapters, which are mainly devoted to the determination of H.

Next follows a concise statement of the various ways of defining the absolute current, and a fairly complete chapter on standard galvanometers. In Chapters IV. and V., and in Chapter XI., to which reference is made, there is given the ablest description of the dimensions of the electric and magnetic units that we have ever read. It is both correct and comprehensible, which is saying a very great deal for an exposition of a subject which, as usually explained, generally leaves even a thoughtful student semi-dazed as to whether the dimensions are the dimensions of the unit, or the dimensions of a quantity measured in the unit. Indeed, the early reports of the Electrical Standards Committee of the British Association were actually wrong on the very subject of dimensions, so that "7" was regularly defined as the ratio of the electrostatic to the electromagnetic unit of quantity instead of as the reciprocal of that expression.

The volt, ohm, ampere, coulomb, watt, and joule are also explained and defined in Chapter V., and Prof. Gray gives Sir W. Thomson's expression "activity" for the rate of doing work. He does not mention, however, that the equally short word " power" is regularly employed with this signification.

Chapter VI. is devoted to the laws of the currents sent by galvanic cells through single and parallel circuits, and through any branch of a network like that of the Wheatstone's bridge. A neat proof is given of the arrangement of a given number of cells that sends the greatest current through a fixed resistance, and the reader is very properly warned against confusing the arrangement which develops maximum power with the most economical arrangement. In Chapter VII. we have a complete description of Sir William Thomson's meters, but, as the book is a scientific treatise (in fact, a very good scientific treatise) and not an instrument-maker's catalogue, we think that the author would have done himself more justice had he described, in addition, some of the other many forms of electric meters in common use at the present day for carrying out the same measurements. Further, in view of the large experience that the author of this book has probably had with Sir W. Thomson's meters, it would have been well had there been a description not merely of the advantages of these instruments, but also of their disadvantages, a subject no one would be more willing to discuss than the inventor himself. On pp. 133-35 is given a very simple proof of the ordinary formula for the quadrant electrometer, but the reader is not here warned that the formula may give an answer many per cent. wrong in practice. On p. 302 it is stated that this formula may be slightly wrong if the aluminium needle of the electrometer be not accurately adjusted relatively to the quadrants, but this, we fear, is rather misleading, since it is further stated that "if the needle hangs at its proper level, and is otherwise properly adjusted, and the quadrants are close, the equation may be taken as accurate enough for practical

purposes," a conclusion regarding which we understand there is grave doubt. In this chapter the very important subject of calibrating instruments by the use of the silver" or the copper voltameters is fully entered into. The large amount of valuable work done on this subject by the author's brother, Prof. T. Gray, of which a description is given, endows this chapter with an authoritative character. Chapter VIII. commences with the construction and use of the various forms of Wheatstone's bridges, the description of the modes of using them, and hints as to the care of a resistance box. The methods for calibrating. relatively and absolutely the wire of a bridge devised by Matthiessen and Hockin, Foster, T. Gray, and D. M. Lewis are discussed at length, and specimens given of the actual results obtained at University College, North Wales, by the use of these methods. The ingenious bridges, which have been arranged by Sir W. Thomson, Matthiessen and Hockin, Tait and T. Gray, for measuring very low resistances, are fully entered into, and the construction of standard coils, the measurement of high resistances, and of the resistance of a battery finish a chapter of especial interest. The method of measuring the resistance of a battery, proposed several years ago by Sir Henry Mance, is condemned by Prof. Gray as being so troublesome as to be practically useless," on account of "the variation of the effective electromotive force of the cell produced by alteration of the current through the cell which takes place when the key is depressed." We think that it should have been stated that this is not a defect especially of Mance's method, but of all methods for measuring the resistance of a battery based on the alteration of a steady current by the alteration of the resistance in the battery circuit. Would it not also here have been well to describe and discuss the condenser method of measuring a battery resistance, as it is the one to which the fewest objections can be raised?

Good as are all the chapters in this book, the next one, Chapter IX., on "The Measurement of Energy in Electric Circuits," is so good that it takes the palm. It commences with the practical methods of measuring the power and efficiency of motors and secondary batteries; the construction and employment of activity meters (wattmeters); and then discusses very fully the laws of alternate currents, the mathematical theory of alternate current generators singly, or coupled in parallel or in series; the theory of the action of an alternate current generator supplying current to an alternate current motor; the true method of measuring the power given to any circuit by an alternate current; and the error produced when an ordinary watt-meter is employed. The work of Joubert, Hopkinson, Potier, Ayrton and Perry, and Mordey on this subject is summed up in a masterly fashion. Chapter IX. is, in fact, the most complete exposition of many problems connected with the all-important subject-the electrical transmission of energy by alternate currentsthat is to be found in any existing text-book, and especially in a small octavo text-book, that can be easily carried in one's coat pocket.

In Chapter X. the measurement of intense magnetic fields is dealt with, and a description is given of ingenious methods proposed by Sir W. Thomson for measuring the force on a conductor conveying a known current placed in the magnetic field, and so determining the strength of the field. The ordinary method of ascertaining the strength of a magnetic field by suddenly withdrawing a coil, of known area and number of convolutions, attached to a ballistic galvanometer, is described. But in order to ascertain the constant of the ballistic galvanometer, the author only gives the old method of observing the swing of the needle when a large coil is turned in the earth's field, a method which necessarily requires for its employment a previous knowledge of the strength of the earth's field at the place. A far simpler method of ascertaining the constant of a ballistic galvanometer is to charge a