"EDIBLE Earth in New Guinea" is the title of a communication by Mr. W. Meigen published in Briefe der Monatsberichte der deutschen geologischen Gesellschaft (1905, PP. 557–564). The earth in question is found on the east side of New Mecklenburg, where it is associated with decomposed coral; its main constituents are oxides of silica and aluminium; there is a smaller quantity of ferric oxide and traces of other substances, including ammonia. Mineralogically, the earth is composed of kaolin, hydrargillite, and ferric oxide; it is a fatty clay of yellowish colour, not unpleasant to the taste, and composed of very small particles. It is used for medicinal purposes, and may well counteract the laxative effects of the fish diet of dwellers on the coast. The article mentions the more important previous contributions to the discussion of the question of edible earths, of which, however, but few analyses have been published. A recent paper was noticed in NATURE of September 27 (p. 543); in vol. xxxiii. of the Journal of the Royal Society of New South Wales was published the analysis of some kaolinite from Fiji.

THE Bulletin of the Johns Hopkins Hospital for October (xvii., No. 187), in addition to articles of medical interest, contains an account by Mr. D. I. Macht of Moses Maimonides, a celebrated Jewish philosopher of the thirteenth century, who was physician to the Sultan Saladin and his successor, and the author of many religious, philosophical, and medical works. In his "Ethics" a complete system of practical hygiene is given which would well compare with the most recent text-books on the subject; lack of exercise, over-eating, alcohol, and excess are summarised as the causes of most diseases. Dr. T. R. Boggs describes a simple method for the quantitative estimation of the proteids in milk. The diluted milk is precipitated with phosphotungstic acid in hydrochloric acid solution, and the volume of the precipitate is read off in an ordinary Esbach albuminometer tube as used in wine analysis. The method is accurate to within 0.3 per cent. to 0.7 per cent., according to controls made by Kjeldahl determinations.

As agriculture in the Virgin Islands is dependent upon small cultivators, progress is hampered by the want of capital. In the annual report for 1905-6 of the experiment station maintained at Tortola, the curator, Mr. Fishlock, notes that the peasants are gradually realising the advantage of planting such permanent crops as cacao, limes, and rubber. The department also fosters cotton cultivation by supplying seed, buying seed-cotton, and preparing the lint for market.

THE July number of the Trinidad Bulletin contains the annual report for the past year, by Mr. J. H. Hart, on the Botanical Department. Seedling canes, rubber, and cacao form the largest items under plant distribution; there was also a considerable demand for young trees of Honduras mahogany, Mimusops globosa, that furnishes balata, and Cedrela odorata, the West Indian cedar. With the view of popularising its cultivation, a large number of plants of Coffea robusta was given away. In a note on the nests of Trigona bees, it is observed that the peculiar trumpet-shaped entrance is connected with the danger to returning bees of being caught by a spider that lurks near the opening.

THE second number of the Memoirs of the Department of Agriculture in India is devoted to the subject of Indian wheat rusts. Three distinct species, Puccinia graminis, black rust, Puccinia glumarum, yellow rust, and Puccinia triticina, orange rust, are commonly found. It was observed in 1904 that the first was most rampant in Central

India, while the latter two predominated further north, and therefore nearer the district where barberries are found. The authors. Mr. E. J. Butler and Mr. J. M. Hayman, have at present no explanation to offer for the origin of the disease year by year. The results obtained by inoculating barley with rust spores taken from wheat plants and vice versa show a considerable degree of specialisation, as very few of the inoculations succeeded.

THE superintendent of the Indian Museum, Calcutta, mentions in his annual report for the year 1905-6 that a number of Tibetan and Bhutanese specimens, chiefly robes, brass ware, and religious objects, was added to the art and ethnological collections, also various agricultural instruments from Assam. The report of laboratory work by Mr. D. Hooper contains, as usual, several interesting analyses. From the shoot of the common bamboo a food product is prepared, known in Assam as gass-tenga, that is eaten with rice; this contains an acid similar to aspartic acid that is probably derived from asparagin. Specimens of the bark of Picrasma javanica, used by the Karens as a febrifuge, yielded a bitter principle allied to quassiin. The analyses of latices from a number of different species of Ficus show that of those examined Ficus elastica alone furnishes rubber of commercial importance.

