6o& 



NATURE 



[February 5, 1920 



Fig. 3 is a photograph of a transmitter arranged 

 for use with sunHght, and mounted so that it can 

 be directed as desired. The lenses are 6 in. in 

 diameter, and the range with sunlight, when a 

 6-in. collecting lens is also used, is about 8 miles. 

 It is impossible at present to say what the ulti- 

 mate limit of range may be. It depends on the 

 apertures of the projecting and receiving optical 

 systems, the brilliancy of the source, and the 

 extent to which amplification bv means of therm- 

 ionic valves may be possible in reception. The 

 selenium cells which the author has used were 

 made by Dr. Fournier d'Albe, and they have given 

 very satisfactory results, the articulation of the 

 speech heard being extraordinarilv perfect. Their 



Fig. 3. 



special sensitivity to red light perhaps accounts 

 for the fact that a small amount of mist between 

 the sending and receiving stations has been found 

 not to interfere greatly with transmission. 



In this short article it has not been possible to 

 give more than a brief description of essential 

 points. Fuller details both of the photophone and 

 of its application to the photographic recording 

 and reproduction of sounds may be found in the 

 Proceedings of the Physical Society of London. 1 



A. O. Rankine. 



AUSTRALIAN RAINFALL AND WHEAT 

 YIELD. 

 T TNTIL large schemes for the conservation of 

 ^ water supply with a view to irrigation have 

 been carried out, the incidence of drought at fre- 

 quent intervals is bound to have a great influence, 

 not only on the sheep runs of the Australian 

 Commonwealth, but also on its wheat crop. It is 

 perhaps surprising that the relation between rain- 

 fall and wheat yield should be to a great extent 

 directly traceable, when we consider to how many 

 indirect influences the yield is exposed. The seed 

 varies in such obvious characteristics as size and 

 hardness, as well as in power of resistance to 

 disease, partly modified by the conditions under 

 which the crop producing the seed has been 



1 Proc. Phys. Soc, vol. xxxi., p. a^j, and paper read Decc'^ber ij, 1919. 



NO. 2623, VOL. 104] 



raised. There is, moreover, no constancy in the 

 soil, which differs from place to place in com- 

 position, in aspect, elevation, and slope, from 

 farm to farm in the amount and choice of fertilis- 

 ing agents, and from district to district in the dates 

 of weather changes and precipitation. There is 

 possible loss by barrenness of seed, by ground 

 pests before germination, by vermin during 

 growth, by stormy, birds, insects, and disease 

 v/hen the grain is in the ear, and much may be 

 shaken out when ripe if the harvest weather be 

 very hot and dry. 



In spite, however, of all these disturbing fac- 

 tors we find from the latest oiBcial publication on 

 the subject strong evidence of direct correlation 

 between the wheat yield per acre and the rainfall 

 of the previous winter. For South Australia and 

 the Northern Territory the correlation coefficient 

 works out at o'6i, with a probable error of o'o7. 

 It must be admitted that the data are far from 

 being homogeneous, comparatively few of the 

 stations yielding figures for the whole period. 

 The publication is entitled " Results of Rainfall 

 Observations made in South Australia and tht' 

 Northern Territory, including all available annual 

 rainfall totals from 829 stations for all years of 

 record up to 1917, with maps and diagrams, also 

 appendices presenting monthly and yearly 

 Meteorological Elements for Adelaide and Dar- 

 win " (Green, Acting Government Printer, Mel- 

 bourne, 1918), and is the fourth of a series. Pre- 

 vious volumes dealt with the Eastern Provinces 

 (Queensland, New South Wales, and Victoria), 

 and two more are contemplated to complete the 

 set by including Western Australia and Tasmania. 



There is a wealth of detail contained in the four 

 hundred or so pages, to say nothing of the 

 seasonal maps and diagrams. The territory 

 covered is large, more than 900,000 square miles, 

 and the annual rainfall varies from 407 in. at 

 Mulloorina in the centre to 6r37 in. at Darwin 

 in the Northern Territory and 45'9i in. at Stir- 

 ling W^est in South Australia. The mean annual 

 rainfall for the Northern Territory (four-sevenths 

 of the whole) is i9'32 in. (thirty-seven years' 

 average), the extremes being 30"28 in. in 1904 

 and i2'30 in. in 1905. For South .Australia the 

 mean is 9^39 in., and the extremes, curiously 

 enough also in consecutive years, 15 in. in 1889 

 and 5'88 in. in 1888. Of forty-seven counties 

 with a long record, twenty-eight had their driest 

 year in 1914, and twenty-five their wettest in 1916. 

 In quite a large number of districts, accordinglv, 

 the wheat yield per acre was lowest in 1914 and 

 highest in 1916. The most conspicuous dry 

 periods were 1895 to "P^ and 191 1 to 1915. It 

 may be noted that rainfall was deficient at Green- 

 wich also for each vear of the first of these 

 periods, but not for the secoiid. 



Conditions at .'Vdelaide, which has the longest 

 meteorological record in the district, are very 

 difTerent from those at Greenwich, but there is 

 some similarity in the rainfall. The wettest dav 

 at Adelaide in seventy-seven years was March 5, 

 1878, with 3'5o in., the number of daily falls 



i 



