October 28, 1920] 



NATURE 



291 



tion of the breeding period, and Dr. Orton inclines 

 to disregard this factor, assigning little or no signi- 

 ficance to it. Salinity variations may be so capricious 

 in relation to the distribution and habits of some 

 animals that these, too, are difficult to correlate with 

 spawning cycles. On the other hand, temperature 

 variations seem to explain many facts, and the dis- 

 cussion deals largely with these changes. There is a 

 temperature constant which applies to most marine 

 species, and breeding proceeds when this condition is 

 satisfied. Growth (and absolute dimensions) will depend 

 to some extent upon the fraction of the lifetime of the 

 animal occupied in breeding. One might therefore 

 expect that a semi-s<dentary or sedentary animal at 

 the extreme northerly limit of its distribution would 

 be generally larger than one in which the tempera- 

 ture of the sea remained near the constant expressing 

 the temperature optimum. It is suggested that the 

 approximately uniform conditions obtaining in some 

 polar, tropical, and abyssal marine areas may be con- 

 nected with prolonged, or even continuous, breeding 

 f)eriods. The geographical distribution of a marine 

 animal is obviously controlled by its phvsiological 

 temperature constant, and deviations from the latter 

 may be the mode of origin of sub-species or varieties. 



Among a number of water-supply papers issued b\ 

 the United States Geological Survey special attention 

 may be directed to several series of river plans and 

 profiles constructed in order to determine the location 

 of undeveloped water powers. Recent volumes include 

 "Profile Surveys of Rivers in Wisconsin " (Paper 417), 

 "Profile Surveys of Skagit River Basin, Washington " 

 (Paper 419), and " Profile Surveys along the Rio 

 Grande, Pecos River, and Mora River, New Mexico" 

 (Paper 421). The maps show the outlines of the river- 

 banks, islands, the position of rapids, falls, shoals, 

 and existing dams, and the crossings of all ferries 

 and roads, in addition to the contours of the banks 

 to an elevation high enough to indicate the possibility 

 ■ of using the streams. Data concerning the volume 

 of the flow have been published in previous papers. 



The Monthly Bulletin of the Weather Bureau of 

 ] the Manila Central Observatory issued by the 

 Government of the Philippine Islands contains de- 

 tailed observations for numerous stations situated in 

 different parts of the group. The monthly results for 

 August, 19 19, are to hand, prepared under the direc- 

 tion of the Rev. Jos^ Algu^, S.J. Rainfall at Manila 

 during the month was 78-10 in., which breaks all 

 records since the foundation of the observatory in 

 1865 ; the previous highest rainfall in August was 

 4315 in. in 1877, and the highest for any month of 

 the year was 57-88 in. in September, 1867. 



Thb exact relations between climate and the growth 

 ,of crops, and the possibilities of accurate forecasts of 

 pro<luction, is the subject of an article bv Mr. T. .\. 

 Blair in the Scientific \fonlhly for October. Mr. Blair 

 shows in the case of niaizo in Ohio that the United 

 States Weather Bureau has found that a July rainfall 

 of less than 3 in. means an average yield of 30 bushels 

 NO. 2661, VOL. 106] 



per acre, and that a rainfall of 5 in. or more results in 

 38 bushels. When the July rainfall is 3J in. the yield 

 is 15,000,000 bushels greater than when it falls short 

 of that amount by half an inch. In the four great 

 maize-growing States of Indiana, Illinois, Iowa, and 

 Missouri the addition of half an inch to a total of 

 2J in. adds 10 bushels per acre to the average yield. 

 .•\ more precise relationship is found in the idea of 

 critical period — that is, certain short periods of time 

 in the growth of any crop, during which its future 

 prospects are largely determined. With some crops 

 this is a single period; with some temperature, and 

 with others rainfall or sunshine, are the most im- 

 portant factors. In the case of maize in Ohio it has 

 been found that the first ten days of .August are the 

 critical period as regards rainfall. The application of 

 this knowledge may be used to increase production in 

 two ways. First, it will be possible to determine what 

 crops are climatically suited to a particular district ; 

 and, secondly, by the use of early or late varieties and 

 b\- the help of fertilisers or other means, the crop 

 may be advanced or retarded so as to bring the critical 

 period into coincidence with favourable weather. 



Science Progress for October contains an article by 

 Dr. .Aston which summaries his work on the atomic 

 weights of the elements as determined by his method 

 of the "mass spectra." These spectra arc obtained 

 on a photographic plate placed in vacuo by deflecting 

 positive rays on to it by passing them first through an 

 electric, then through a magnetic, field in such a way 

 that all the rays corresponding to an element of given 

 mass are concentrated in a short line on the plate 

 and those of different masses in other parallel lines. 

 The separation of the lines is sufficient to fix the 

 atomic weights to considerably less than a tenth of a 

 unit. By this method it has been possible to demon- 

 strate that boron consists of two, neon, silicon, and 

 chlorine of two or three, bromine of two, krypton of 

 six, and mercury of at least two isotopes— that is, 

 elements of the same chemical properties but of 

 different atomic weights. Dr. .Aston concludes that 

 the atoms have weights which are all integral 

 numbers, and that observed fractional atomic weights 

 are du« to mixtures of two or more isotopes. 



Electric furnaces in which the heating element 

 consists of a metallic wire or strip are now exten- 

 sively used both in the workshop and in the labora- 

 tory, and possess the advantages of cleanliness, accuracy 

 of temperature control, and small working costs. A 

 special form, possessing several new and useful 

 features, and intended mainly for workshop use, is 

 announced by .Automatic and Electric Furnaces, Ltd. 

 The furnace-chamber or muffle is flat on the top and 

 bottom, while the sides are rounded, this shape being 

 specially suited, to secure complete contact with the 

 heating coil wound on its exterior. In order to pre- 

 vent destruction of the coil through inadvertent over- 

 heating, a cut-out consisting of a loop of wire melting 

 at 960° C. is inserted in tf^e furnace. The heating 

 current passes through the loop and is cut off by the 

 melting of the latter, a rc<l pilot-lamp being simul- 

 taneously lit up to serve as a warning. Several sizes 



