that had been called virulent were now more often called parasitic; and it may justly be said that Pasteur's researches were the efficient beginning of the vast science of bacteriology-vast alike in natural history and in pathology and in its intimate relations with organic chemistry.

Besides ascertaining the micro-organisms appropriate to several diseases, Pasteur, still working on the lines which he had followed in his studies of fermentation and of the diseases of beer and wine, found various means of "cultivating" the germs, separating them, multiplying them, and then testing their different influences on different animals, or on the same animals in different conditions, or after various changes induced in themselves. Among these changes, the most important and most fruitful in its further study were the various means of "attenuation" by which the virulence of disease-producing micro-organisms can gradually be so diminished that at last they can, without harm, be inoculated or injected into an animal which they would have rapidly killed if similarly inserted in their natural state. And some of these injections were shown to be better than harmless, for, by conveying the disease in a very mild form, they rendered the animal, for some considerable time, insusceptible of that same disease in a more severe form; they conferred an immunity similar to that given by mild attacks of the contagious fevers which, as it is commonly and often truly said, "can be had only once." Or, as Pasteur held, the inoculation with the attenuated virus was similar to vaccination, which gives protection from small-pox by producing similar disease in a milder form. Hence began the practice of "protective inoculation" for many diseases besides small-pox.

In studying the methods of attenuation, Pasteur found many facts which are not only valuable in bacteriology, but are likely to help to the knowledge of important principles in general pathology. To cite only a few examples he found marked differences among the micro-organisms of different ferments in their degrees of dependence on air. The great majority need oxygen for the maintenance of life; but unlike these, which he named aërobic, were some anaerobic, the first examples ever known of organisms capable of living without oxygen. He showed that the bacilli of anthrax, being aerobic, soon perish and disappear in the blood of the animals that have died of the diseases due to them, and that in the same blood the anaerobic septic bacilli needing no oxygen now appear and multiply. In anthrax, also, he showed that the attenuation may best be attained by keeping the cultivated bacilli at a high temperature (about 42° C.) for a certain number of days, regulated according to daily tests of the reduction of their virulence. In the end they become incapable of killing even mice, and are protective for sheep and cattle, and other animals, which in their natural intensity they would rarely fail to kill. In chicken-cholera, the disease for which the first experiments in protective inoculation were made, he showed that the due attenuation can be obtained by a series of successive cultivations of the micro-organisms in pure air, provided that intervals of several days or weeks are allowed between each two of the cultivations in the series. In experiments on the transmission of the virus of a disease of one species through

a succession of animals of another species, he showed that the virulence of the bacilli of swine-erysipelas was increased by transmission through pigeons, but diminished by transmission through rabbits. And, as to the varying susceptibility of the same animal under different conditions, a fact so commonly observed in man, he showed that chickens, which are ordinarily insusceptible of anthrax, could be made susceptible by lowering their temperature. They became again susceptible when their natural temperature was restored; and when apparently dying of anthrax in the cold, they recovered if warmed.

It was a step far beyond what had been obtained by protective inoculations when Pasteur invented and proved the utility of his treatment of rabies. Here he proved that when a virus has been inoculated or in any way so inserted that it may justly be deemed sure to destroy life, this result may, at least in the case of rabies, be prevented by a daily or otherwise gradual series of inoculations, beginning with the same virus very attenuated, and diminishing the degree of attenuation till it is used in such intensity as, without the previous graduated inoculations, would certainly have been fatal. The results of the treatment of rabies on this principle are well known; its success is certain, and is enough to justify the hope that by similar treatment, whether with virus simply attenuated, or with some "lymph" derived from a cultivated virus, or from the chemical products of its action on the liquids in which it has grown, other specific diseases may be similarly controlled. This is especially probable for those in which, as in rabies, there is a clear interval between the entrance of the virus and the first outbreak of the disease; and it is becoming very probable that tuberculosis will be one of these. But it would be useless to imagine the probabilities of what will now follow from the researches that have already followed the discoveries of Pasteur.