ATTENTION is directed in the Journal of the Society of Arts (vol. liv., No. 2812) to the soda lakes of Mexico on the great desert south of Yuma. These vast lakes of crystals of carbonate of soda are within 3000 yards of the sea. They are the property of the Mexican Government, and it is believed that they may become sources of enormous income to the country.

THE British Commercial Agent in the United States reports that the plan of storing coal under water is being adopted at a new plant west of Chicago. Twelve large cement-lined pits have been constructed with a bottom of clay soil. Their storage capacity is 14,000 tons. A 12-inch water pipe opens to the pits near the top, so that the coal can be flooded when required.

IN the Engineering Magazine (vol. xxxii., No. 1) Mr. Alfred Sang urges the practical value of industrial museums as exemplified by the Conservatoire des Arts et Métiers in Paris, and what was originally the Patent Office Museum at South Kensington. While satisfactory results must depend upon a board of management composed of experts in the various branches of science and of industry represented, the author gives examples of exhibits that would prove of special benefit to students.

IN the Journal of the Franklin Institute (vol. clxii., No. 4) Prof. Alfred J. Henry, of the U.S. Weather Bureau, gives an account of weather forecasting by synoptic charts. The method is based on two well-established facts, the general eastward drift of the atmosphere in temperate latitudes in the northern hemisphere, and the close relation that subsists between the weather and the distribution of atmospheric pressure. Within recent years there has been an appreciable gain in the accuracy of the forecasts. The period covered by the forecasts has been extended from eight to forty-eight hours, and instead of forecasts expressed in very general terms for large areas, definite forecasts are now made for all the larger towns and for each of the States and territories. The most important gain however, is in the adaptation of the forecasts to the needs of special industries, the perfection of the system of flood warnings, and the general improvement in the warnings of severe storms and cold waves.

66

printed edition of Euclid's Elementa Geometriæ (1482); the first edition of de Caus's "Les Raisons des Forces mouvantes (1615), to which, according to Arago, is due the invention of the steam engine; Daguerre's description of his invention of the Daguerreotype process of photography (1839), and the earliest works on ballooning. Bibliophiles and librarians looking out for scientific works of great rarity and interest, or for volumes of Proceedings of scientific societies and standard books on the exact sciences, will find it an advantage to consult the interesting catalogue the first part of which Messrs. Sotheran have just published.

MESSRS. GEORGE PHILIP AND SON, LTD., will shortly issue a novel perpetual calendar invented by the Rev. J. W. Wiles. It is claimed that by a simple arrangement the calendar will show the day of the week of any day in any year from the beginning of the Christian era to the end of time.

MR. W. A. SHENSTONE, F.R.S., has revised, and in some instances amplified, the essays he recently contributed to the Cornhill Magazine, and they will be published by Messrs. Smith, Elder and Co. to-morrow under the title of "The New Physics and Chemistry: a Series of Popular Essays on Physical and Chemical Subjects."

MESSRS. ARCHIBALD CONSTABLE AND CO., LTD., will publish very shortly a volume by Prof. E. Ray Lankester, F.R.S., entitled "The Kingdom of Man," containing a statement of the present position of scientific knowledge and the promise of the future.

THE second quarterly number of Science Progress in the Twentieth Century has now been published by Mr. John Murray. The ten articles included in this issue of the new scientific quarterly review range over many departments of science, and should appeal to a wide circle of readers.

THE first parts of two works of science which are being published serially by Mr. Fritz Lehmann, Stuttgart, have been received. "The Macrolepidoptera of the World," by Dr. Adalbert Seitz, is to be completed in 100 parts, and "Das Mineralreich, by Dr. Reinhard Brauns, in seventy-five parts. Both works are illustrated by excellently produced coloured plates. Messrs. Williams and Norgate are the agents of the publishers in this country.

