stant, is greater than 1; species Q, gases for which k is less than 1. On looking at the page of NATURE referred to, it will be seen that Perry questioned or even denied the possibility of a gas of species Q. His theorem is-A finite spherical globe of gas, given in equilibrium with any arbitrary distribution of temperature having isothermal surfaces spherical, has less heat if the gas is of species P, and more heat if of species Q, than the thermal equivalent of the work which would be done by the mutual gravitational attraction between all its parts, in ideal shrinkage from an infinitely rare distribution of the whole mass to the given condition of density.

§ 6. From this we see that if a globe of gas Q is given in a state of convective equilibrium, with the requisite heat given to it, no matter how, and left to itself in waveless quiescent ether, it would, through gradual loss of heat, immediately cease to be in equilibrium, and would begin to fall inwards towards its centre, until in the central regions it becomes so dense that it ceases to obey Boyle's law; that is to say, ceases to be a gas. Then, notwithstanding Perry's theorem, it can come to approximate convective equilibrium as a cooling liquid globe surrounded by an atmosphere of its own vapour.

7. But if, after being given as in § 6, heat be properly and sufficiently supplied to the globe of Q-gas at its boundary, and the interior be kept stirred by artificial stirrers, the whole gaseous mass can be brought into the condition of convective equilibrium. §8. In the course of the communication to the Royal Society of Edinburgh, curves were shown representing the distributions of density and temperature in convective equilibrium for four different gases, corresponding to the four values of k

Gas (1) k=13 (approximately the value of k for the monatomic gases, mercury vapour according to Kundt and Warburg, argon, helium, neon, krypton, and xenon).

Gas (2) k=1 (approximately the value of k for seven known diatomic gases, hydrogen, nitrogen, oxygen, carbon monoxide, nitric oxide, hydrochloric acid, hydrogen bromide).

Gas (3) k=1 (approximately the value of k for water vapour, chlorine, marsh gas, bromine iodide, chlorine iodide).

Gas (4) k=14 (approximately the value of k for sulphur dioxide).

Four of these curves agree practically with curves given by Homer Lane for k=13_and_k=13, in his original paper to the American Journal of "Science, July, 1870.

89. In a communication to the Edinburgh Royal Society of February, 1887, "On the Equilibrium of a Gas under its own Gravitation only,' "I indicated a graphical treatment of Lane's problem by successive quadratures, which facilitated the accurate calculation of numerical results, and was worked out fully for the case k by Mr. Magnus Maclean, with results shown in a table on p. 117 of the Proceedings of the Royal Society of Edinburgh, vol. xiv., and on p. 292 of the Phil. Mag., March, 1887. The numbers in that table expressing temperature and density are represented by two of the curves now laid before the society. The other curves represent numerical results calculated by Mr. George Green, according to a greatly improved process which he has found, giving the result by step by step calculation without the aid of graphical constructions.

The mathematical interpretation of the solution for Perry's critical case of k-1, and for gases of the Q-species, is exceedingly interesting.

The communication included also fully worked out examples of the general solution of Lane's problem

for gases of class P of different total quantities and of different specific densities.

§ 10. In my communication to the Royal Society of Edinburgh, of February, 1887, I pointed out that Homer Lane's problem gives no approximation to the present condition of the sun, because of his great average density (1.4). This was emphasised by Prof. Perry in the seventh paragraph, headed Gaseous Stars," of his letter to Sir Norman Lockyer on Life of a Star" (NATURE, July 13, 1899), which contains the following sentence :--

66

"The

"It seems to me that speculation on this basis of perfectly gaseous stuff ought to cease when the density of the gas at the centre of the star approaches 0.1 or one-tenth of the density of ordinary water in the laboratory." KELVIN.

THE PROBLEM OF THE RHODESIAN RUINS.

TH

1

The

HE recent investigation of some of the famous ruins of Rhodesia, conducted in 1905 by Dr. D. Randall-MacIver on behalf of the British Association and the Rhodes trustees, has resulted in an entirely fresh view of their origin and age. hitherto generally accepted view, that these buildings were erected in very ancient days by a Semitic people, whose search for gold led them thus far afield, has received a serious check. Dr. MacIver's researches, conducted upon the lines of archæological investigation, point to the buildings in question being of comparatively recent date, not earlier, in fact, than late mediæval times. This result is the more striking when we remember that his previous researches have been mainly archæological, conducted chiefly in Egypt, and that, in consequence, we might expect a certain degree of bias in favour of retaining the ruins within the sphere of archæology. That a trained archæologist has been unable to find evidence of high antiquity upon the sites investigated is at least a strong point in favour of his argument.