It hardly need be said that this summary of Pasteur's life and works, and of the chief results to which they have led, can give no fair estimate of the number and the variety of his experiments and observations. Only a complete personal study of his published works, and especially of those in the Comptes rendus de l'Académie des Sciences, can give this. Yet even a mere summary may indicate the most notable points that may be studied in his scientific character: of his charming personal character there is no need to speak here. Clearly, he had a native fitness and love for the study of natural science, and these were well educated, and have been manifest in his whole life. But with this loving devotion to science, he has shown not only a very rare power both of thinking and of observing, but that spirit of enterprise which stirs to constant activity in the search after truth, especially by way of experiment. With the power of accurately thinking what is likely to be true, he shows a happily adjusted ingenuity in the invention of experiments for tests of thoughts, and the habit of doubting the value of any scientific thought, even of his own, which does not bear experimental tests. Especially, the thoughts of what may be true in biology seem to have been always submitted, if possible, to tests as strict as those that may be used in chemistry and physics; and they appear to have been repeated and varied with ad

also have been disturbed by volcanic eruptions. These have occurred before, during, and since, the epoch of the greatest extension of Lake Bonneville. Craters of scoria, as well as flows of lava, remain as monuments-the former occasionally three or four hundred yards in diameter, the latter sometimes a little more than three miles long; the materials are all basaltic. Rhyolitic lavas also occur in the region, but these are long anterior in date to the epoch of the lake. Organic remains are not common in the marls and other deposits of the old lake-bed. This is not surprising in the later period of its history, but they might have been expected in greater abundance in that when the waters were still fresh. The earliest deposits do not carry us back beyond the Pleistocene age, so that, geologically speaking, both the formation and desiccation of the lake are modern events.

This is a bare outline of the last pages of the story of a remarkable district in America, the like of which can be found in more than one other locality on the earth, though, perhaps, in none of them is the record so clearly preserved. Mr. Gilbert's memoir is not only a most careful description and full discussion of the various phenomena presented by this singular dried-up region, but also he turns, not seldom, to questions of wider import, on which, however, want of space forbids us to touch. Moreover, the second chapter of the volume is occupied by a very full discussion of the various topographical features of lake shores. Perhaps in this the author errs occasionally on the side of prolixity, but he brings together so much valuable information that the book will be indispensable to all who wish to study the history and phenomena of lakes and inland seas. We lay it down with a deep sense of gratitude to him for the loving labour which he has evidently bestowed upon this memoir, and will only add that, high as the standard already attained by the publications of the American Geological Survey may be, this monograph, especially in the work of the printer, and in the number, interest, and excellence of its illustrations, more than attains to it. T. G. BONNEY.

ON DUCKS AND AUKS.

On the Morphology of the Duck and Auk Tribes. By W. Kitchen Parker, F.R.S. With Nine Plates. (London: Williams and Norgate, 1890.)

WHE

7HEN the grave, a few months ago, closed over the remains of W. Kitchen Parker, there still remained with those who knew him the memory of an excellent man and brilliant anatomist. None save one devoted to the science could have worked on as he

did, often amid many cares and troubles, feeling a high delight in his work, and considering the attainment of knowledge to be its own reward. From a very early period of his life all his spare moments were devoted to anatomical research, and the hour of death overtook him while as yet, though full of years, he was labouring

still.

One of his latest, if not his last work, lies before us. It treats of the morphology of the Anatidae and the Alcidæ, and has been published by the Royal Irish Academy, as one of its Cunningham Memoirs." It may be necessary to add that these memoirs are

published from the resources of a fund left to the Irish Academy by a Mr. Cunningham, and that great care is taken that the memoirs published therein shall be of a high order of merit.

Most of the materials for this memoir had been in the late Dr. W. K. Parker's hands for many years, but those which he needed to complete his investigations had been only obtained during the last few years from several friends. In a brief introduction he states that for many years after the publication of his first two papers on the osteology of birds (1850-60), his attention was directed solely to the skull. The burden of the anatomy of the skull was placed upon his willing shoulders by Prof Huxley, who then by degrees tempted him into the investigation of the organs of support, which have proved to be of as great interest and profit as the anatomy of the skull itself.

The two families of birds whose morphology is treated of in this memoir are very distantly related, and the true position and genealogy of the duck tribe present as tough a problem as those of the auk tribe; indeed, herein is enough, Mr. Parker writes, to task the ingenuity and strength of two or three generations of biologists. The cranium in Cygnus and its vertebral column are described from an early stage; the wings of Cygnus and Anas and the hip-girdle of Anas in various stages of development are noticed. Among the auks, Uria troile has been selected for description.

In a summary of nearly seven pages, it is pointed out that the Anatidæ manifestly converge towards the Galliaaceous group; that they have the Struthious division of the Ratitæ obliquely below them; whilst the Alcidæ are related to a large and varied group of existing families, but in their ancestry belong somewhere between those two extremely dissimilar extinct families, the Hesperornithida and the Ichthyornithidae. The revelations made by the precious remains of those two toothed types throw a bright light on one side of these questions of origin and relationship, but intensify the darkness of the other side.