GREENWICH OBSERVATORY AND THE POWER STATION.At the meeting of the Astronomische Gesellschaft recently held in Jena (September 12-15) Dr. Foerster directed attention to the erection of the large generating station near to Greenwich Observatory, and the consequent interference with the work of the institution. After Prof. Dyson had described the unfavourable position in which the observatory is situated, a resolution having the following effect was passed :-That the convention of the International Astronomische Gesellschaft, meeting in Jena, in view of the communication made in the latest report of the Greenwich Observatory, expresses the hope that the loss which would be occasioned if the observatory were removed may be averted. The resolution also expressed the hope that, as Greenwich has succeeded in establishing itself as the standard place, all future proposals to remove it may likewise be averted (Astronomische Nachrichten, No. 4127).

LUNAR CHANGES.-In No. 588 of the Astronomical Journal Prof. W. H. Pickering discusses Mr. Stebbins's observations of the lunar crater Linné, made during the eclipse of the moon which took place on February 8, 1906, and compares them with the similar observations made at the same time by Prof. Frost. Although some slight doubt exists as to the precision of one or two of Mr. Stebbins's measures, the curve showing the change in diameter of the spot surrounding Linné, according to his observations, agrees in general with the similar one obtained by Prof. Frost. Both show a substantial increase in the diameter immediately after the passing of the earth's shadow. Prof. Pickering ascribes this increase of diameter to the deposition of hoar-frost, or something analogous to it, caused by the drop in temperature consequent upon the screening off of the sun's rays by the opaque body of the earth. This phenomenon has now been observed by six observers working quite independently, several of whom originally prejudiced against it, therefore Prof. Pickering considers that it may be accepted as confirmed.

were

The variation of the diameter of the spot during the ordinary course of lunation has similarly been confirmed by several observers, one of whom, Dr. C. W. Wirtz, discusses his observations at some length in No. 4118 of the Astronomische Nachrichten.

ECLIPSE OBSERVATIONS.-In No. 9, vol. xxxv., of the Memorie della Società degli Spettroscopisti Italiani, Prof. Riccò concludes his account of the eclipse observations made by the Italian expedition to Alcala de Chivert in August, 1905. Among other matters he discusses white prominences," and describes those seen during the eclipse in question as faint and indistinct, especially in the lower parts, and appearing as little more than a whitish shadow projected on to the background of the corona. He also suggests that these objects are in nature somewhat of an intermediate stage between the prominences and the true coronal streamers.

Estimating the height of the various layers of the solar atmosphere by two independent methods, Prof. Riccò found that that which he calls the "6 reversing layer, or the stratum producing the so-called "flash spectrum, extends to some 3" or 2000 km. (1250 miles). That part of the chromosphere which emits D, and F especially has a height of about 7" to 9", whilst the calcium vapours of the chromosphere extend to about 15" from the base. Photographs taken on special plates with a prismatic camera show that the maximum brightness of the continuous spectrum of the corona occurs in the yellow and red regions.

THE ZODIACAL LIGHT.-During the past summer Prof. Barnard, at the Yerkes Observatory, made a number of observations of the zodiacal light, the results of which he Journal. On June 22 now publishes in No. 2, vol. xxiv., of the Astrophysical he paid special attention to the phenomenon, and found it to be much more extensive than he had previously supposed. He concludes that the light extends at least 65° north and south of the sun (assuming the southern extent to be the same as the northern), a value considerably larger than that arrived at by Prof. Newcomb, observing in Switzerland, in the summer of 1905.

This

THE MIRA MAXIMUM OF 1906.-In No. 4110 of the Astronomische Nachrichten Prof. Nijland publishes the results of his observations of Mira made during the period August 24, 1905, to February 24, 1906. The curve accompanying the paper shows that a sharp maximum occurred on January 3, when the star's magnitude was 3.9. was preceded by a very flat minimum of about the ninth magnitude, extending from the commencement of the observations until November 9, 1905, and then a steep ascent to the maximum. The lowest magnitude, 9.05, occurred on September 23, 1905.

METEOROLOGY OF THE NILE VALLEY.1 THE Egyptian Survey Department, constituted

some

years ago, is adding largely and rapidly to our knowledge of the hydrography, geology, and meteorology of the Nile basin. The director-general, Captain Lyons, R.E., has prepared and issued a monograph dealing very fully with the physiography of the Nile basin. In this work, which was reviewed in NATURE of September 6 (vol. lxxiv., p. 461), he combines the results of former observers and investigators with the data accumulated during the past ten or twelve years by his department. It is a storehouse of information relating to that most remarkable, and until recent years most mysterious, of rivers.

