September 15, 1910] 



NATURE 



343 



ni Australian waters H.M.S. Fantdme laid a line between 

 Thursday Island and Cartier Island, the latter being 

 formally annexed by Captain Pasco on March 17. 



The connection between non-periodical variations of 

 rainfall and famines in German East Africa is discussed 

 by Dr. E. Kremer in No. i, year iqio, of Aus dem Archiv 

 tier Deulschen Seewartc. Normal annual rainfall values 

 can only be safely calculated from thirty to forty years' 

 observations, and monthly values require a much longer 

 period. These are not available for the district in ques- 

 tion ; the author directs attention to the fact, but he makes 

 the best possible use of the short series at his disposal 

 and of the information obtainable from travels and ex- 

 peditions. The German colony extends from latitude 

 i°— 12° S., and longitude 29°— 405° E., and the rainfall 

 may be divided generally into one rainy period in the 

 south and two rainy periods in the north. The greater 

 part of the country has less than 40 inches of rain per 

 annum ; only the central and northern coastal districts 

 and the mountainous parts receive more, while the whole 

 of the interior tableland, to about longitude 33°, receives 

 less than 30 inches. The author discusses the possibility 

 of predicting famines or scarcity on the lines successfully 

 followed in India, and shows, e.g., that scanty rainfall, or 

 its unfavourable distribution, in Zanzibar is followed by 

 famine over a large part of the East .\frican coast, and 

 further that the distribution of atmospheric pressure in 

 the months preceding the rainy periods is intimately con- 

 nected with the amount of subsequent rainfall. 



The Royal Meteorological Institute of the Netherlands 

 has issued a paper, by Dr. P. H. GalM, in which a large 

 number of wind and current observations, chiefly from 

 the Indian Ocean, are examined with the special purpose 

 of testing the theories of Nansen and Ekman with regard 

 to the effect of the deflecting force of the earth's rotation 

 in changing the direction of drift currents and to the 

 speed of current set up by wind action. Dr. Gall^ finds 

 that when the drift influence is alone at work in pro- 

 ducing surface current, the angle between the wind and 

 the resulting current is between 40° and 50°, substantially 

 in agreement with the theoretical values. A wind velocity 

 of I metre per second produces an average speed of 

 current of between 4 and 5 centimetres per second. 



Mr. H. T. Ferrar contributes an interesting' paper on 

 " The Creation of an Artificial Water-table in Egypt " to 

 the July number of the Cairo Scientific Journal. The 

 author is of opinion that the modern irrigation construc- 

 tions in themselves have had little direct effect in raising 

 the level of the subsoil water in the Nile delta, but he 

 deals specially with the changes induced in the irrigation 

 system of Lower Egypt by the works of Mohammed Ali 

 Pasha, who in 1820 excavated a number of deep perennial 

 canals capable of discharging the low-level summer supply 

 of the Nile. Observations made at more than one hundred 

 and fifty experimental tube wells support the view that 

 there are two water tables in Lower Egypt : (i) a 

 natural water-table, which is independent of the works of 

 man except locally where extra permeability allows a 

 constant supply of irrigation water to be added ; and (2) an 

 artificial water-table which was created by the act of the 

 introduction of perennial irrigation by Mohammed Ali 

 Pasha. It is thought that the artificial water-table (2) has 

 gradually become higher, owing mainly to excessive water- 

 ing of crops, until at the- present day it has a deleterious 

 effect upon the fertility of the soil. Mr. A. Lucas offers 

 some criticism of Mr. Ferrar "s conclusions in the August 

 number of the same journal. 



NO. 2133, VOL. 84] 



J The Journal de Physique for September contains the full 

 text of a communication made to the French Physicaf 

 Society in March last by M. C. Fery, describing a spectro- 

 graph with a prism having spherical faces so designed that 

 llie image of the slit is in focus on the photographic plate. 

 Both faces of the prism are concave to the incident light, 

 and the back surface is silvered so as to reflect the light. 

 The slit, the centres of curvature of the two faces, and 

 the spectrum produced are on a circle which has the radius 

 of the first surface of the prism for diameter. With a 

 prism of quartz a metre from the slit and screen, the 

 spectrum photograph obtained is 22 centimetres long, and 

 the definition is very good throughout. Like the Rowland 

 concave grating, the apparatus is astigmatic. 



A NEW form of colorimeter has been sent for our inspec- 

 tion by Messrs. E. B. Atkinson and Co., Hull. The liquid 

 to be tested is placed in a rectangular glass cell of a few 

 cubic centimetres capacity, by the side of which works a 

 vertical hollow glass wedge filled with a standard coloured 

 solution. By means of a screw the wedge is raised or 

 lowered, thus bringing various thicknesses of the standard 

 colour into the same field of view as the cell. Both 

 liquids are viewed horizontally through a slit against a 

 white background, and the position of the wedge is 

 adjusted until a depth of tint is found equal to that of the 

 liquid under examination. A graduated scale shows how 

 much the wedge has been raised, and by reference to a 

 chart the corresponding proportion of colouring ingredient 

 in the liquid is obtained. Wedges calibrated for certain 

 purposes— e.^., the colorimetric determination of iron, 

 copper, ammonia, hcemoglobin, iodine — are supplied by the 

 makers ; but the user can himself readily calibrate the 

 instrument for his own particular purposes. For example, 

 a medical man could with little trouble construct a standard 

 haemoglobin chart for his own district. Other advantages 

 claimed for this " universal " colorimeter are (i) saving of 

 time, a reading requiring only about a minute ; (2) exact- 

 ness, since the two colour surfaces compared are contiguous ; 

 and (3) economy of cost. The apparatus appears to be 

 especially suitable for operations involving a number of 

 colorimetric estimations of the same type. 



The fifteenth number (No. 3 of 1910) of the Italian 

 review Scientia, now in its fourth year, is largely occupied 

 with philosophy. The ideas of Poincar^, Bergson, 

 Einstein, and the Pragmatists naturally form the text of 

 discussions, such as that by F. Severi on " Hypothesis 

 and Reality in Geometrical Science," or Chwolson's " Can 

 we apply Physical Laws to the whole Universe?" or 

 F. Enriques's criticism of Pragmatism. The theory of two 

 star-streams interpenetrating one another is described bv 

 Mr. A. S. Eddington, of the Royal Observatory, Green- 

 wich. M. Guignebert sketches the rise of Christianity in 

 accordance with the sifted conclusions of recent research. 

 Mr. Abegg's article on " Chemical Affinity " has a 

 pathetic interest, from the fact that while it was in the 

 press the author met his death in an aeronautic accident. 

 The useful rasscgiie of various sciences are continued ; 

 Mr. E. S. Russell reviews the claims of epigenesis versus 

 evolution. The advantages of a " mnemonic " theory of 

 heredity, in accordance with the suggestions of Semon, 

 Rignano, and Francis Darwin, involving an " interiorisa- 

 tion of external stimuli," is well put. M. Landry supplies 

 his annCial review of economic research. M. le Comte 

 de Baillehache defends his system of electric units — a 

 dimensional system — put forward last year in his " Unites 

 electriques. " Reviews of scientific periodicals, notes of 

 scientific meetings, and critical notices of books, make up 

 an issue that is very level in quality. The books reviewed 