Though both groups are adapted for an aquatic life, they are very sharply defined from each other. The ducks are more or less terrestrial, but are also swimmers ; while the auks are not adapted for a land life, but are at home and at ease in the denser or rarer medium-they can dive and fly.

From the ontological standpoint, it will be conceded that that which has dominated the whole bird form is the wing, and embryology shows that this is merely the modified fore-paddle of a low gill-breathing amphibiana nailless fore-paw. But the nails or claws do appear; yet, in the wing, they are out of place; and this reptilian stage is only transient. If the bird is, indeed, the child of the reptile, it must forget its father's house; it must proceed beyond its progenitor. But if we are willing to see the bird's wing grow, not out of a perfect and typical cheiropterygium, but out of an ichthyopterygium in an unsettled state, ready for transformation into the higher type of limb, then the difficulty is solved. It was a fish paddle; it was not to become a fore-foot; it did change into the framework of a bird's wing; in that respect it is a perfect thing; as a paw, it is an abortion. But an organism moves together in all its parts, if it moves at

all; and thus we see that, in correlation to the profoundly modified fore-limb, every other part of this feathered creature has suffered changes.

The whole memoir is devoted to a detailed account of the changes which are thus brought about during this beautiful metamorphosis, the interest of which is increased by the peculiarly fascinating manner of their description, and to which in a brief notice it would be impossible to do proper justice.

OUR BOOK SHELF.

A Dictionary of Metric and other Useful Measures. By Latimer Clark. (London: E. and F. N. Spon, 1891.) THIS dictionary will be found to be a most valuable and useful vade-mecum by all those who have occasion to employ metric and other physical measures. The arrangement of the tables in this form is the most convenient that could have been adopted, and for uniformity and facility of reference could hardly be excelled. One great feature, which is generally lacking in ordinary sets of tables, is the setting forth of the relations of the different metric units to each other: thus, for instance, on looking under the heading gramme-centimetre, we find its equivalent in kilogramme-metres, foot-grains, foot-pounds, joules, ergs, &c., while the latter are indexed under their respective titles. Not only have the French measures with their factors for conversion into British measures been given, but physical, electrical, and other modern units which are so numerous and indispensable.

With regard to some of the fundamental units we may mention that the value of the cubic inch of water, adopted here, is that which "was recently determined with great care by the Standards Department of the Board of Trade"; and in consequence of its being recently legalized, the values of the cubic foot, gallon, &c., have been revised. Throughout the work the logarithms of all the chief factors have been inserted, and at the end there is a short table of logarithms and anti-logarithms adapted for use with any number of figures up to five.

A Text-book of Geometrical Deduction. Book I., corresponding to Euclid, Book I. By James Blaikie and W. Thomson. (Longmans, Green, and Co., 1891.) THIS work forms an excellent supplement to the first book of Euclid, and by its means a systematic course of training in the art of solving geometrical deductions can be obtained. The arrangement adopted is good, and of a very progressive character. The propositions are divided into sections, and each section is subdivided into three parts: in the first a deduction is worked out in full to serve as a guide to the student; deductions similar to the one already mentioned then follow, in which the figures are in each case given and such notes as are deemed necessary for a beginner. In the last part no figures or notes are added, but occasionally references are given to the propositions on which the proofs depend. The deductions in the last two parts should be written out by the student, and the proofs made to depend on the preceding propositions of Euclid. Additional parts, corresponding to the remaining books of Euclid, are in preparation, and if they are up to the standard of the present one, the series will be found generally useful.

Elementary Algebra. By W. W. Rouse Ball. "Pitt Press Mathematical Series." (Cambridge: University Press, 1890.)

IN this book all those parts of the subject which are usually termed "elementary" are dealt with. It is a sound and well-written treatise. No deviation of importance has been made in the general order of arrange

ment that has been lately adopted, but many articles and examples which might profitably be left for a second reading have been marked with an asterisk. Permutaexponential theorem-subjects which are sometimes intions and combinations, the binomial theorem and the cluded in an elementary treatise, and sometimes excluded-have here only been lightly touched upon, and will serve as an introduction to the more detailed discussions contained in more advanced text-books. Numerous examples are interspersed in the text of each chapter, and here and there are papers and questions that have been set in various examinations. The table of contents is fuller than usual, and will enable the student to find readily any particular article to which he may wish to refer.