We propose to give a brief statement, based on the information contained in the monograph, of the more important features of the meteorology of the Nile Valley and their relations to the physiography of the whole area. The river obtains its supplies from two collecting areas, one the equatorial lake plateau (between lat. 5° S. and lat. 5 N., and long. 28° and 35° E.), and the second the Abyssinian mountain and plateau area (between lat. 7° N. and 14° N., and long. 35° and 40° E.).

The former is the larger catchment basin, and includes the Victoria, Albert Edward, and Albert Lakes, which serve as reservoirs to store the rainfall of the whole region. The Victoria Lake (equal in area to Scotland) is about 4000 feet above the sea, and is slightly lower than the mean level of the plateau. The ground rises slightly to the south and east, and rapidly to the west to the elevated peaks of Ruwenzori, which separate it from the rift valley, in which are situated the Albert Edward and Albert Lakes connected by the Semliki River. The catchment area of the Victoria Lake is only of comparatively small extent, not more than twice the area of the lake, the level of which hence varies very slightly with the season. The Victoria Nile, which issues from the north of the lake, is precipitated over the Ripon Falls, and thence passes over flat, marshy ground to the Choga Lake Swamp, and descends by a series of rapids, and finally by the Murchison Falls, to the lower level of the Albert Lake at its northern extremity in lat. 2° N.

The Albert Edward and Albert Lakes, with their tributaries, appear to collect a larger volume of water than the Victoria Lake. The Victoria Lake discharges by the Victoria Nile a nearly constant amount, averaging 500 cubic metres per second, and the Albert Lake amounts varying between 500 and 1100 cubic metres per second. The discharge of the lake system is carried off northwards from the Albert Lake by the Bahr-el-Jebel, or Albert Nile, as it is called by Sir William Willcocks. It descends rapidly from a level of 2300 feet to 1500 feet at Gondokoro (lat. 5° N.), in a narrow channel with numerous falls and rapids, and thence to Lake No (lat. 93° N.) through an extensive flat and swampy region. It is joined at Lake No by the Bahr-el-Gazal, and about eighty miles further down stream by the Sobat. The former drains a large portion of the Soudan, its head-waters being chiefly in the equatorial belt. The Sobat is formed partly by drainage from the same belt and partly from the southern face of the Abyssinian plateau.

Between Lake No and Khartoum, the main stream is now known as the White Nile. The discharge of this river below Lake No varies to a slight extent during the year, and averages only 350 cubic metres per second, and hence considerably less than the supply passing into the "The Physiography of the River Nile and its Basin." By Captain H. G. Lyons, R. E., Director-General Egyptian Survey Department.

river from the Albert Lake. The difference represents the loss by evaporation in the extensive swamp region through which these streams flow. That of the Sobat is only considerable during the rainy season, from April to December, ranging between 380 cubic metres and 1470 cubic metres per second. The White Nile below the junction of the Sobat (lat. 91° N.) to Khartoum (lat. 151° N.) receives no affluent, and flows in a broad valley as a wide stream of moderate velocity. This part of the Nile plays a subordinate but important rôle with respect to the Nile floods. From May to September the flood water brought down by the Sobat River is ponded up or held back in this reach of the Nile, and hence does not contribute to the Lower Nile flood. Captain Lyons states that this action stores up an average of about 1500 million cubic metres from the Sobat flood, which is supplied to the Nile in October, November, and December, thus prolonging the period of the Nile flood, and delaying the fall of the Nile to its low-water stage.

The main flood water of the Nile is brought down by the Blue Nile and the Atbara from the Abyssinian plateau. The rainfall occurs between June and September, and is immediately discharged down the hills into the valleys, the greatest portion down the Blue Nile, which joins the White Nile (there forming the Nile) at Khartoum. The maximum flood of the Blue Nile is about 12,500 cubic metres, and of the Atbara 5000 cubic metres, per second.

The Nile flood proper is hence due solely to rainfall in the Abyssinian and adjacent Soudan area. It commences in June, and reaches its maximum about the end of August or beginning of September. The maximum height of the Nile flood, or the total discharge during the flood period, may hence be accepted as a measure of the total rainfall over that area, just as the variations of the Victoria and Albert Lakes represent the seasonal variations of the rainfall in their catchment areas.