Dr. MacIver made excavations on seven sites in various parts of Rhodesia, these being :-(1) Inyanga, on the Cecil Rhodes estate, sixty miles north of Umtali; (2) the Niekerk ruins to the north-west of Inyanga; (3) a site three miles south of Umtali; (4) Dhlo Dhlo, in the Incisa district; (5) Nanatali, sixteen miles east of Dhlo Dhlo; (6) Kami, fourteen miles west of Bulawayo; and (7) Great Zimbabwe, in the Victoria district; the site which hitherto had received the greatest attention. These sites were well selected as being distributed over a wide area, and, moreover, as differing considerably from one another both in general character and in special features, as also in the greater or less degree of elaborateness in their structure. It may be remarked at once that the distinctive features observable in comparing the different buildings are often no less remarkable than are the points of similarity. No two seem to be alike, and the divergences and specialisation render their individuality very striking.

The principal questions to be determined in regard to these remarkable buildings were: By what people and at what period were they erected? The controversy, which is still active, centres mainly upon these two main points, and the older theory of their Semitic origin and great antiquity, urged by Mauch, Bent, Keane, Hall, and others, is being maintained steadfastly and strenuously by several authorities. Dr. MacIver in the title of his book, "Mediæval Rhodesia," has hoisted his fighting flag. His conten

1"Medieval Rhodesia." By Dr. David Rardall-MacIver. Pp. xv+:06. (London: Macmillan and Co., Ltd., 19c6. Price 20s. net.

tion is that none of these buildings are referable to an earlier period than mediæval or post-mediæval times. He argues that none of the objects hitherto discovered in excavating within the area of the ruins would be recognised by an archæologist as more than a few centuries old; and that the objects, when not immediately recognisable as mediæval imports, are of characteristically African type." Inyanga and the Niekerk ruins do not appear to have produced any but native African objects, and at Umtali a fragment of glazed stoneware was the only foreign object found. At the better-known sites, Dhlo Dhlo, Kami, Nanatali, and Zimbabwe, a fair number of imported objects have been found, but here again Dr. MacIver holds that in no case is there evidence of a pre-mediæval antiquity. As far as possible, he endeavoured in his excavations to reach the lowest strata, and to explore the levels which must be contemporary with the earliest portions of the walls of the buildings, and the objects found therein were naturally considered by him of the highest importance.

It was at Dhlo Dhlo that he discovered his most valuable piece of evidence. The absence of objects of foreign workmanship and of known date at the Inyanga, Niekerk, and Umtali sites rendered impossible the assignment of any definite period to the buildings there, although the negative evidence may be held to indicate the lack of foreign influence, which itself may possibly be regarded as pointing to these sites being earlier than the others which were examined, a view which is held by the author on structural grounds. At Dhlo Dhlo, on the other hand, numerous imported objects were found, and in excavating one of the platforms upon which a dwelling had been erected, and which Dr. MacIver asserts most positively is contemporaneous with the earliest portion of the building, he came across a piece of blue and white Nankin china in the unbroken cement floor of the dwelling. This fragment is shown (No. 20) in the illustration reproduced. If this cement floor was, as he maintains, erected at the same time

as the oldest walls of the main building, we must certainly admit the validity of his contention that the building cannot antedate the fragment of porcelain, and that the date of erection, therefore, cannot be pushed back beyond late medieval times. His critics appear willing to admit the validity of his argument as regards Dhlo Dhlo, but they urge that the buildings on this site are relatively late, and that this dating will not hold good in the case of the buildings at Great Zimbabwe, which they regard as much earlier.

Dr. MacIver regards the principal buildings, such as the so-called "Elliptical Temple "at Zimbabwe, as being fortress-kraals, and urges that the "Elliptical Temple" itself was the fortified residence of the Great Chief, or. Monomotapa, whose sway extended over an enormous area and a very extensive population. To understand how architectural feats, such as the finer Rhodesian buildings at Dhlo Dhlo, Nanatali, and Zimbabwe, can have been achieved by the precursors of the modern South African natives, it is necessary to assume that in those days there was organisation of a far higher character than has obtained. in recent years, organisation under great chiefs whose power and intelligence were of a relatively high order. This would appear, from the Portuguese and other records, to have been the case in the days of the Monomotapan empire of the Middle Ages down to the close of the sixteenth century. The Monomotapa, or paramount chief, may well have resided at Zimbabwe, and he is recorded to have had captains in various fortresses elsewhere. The organisation of labour implied by the elaborate and decorated stone architecture is certainly remarkable, more particularly when we compare these edifices with the results of the constructional efforts of the modern Kafir peoples; but under an intelligent and powerful ruler, and under stable conditions of life, a degree of culture may have been reached far higher than it is possible for smaller communities under lesser chiefs to maintain. It seems well within the bounds of probability that under such conditions even the finer buildings may have been erected by the more progressive and united precursors of the present native inhabitants of Rhodesia.