A Ride through Asia Minor and Armenia. By Henry C. Barkley. (London: John Murray, 1891.) THE "ride" described in this book came off in 1878, but the author writes so brightly that only very exacting readers will complain of any lack of freshness in his narrative. His journey from Constantinople occupied ninetysix days, of which fifty-three were spent in the saddle. He rode fourteen hundred miles, the average distance done each day being about twenty-two and a half miles; and, says Mr. Barkley, "if the miserable mountain roads are taken into consideration, I think this was very fair work for a lot of ponies." Apart from the personal incidents of the journey, Mr. Barkley was interested chiefly in the character, manners, and customs of the inhabitants of the districts through which he passed; and on these subjects he records a good many acute observations. It is worth noting that he speaks in high terms of the spirit of hospitality displayed in the parts of the Turkish dominions he has visited. Of course, the Turk is most hospitable to the Turk, and the Christian to the Christian; but" it often happens that the Turk receives the Christian as his guest, and the Christian the Turk." If a respectable traveller finds a want of hospitality on the part of either Turk or Christian, Mr. Barkley cannot but think it is the traveller's own fault.

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.]

Prof. Van der Waals on the Continuity of the Liquid and Gaseous States.

WITH regard to Mr. Bottomley's criticism, I should like to add to what Prof. Rücker has said that Prof. Van der Waals's book is not in any sense a treatise on the continuity of the liquid and gaseous states, but a thesis wherein is put forward the author's own work on which he claims a doctor's degree.

The preface explains that, in the attempt to determine the value of one of Laplace's capillary constants, the author was forced to proceed by theory, and that the course of these theoretical investigations led him to see that there must be continuity between the gaseous and liquid states. He was, in fact, led to his well-known characteristic equation for a substance in a fluid state, an equation in no way depending on the character of the fluidity.

This characteristic and its application are for this thesis the important things, and Prof. Van der Waals proceeds therefore to show that the results deducible from it are in complete agreement with Dr. Andrews' experiments and Prof. James Thomson's suggestions. It is not a point with him to discuss the question of continuity except as bearing on his characteristic; but this continuity is doubtless taken for the title of the thesis as being the most important deduction from his theory.

There is no question of priority: the author gives full information as to where the experiments bearing on the subject

are recorded, and only claims to have shown this continuity as a consequence of known laws.

Prof. Van der Waals has been unfortunate in that the English dress in which his thesis appears is a translation from a translation. A literal rendering would have shown that he took his descriptions and diagrams from Maxwell's "Theory of Heat' because this is a "little book which is certainly in the hands of every physicist": it would have prevented the insertion of that footnote on p. 416 alluded to by Mr. Bottomley, since the text runs, "That Maxwell joins the points c and G by a straight line I do not think happy. It is apt to lead off the track and not on to it." The first of Mr. Bottomley's quotations-and with this, I might add, the scientific part of the preface of the original concludes-should read: "These considerations have led me to perceive continuity between the gaseous and liquid states, the existence of which, as I saw later, had been already surmised by others." Vermoed (surmised) certainly seems a weak word in the light of Dr. Andrews' experiments, but it may possibly point to an earlier date for Prof. Van der Waals's theoretical conclusion than that of his thesis.

The Flying to Pieces of a Whirling Ring. DR. LODGE having set the ball of paradox rolling, perhaps I may be allowed to point out some of the paradoxes of his critics on the subject of revolving disks, of the well-known "grindstone problem." Prof. Ewing refers to two treatments of this problem, which, however, stand upon quite different footings. Prof. Grossmann's discussion reduces the problem to one in two-dimensions, and leaves an unequilibrated surface stress over both faces of the disk. Even if the disk be moderately thin, the solution cannot be considered satisfactory till the degree of approximation has been measured by comparison with the accurate solution of the problem. But Grossmann's method is precisely that of Hopkinson (Messenger of Mathematics, vol. ii., 1873, p. 53), except that the latter has dropped by mischance an r in his equation (1) [or Grossmann's (6)]. This slip I pointed out in 1886; and Grossmann's results, such as they are, flow at once from Hopkinson's corrected equations. Between Hopkinson and Grossmann this theory has several times been reproduced in technical books and newspapers without comment on its want of correctness. Such first-class technical authorities as Ritter and Winkler have also given quite erroneous solutions of the "grindstone problem."