The Nile below the junction of the Atbara (lat. 18° N. to lat. 34° N.) receives no affluents, and flows in a comparatively narrow valley, over which the flood waters, with their rich alluvial contents, are distributed by means of a vast system of canals.

The Nile basin may hence be divided into three areas or regions, not differing greatly in breadth from south to north. The most southerly is the equatorial lake belt between lat. 5° S. and 5° N., an intermediate region between lat. 5° N. and lat. 18° N. includes the Soudan and Abyssinia, and the northerly region comprises the lower Nile basin from lat. 18° N. to the Mediterranean in lat. 34° N. The low river supply (January to May) is chiefly due to discharge from the equatorial lake area, and the summer flood supply to discharge from the Abyssinian region.

The following gives a sketch of the more important features of the meteorology of the Nile basin, based on the important information and data of Captain Lyons's monograph.

Temperature is remarkably uniform in the equatorial lake region. Thus at Entebbe, on the north shore of the Victoria Nyanza, it ranges only between a mean of 72°-7 in January and 70° in July. In the Nile basin north of about lat. 5° N. temperature is lowest in January, and attains its maximum in May in the southern half of the valley south of Khartoum, and in July in Nubia and Egypt. The annual range of temperature increases northwards from the equatorial belt to northern Egypt. The greater part of the Nile basin is within the tropics, and is throughout the whole year characterised by high temperature. That portion of it between lat. 15° N. and lat. 180 N. (in which are the meteorological stations of Khartoum, Berber, and Dongola) is the hottest and driest area in the Nile basin. It has an elevation of about 1200 feet. the south is the comparatively damp and cooler region of the Bahr-el-Gazal, the Albert Nile, and the lake plateau, whilst to the north the valley descends slowly to the relatively cool Mediterranean coast. This the Soudan hot area is one of the hottest regions in the world. The following gives a comparison of the mean monthly maximum temperature of Berber in that area, and of Jacobabad, the hottest station in India, and also of Massawa, on the Red Sea, in the same latitude as Berber :

To

The data show that at the hottest period, from May to September, the high-temperature conditions are as intense in the Soudan hot area as in Upper Sind, and are more prolonged and persistent. This hot area plays a very important part in the meteorology of the Nile basin. It is throughout the whole year much hotter than Lower Egypt. The difference between the mean day temperature at Berber and Alexandria increases from 8° in January to 16° in April and May. It thence diminishes under the influence of the monsoon rainfall in the Soudan region to 8° in August, and increases to a second maximum (12°) in November. It is undoubtedly due to the presence of this permanent hot area in the central Nile basin that northerly winds prevail almost continuously in the northern half of the basin (i.e. north of Berber). The Massawa data also indicate that the hottest portion of the Red Sea is from 10° to 16° cooler during the day hours from March to October than the land area to the west. As the width of the Red Sea in lat. 15° to 20° N. is about 300 miles, it is evident that the presence of this relatively cool area will modify considerably the air movement and pressure distribution in the adjacent land areas, more especially the Soudan comparatively low-lying area.

Much less is known of the pressure distribution than of temperature in the Nile basin. Barometric observations are being taken at a considerable number of stations. When the elevations of the observing stations have been accurately determined by the Survey Department, it will then be possible to give, for the first time, a satisfactory statement of the changes of the distribution of pressure during the year. It is to be hoped that this information will be available in Captain Lyons's monograph on the meteorology of the Nile basin, which we believe he has under preparation. Comparison of the temperature conditions of northern India and of the Nile basin suggest the probable pressure scheme. Pressure in January and the following three or four months is probably lowest in the interior regions of Africa to the south of the equator. An independent local low pressure begins to form in the Soudan hot area in March, and intensifies to some extent in April and May. This low-pressure area limits the advance of the monsoon winds in that region in the same manner that the low-pressure area in Baluchistan and Sind, and the Himalayan mountain barrier, limit the northward extension of the south-west monsoon winds in India. During the period from June to September, an extensive low-pressure area extends from the Soudan across south-west Asia to Upper India, but it is probable that the Soudan depression, due to the local thermal conditions, maintains an independent existence from the Upper India depression, and is separated by a belt of somewhat higher pressure across the Red Sea. This is not confirmed as yet by observation. Captain Lyons, however, indicates in the chart of the mean distribution of pressure in northern and central Africa in July his conviction that a local belt of low pressure stretches across central Africa between lat. 12° N. and 18° N. This either fills up in October and November or is transferred southwards.