Even more remarkable, in some respects, than the huge "fortified kraals" are the terrace walls on the Niekerk site described by Dr. MacIver. These stonebuilt walls form irregular concentric rings round the hills upon which the villages were situated, and although structurally simple, cover an enormous area extending in close formation over a space of upwards of fifty square miles. They do not appear to have been erected as supporting walls for agricultural terraces, nor to have been connected with an irrigation system, and, in the absence of evidence to the contrary, one must assume that their purpose was defence, though one accepts this view somewhat reluctantly, for, when regarded as an elaborate system of defensive girdle walls, one cannot but admit that their practical value is hardly commensurate with the enormous labour expended upon them. They recall to one's mind the sementera walls of Luzon, in the Philippines, which also form long, irregular, though concentric alignments up the slopes of the hills, following their contours, covering, too, a very large extent of country. In the case of the sementeras there are transverse walls dividing up the terraces into sections. They are purely for agricultural purposes, and are mostly, though not all, connected with a wonderful system of irrigation. It might be of use to compare the sementera system with the Niekerk terrace walls, on the chance of a clue to the

mon to all such works, but most of them, we fear, would be classed by Lamb among the books which are no books. It is not so with Mendeléeff's "Principles." In its insight, in its grasp of detail and of principle, in its extraordinary power of coordination, in its suggestiveness, and in its wealth of speculation, it is a book among books, and may be read with profit and a pleasure occasionally tinctured with amusement by every true student, no matter how old. To those who had the good fortune to know its author person. ally it reflects the man in every page. Even the footnotes are instinct with character and originality. Mendeléeff's "Principles" may be said to stand in the same relation to the chemistry of the latter hali of the nineteenth century that Dalton's "New System" did to the chemistry of the earlier half. Each work was the definite and orderly presentation of the doctrine and philosophy of its author.

To Beilstein's life and services to chemical science we have already made reference; of Menschutkin, whose death has only just been announced, we hope to speak later. Our immediate concern is with the most distinguished of the eminent triumvirateDmitri Ivanovitsch Mendeléeff. The chief facts of Mendeléeff's personal history have been given in No. xxvi. of the series of "Scientific Worthies," which appeared in these columns so far back as 1889. It is sufficient here to recall that he was a Siberian, born at Tobolsk on February 7th (N.S.), 1834. He died, therefore, within a week of his seventy-third birthday. He was the seventeenth and youngest child of Ivan Paolowitsh Mendeléeff, Director of the Gymnasium at Tobolsk, who, shortly after the birth of his son Dmitri, became blind and lost his position. The family thereby became practically dependent upon the mother, Maria Dmitrievna Mendeleeva, a woman of great energy and force of character, who established a glass works in the town, on the profits of which she brought up and educated her large family. The story of Mendeléeff's youth and early struggles is given in the preface to his great work "On Solutions," which he dedicated to his mother's memory in a passage of singular beauty and power. At the age of sixteen he was sent to St. Petersburg, but, owing to official restrictions, he was prevented from studying chemistry under Zinin at the University, as he had intended, and was transferred to the Pedagogical Institute, where he came under the influence of Woskresenky in chemistry, and of Lenz in physics. Whilst at the institute he wrote his first paper on "Isomorphism," and after serving in the Gymnasium at Simferopol and at Odessa, he gained his Magister Chemiae in 1856, and was made a privat-docent in the University of St. Petersburg.

There is hardly a department of chemistry in which Mendeléeff did not labour, at one time or other, during the thirty years of his activity as a teacher. Chemical mineralogy, chemical geology, and the chemistry of aliphatic substances in turn, and apparently with equal zeal, attracted his attention. It is to this catholicity and power of taking broad and comprehensive views of the operations of chemistry that Mendeléeff owes his eminence as a chemical philosopher. But it is in the domain of physical chemistry that his fame as a worker chiefly rests. His early papers on the thermal expansions of liquids above their boiling points up to temperatures at which their cohesion and latent heats are nil, and at which the liquid becomes gaseous independently of pressure and volume, anticipated the researches of Andrews, and were, in their turn, a development of the observations of Cagniard de la Tour, Wolff and Drion.