Prof. Boys refers to Clerk Maxwell's solution. Unfortunately the editor of his scientific papers has given no word of warning about the difficulties of that solution. It involves the paradox of an equilibrated shearing stress on the faces of the disk, and this stress is comparable with the stress which Maxwell supposes to burst the stone (see "History of Elasticity," vol. i. p. 827). Thus both the solutions suggested by Profs. Ewing and Boys suffer from the same defect of unequilibrated stress on the faces. Their difference leads to the fact that Maxwell's causes a hollow disk to burst first at the outer rim, and Grossmann's at the hole.

The solution by Mr. Chree, to which Prof. Ewing refers, seems to me to lie on a higher plane than the other two, and to have been better worth reproducing than Grossmann's, although it cannot be considered as final. Mr. Chree recognizes that for his form of solution normal stresses over the faces of the disk would be necessary, and he proceeds to find their values. Grossmann failed to notice this paradox of his supposed solution, and therefore gives no measure of the amount of its error. Some years ago Mr. Chree kindly provided me for lecture purposes with a solution of the disk problem in which the stress on the disk face was zero over a circle of given radius. This was a closer approximation to the facts of the case, but as the stress was still unequilibrated at other points of the face the solution was not of course final.

If all these solutions are therefore paradoxical, where is the correct one to be sought? I fear it has yet to be worked out. Some progress can easily be made with it. It involves four series of Bessel's functions, two of either type, but the surface conditions lead to equations so complex that they will, I think, puzzle the ingenuity of our best Cambridge analysts. When solved, the work to be of practical value must be reduced to numerical tables and not left in the form of infinite series-a type of solution of elastic problem which is so common and yet so technically useless. An Italian has recently solved, by a finite number of definite integrals, the problem of the elastic spherical

shell under given surface forces: possibly something might be done for the grindstone problem in the same direction. At any rate, my object in writing to NATURE is to point out that the solutions referred to by Profs. Ewing and Boys are incorrect, and to express a hope that no competent analytical elastician will, owing to these paradoxical solutions, hesitate to try his hand at a very important problem. I am quite certain that no real solution (the paradoxical are myriad) exists prior to 1860, and pretty nearly certain that none has been achieved since, although my bibliography of papers on the strength of materials for the last twenty years is not so complete as I could wish. University College, March 20. KARL PEARSON.

Deductions from the Gaseous Theory of Solution.

FROM the gaseous theory of solution, Prof. Orme Masson has concluded (see NATURE of February 12, p. 345) that there must be some temperature above which two mutually soluble bodies will be infinitely soluble in each other. This, no doubt, is a fact, and it may be interesting to show that precisely the same conclusion can be drawn from the hydrate theory of solution.

Take first the case of a solution from which a solid separates on cooling. The body which separates, say solid water, does so owing to the tendency of its molecules to coalesce and form solid aggregates; and their tendency to do so is, we know, increased by lowering the temperature: on introducing any substance which possesses an attraction for the water molecules, the attraction of these for their fellows will be in part counterbalanced, and to get them to coalesce a lower temperature will be necessary, and the lower will this temperature be, the more foreign substance there is present; thus the freezing-point of the water will fall as the amount of, say, any salt present in it is increased, as in ADC, Fig. 1. Similarly, if

we start with the pure salt at B, its freezing-point will be lowered by the addition of water, giving us a curve such as BFEC, which meets or cuts the first curve at some point c-the miscalled cryohydric point. This is precisely what does occur; the woodcut in fact represents the crystallization of water and the hexhydrate of calcium chloride from solutions of this salt, and may be taken as a typical example of the figures obtained in all cases. A solution of the composition D will be the one containing the most water of any which can exist at the temperature t, while E is the one containing the most salt at this temperature, all solutions of intermediate composition being capable of stable existence at t. At t1 any solution weaker than F will be able to exist, since there is no inferior (i.e. for weak solutions) limit of stability, while above в there is neither superior nor inferior limit, and the two substances will be infinitely soluble in each other.

The same general results will obtain when the substances separate on cooling in the liquid instead of the solid condition, but they may be expressed in another form. From Fig. I we see that the maximum amount of B which the liquid A can hold at different temperatures is represented by CB, and that this maximum increases with the temperature; it may be represented by AC, Fig. 2; similarly, the maximum amount of A which B can contain is represented by CA, Fig. 1, or BC, Fig. 2, and we thus get in Fig. 2 a double curve which shows that at any tempera