The air movement in the Nile basin is on the whole comparatively simple. It is almost continuously from north to the north of lat. 17° or 18° N. (Berber), and is hence a drift up the valley due to permanent temperature and pressure differences between the east Mediterranean and Upper Nile valley. Also in the extreme south of the

basin (in the basin of the Victoria and Albert Lakes) it is, so far as is indicated by the available data, almost equally persistent, but from the opposite direction, that is, from south and south-east. That region is hence, during nearly the whole year, within the sphere of the south-east trades. The movement is apparently for a short period in the early months of the year light, variable, and irregular, but chiefly from north.

The air movement in the intermediate region between the equator and lat. 16° N. to 18° N. is typically monsoon. During one period of the year dry land winds (from the north) prevail, followed during the remainder of the year by humid oceanic winds (from south or west). The in fluence of the Soudan hot area begins to be shown in winds alternate March, and between northerly and southerly directions in April and May. Thus at Khartoum the percentage of steadiness decreases from about 90 per cent. in January to 40 per cent. in May. In the beginning of June a change similar to that occurring in India in that month is initiated. Steady winds, the continuation of the south-east trade winds, which have previously given heavy rain to the equatorial lake area, prevail during the next three months. The direction of the air movement rapidly changes in proceeding northwards from south to west, determined by the position of the Soudan low-pressure area and action due to the earth's rotation. The current hence advances directly to the Abyssinian mountain or plateau area, the axis of which runs due north and south, its forced ascent over which gives rise to the heavy precipitation over the greater part of the plateau. No rain falls at this time in the Red Sea coast districts on the lee side of the plateau. The plateau hence plays (but much more completely) the same part for the Abyssinian branch of the south-west monsoon current that the West Ghats play with respect to the Bombay branch. This movement holds steadily until September, when the monsoon current contracts southwards, and light, northerly winds extend slowly to the neighbourhood of the equator. There is hence a clearly marked monsoon alternation of winds and of season (dry, and wet) in the intermediate area between lat. 5° N. and lat. 18° N.

The distribution of the rainfall in the Nile basin is very clearly exhibited in a series of monthly charts in Captain Lyons's monograph. A chart showing the amount and distribution of the average annual rainfall would have been a useful and valuable addition to the series. Charts of annual or seasonal rainfall are, as a rule, even more valuable for comparison than charts of monthly rainfall.

The air movement has shown that the Nile basin may be divided meteorologically into three areas, viz. the area of dominant northerly winds (north of lat. 17° N.), the area of alternating monsoon winds between lat. 17° N. and the equator, and the area of dominant south-easterly winds south of the equator. The rainfall differs greatly in its characteristic features in these three areas. In the northerly region it occurs during the winter months, as in Syria, the Euphrates valley, and the Iran plateau, and is small and very variable in amount. The average annual fall at Alexandria and Suakim is about. 5 inches, at Port Said 2 inches, and at Suez inch. In the intermediate monsoon region practically no rain falls from November to April. Thunderstorms occur in May, chiefly in the southern districts, and frequent heavy rain from June to September or October, according to position. The rainfall is heaviest on the western and central portions of the plateau. In the Himalayas the rainfall is, as a rule, heaviest at an elevation of about 4000 feet. The Abyssinian data are too scanty to show whether there is any line of maximum rainfall lower than the level of the interior plateau or higher mountain ranges, where the highest elevations exceed 15,000 feet.

The precipitation in the equatorial lake region has a double maximum and minimum in its annual variation, related, as Captain Lyons points out, to the apparent movement of the sun. The rainfall is small in amount during the period of heavy rainfall in the monsoon region from June to September. It is heavy from October to December, and again in March and April, and is light to moderate during the intervening months of January and February, and moderate in May.

1