At about this period he was attracted to the special line of inquiry and of speculation which was the dominant and most striking feature of his scientific activity, and which eventually culminated in the great generalisation with which his name is inseparably connected. It is easy to detect in these early attempts at tracing the relations between the physical and chemical properties of substances and their molecular and atomic weights the germs of the conception which eventually tock shape as the Periodic Law. His work on specific volumes was begun in 1855, and was continued by him in Heidelberg, where he went in 1859, and where he remained until 1861. Germany would appear to have exercised no permanent influence on Mendeléeff. He worked alone, and seems to have derived nothing from personal contact with Bunsen and Kopp. It is significant of his perspicacity that he should at this time have clearly appreciated and publicly declared his belief in the value of Gerhardt's work on the determination of the chemical molecule at the very period, in fact, when the whole weight of German authority was directed against the doctrine of the new French school. Returning to St. Petersburg, he became professor of chemistry at the Technological Institute. In 1866 he was transferred to the University, and in 1890 he was appointed head of the Standards Department.

The same faculty of perceiving the underlying basis of a physical generalisation is seen in the notable paper which he communicated to our Chemical Society in the year following his election into that body as an honorary foreign member, in which he developed a general expression for the expansion of liquids under constant pressure, analogous to that which expresses Dalton's law of the uniformity of expansion of gases. The formula V=1+ kt applies only to a so-called ideal gas; in like manner, Mendeléeff's expression is to be regarded only as a first approximation-that is, as applicable only to ideal liquids. In the case of actual liquids the deviations from the ideal form increase not only as the liquid approaches the point of change of state of aggregation, but also augment with diminishing density, increasing cohesion, and diminishing molecular weight, just as Mendeléeff himself showed that the deviations from Dalton's law were related to the molecular weights of the gases. Subsequent observers, by applying van der Waals's theory of the relation between pressure, volume, and temperature, have shown that the development of Mendeléeff's formula affords a simple and ready method of calculating the critical temperature of bodies from their thermal expansions as liquids-in other words, of reaching the same constant by a method analogous to that employed by Mendeléeff himself to the observations of Kopp and Pierre.

Mendeléeff signalised his connection with the University by the publication of his "Principles of Chemistry," which has passed through many editions in Russia, and has been translated into German and English. It is not easy to avoid speaking of this work in terms which savour of hyperbole. Most treatises on chemistry owe a great deal to their predecessors. Indeed, there is probably no form of literature which so obviously proceeds on strictly evolutionary principles. But Mendeléeff's great work is a thing apart-something sui generis. The bare facts of chemistry, in greater or less detail, are com

Mendeléeff's work on the relative densities of aqueous solutions of alcohol takes its place as a classic alongside the works of Blagden and Gilpin, and of Drinkwater and Fownes in this country, and, as in the case of these observations, has been utilised by Continental Governments for the purposes of revenue. These determinations were applied by Mendeléeff to the elucidation of a theory of solution, and in a paper, also communicated to our Chemical Society, he sought by means of them to reconcile Dalton's doctrine of the atomic constitution of matter with modern views re

specting dissociation and the dynamical equilibrium of molecules. How far this attempt will be ultimately successful time alone can show. Mendeléeff had little sympathy with the theory of electrolytic dissociation, which, he declared, was not in harmony with the facts of observation, and was of little use in facilitating our comprehension of the true nature of solution. was he more predisposed towards the conception of electrons, although perhaps his belief in the integrity of the atom was hardly so fundamental as that of Dalton, who would have gone to the stake rather than recant his declaration : Thou canst not split an

atom! "

