September 23, 1921] 



SCIENCE 



277 



factory. A combination of this method with 

 the first one above is often helpful, and has 

 been used in Chile. 



Third, " the corrections at the standard sta- 

 tion may be multiplied by a factor propor- 

 tional to the daily range (or to the difference 

 between midday and evening observations) at 

 the station to be corrected." This method de- 

 pends upon properly correcting for any differ- 

 ence in daily range which may exist between 

 nearby stations, as, for instance, between hill 

 and valley stations where the daily ranges may 

 be quite different. But care must be used to 

 see that the climates of the two places are 

 similar, for the daily amplitude of variation 

 may be the same, but the shape of the curve 

 different. This was found to be the case in 

 applying the method to Los Andes (valley) 

 and Santiago (coastal slope), Chile, in which 

 the amplitude is the same but the sharp night 

 minimum at Los Andes was markedly differ- 

 ent from that at Santiago. 



The fourth, and last, method is that in which 

 " all direct use of standard stations may be 

 avoided and the reduction to true mean based 

 on considerations connected with the general 

 phenomena of the diurnal variation of temper- 

 ature." This makes no use of standard sta- 

 tions but bases the whole system upon a nor- 

 mal diurnal curve which may be expressed by 

 the Fourier series 



Tk = T + a-^ sin (tk + A^) + a, sin (2<ft+\4,), 

 in which the coefficients of the first and second 

 harmonics may be calculated as follows: 



a^ — —0.72 + OA4: (M — m), 

 a- = -I- 0.54 -f 0.08 (M — m). 



This method has been thoroughly tested and 

 has given satisfactory results, but is chiefly 

 valuable when the stations are so situated that 

 the other methods are not applicable. 



A. Buchan favored the mean of the daily 

 maximum and minimum as the most satisfac- 

 tory expression of true mean temperature, and 

 the view has become widely accepted. But 

 there are objections, not only from the instru- 

 mental standpoint — for maximum and mini- 

 mum thermometers are more likely to develop 

 systematic errors than are dry-bulb thermom- 



eters — but also because the ratio of this mean 

 to the true mean is dependent upon location, 

 time of year, and cloudiness. For that reason, 

 the author advises meteorologists in English- 

 speaking countries to make less use of the 

 mean of the daily extremes and more of suit- 

 able^ combinations of the observations at fixed 

 hours, such as the first formula above. 



C. LeRoy Meisinger 

 Washington, D. C. 



SPECIAL ARTICLES 



ON THE ELIMINATION OF THE X-CHROMO- 



SOME FROM THE EGG OF DROSOPHILA 



MELANOGASTER BY X-RAYS 



Some of the results of the investigation of 

 the biological effects of X-rays point to a 

 specific effect of the rays on the nuclear mat- 

 ter of the dividing cell and especialy the germ 

 cell. 



1. Actively growing tissue whether normal 

 or pathological is the most susceptible to X- 

 rays. 



2. It has been found comparatively easy to 

 sterilize a number of different species without 

 apparently otherwise injuring them. 



3. Perthes'- has found that X-rays have a 

 destructive effect on chromosomes in the ova 

 of Ascaris megalcephala. 



The experiments to be described were per- 

 formed with a view to determining if X-rays 

 by affecting the X-chromosome could disturb 

 the inheritance of a sex-linked character. 

 Wild type (red-eyed) females of Drosophila 

 melanogaster, homozygous for red-eye, were 

 X-rayed soon after emerging from the pupa 

 with a dose just under the sterilization dose 

 and mated to white-eyed males. Sisters of the 

 rayed females were used as controls and mated 

 to white-eyed males. The white eye color is 

 sex-linked and recessive to the red and when 

 there is no non-disjunction the offspring in 

 tht first generation of a cross between a 

 homozygous red-eyed female and a white-eyed 

 male are all red-eyed. 



In all, four experiments have been per- 

 formed. In three of these, numbers 71, 76, and 

 77, thirty-five virgin females, homozygous for 



1 Perthes, Vent. Med. WocJi., 30, 1904. 