66

Nor

The story of the rise and development of the Periodic Law is so well known that it is unnecessary now to dwell upon it. By a good fortune, which some may regard as evidence of predestination, Mendeléeff lived to see the verification of his predictions in the discovery, in rapid succession, of gallium, scandium, and germanium; and no seer ever prophesied more truthfully. It was the astonishing accuracy of Mendeléeff's prognostications, and the apparent boldness and confidence with which they had been uttered, that profoundly impressed the whole scientific world, and secured for his generalisation a respect and acceptance for which otherwise it would have had long to wait. This generalisation is now woven into the fabric of modern chemistry, and is universally accepted as the only rational basis of classification. Like many other great natural truths, we are able, on looking back, to discern its germs in the tentative efforts of previous thinkers who more or less dimly appreciated the significance of the facts upon which it is based, but it is perfectly certain that Mendeléeff knew nothing of the prior work of De Chantcourtois and of Newlands, and was no more influenced by it than was Dalton by Richter or by the "Comparative View of the Phlogistic and Antiphlogistic Theories" of William Higgins. In the memorable Faraday lecture which he gave to the Chemical Society in 1889, Mendeléeff, with a true nobility of mind and a modesty which revealed the real greatness of the man, gave adequate expression to his appreciation of the efforts of his predecessors, claiming for himself only courage and intrepidity in placing the whole question at such a height that its reflection on the facts could be clearly seen."

66

own

The Periodic Law has so far stood the test of experience, and each new extension of the science is consistent with its previsions. The inert gases of the atmosphere find their place in the system, and the only radioactive substance the chemical properties of which have been sufficiently investigated has an appropriate position among its correlated elements. In the old days the followers of Stahl sought to make the conception of phlogiston an all-embracing doctrine. Mendeléeff anticipated these attempts as regards his generalisation by showing that even the universal ether may be included within his system. In his last paper, published in 1902, entitled "An Attempt towards a Chemical Conception of the Ether," he starts with the assumption that the ether possesses mass, and that it has an atomic weight many times less than that of hydrogen, something of the order of 10-6 when H=1; that it is monatomic like argon and helium, and that by its small density and extremely rapid motion it permeates all matter and space. The ether thus becomes, not an affection of matter, but a distinct entity capable of being attracted by elements in proportion to the weights of their atoms, and he held that the phenomena of radio-activity could be explained by the gradual emission of this ether from such substances as uranium and thorium which have the highest atomic weights of the elements.

The truth embodied in the Periodic Law has led many to suppose that this generalisation lends sup

port to, and is indeed the proof of, the validity of the assumption of a primordial matter. Mendeléeff himself declined to see that such an inference was warranted. He saw nothing in the law inconsistent with the idea of the individuality of the elements, holding that until it could be definitely shown that one element could be transformed into another, or that ether and matter were mutually convertible, the elements must be regarded as distinct and separate entities, immutable and unchangeable.

Mendeléeff not unfrequently visited this country, and was personally known to many British chemists, to whom he was always welcome. His tall and commanding presence, his fine head, with its tangle of long, wispy white hair, his expressive features, his guttural utterance, the wisdom and originality of his talk, his shrewdness and sense of fun, all stamped him as an uncommon and strong personality, which immediately made its presence felt in any company in spite of the innate modesty of the man. Of wide liberal views, intensely national, and a great power in the University, Mendeléeff was doubtless a thorn in the side of bureaucratic Russia, and it was currently reported that the frequent foreign missions on which he was sent were so many covert attempts to keep him at arm's-length.

Every scientific honour that this country could pay was awarded to him, and he was profoundly touched and deeply grateful for the sympathy and appreciation thus extended to him. On the occasion of his delivering the Faraday lecture it fell to the writer's duty, as treasurer of the Chemical Society, to hand him the honorarium which the regulations of the society prescribe, in a small silken purse worked in the Russian national colours. He was pleased with the purse, especially when he learned that it was the handiwork of a lady among his audience, and declared that he would ever afterwards use it, but he tumbled the sovereigns out on the table, declaring that nothing would induce him to accept money from a society which had paid him the high compliment of inviting him to do honour to the memory of Faraday in a place made sacred by his labours. T. E. THORPE.

loss.

on

Born at Campobello di Mazzara (Sicily). December 4, 1862, Mascari proceeded in due course to the University of Palermo, where he took the engineering course and obtained his degree in that faculty in 1887. It was while there that he developed the predilection for astronomical investigations, and, under the guidance of Prof. Riccò, worked with that activity and intelligent ability which were the outHe was later standing features of his whole career. appointed to the position of assistant to the Piazzi Foundation, and thus was fortunate enough to be able to continue his association with Prof. Riccò-an association which has proved of inestimable benefit to the study of solar physics.

In 1892 Mascari was appointed first assistant at the Observatory of Catania, where the solar prominence observations, commenced by Tacchini at Palermo in 1872, were continued. Probably only those who have had to use these Italian observations in discussions of collateral phenomena are aware how well this task was performed, and how much the science of solar physics owes to the indefatigable labours