eb. Mar. 



April May 



June July 



Aug. Sept. 



Oct. Nov. 



69-5 



60-6 



527 



55 9 



63-7 



1743 



1631 



1478 



1600 



1716 



1796 



1645 



1511 



1565 



1698 



242 G. H. KNIBBS. 



ning the results for pairs of months. In this way we obtain 

 the two upper lines in the following table : — 



Dec. Jan. 



Temp. Fahr. 70-0 



Suicides per 10 ,„ 



10.000 1832 



Calculated 1805 



This correlation is very approximately expressed by 615 

 + 17t (where t is the temperature Fahrenheit), a formula 

 which gives the results in the last line. Here it may be 

 pointed out that Australia differs very remarkably from 

 Europe in this respect, viz., that the range of temperature 

 throughout the year is decidedly smaller in Australia. Thus 

 a mean for the various countries of Europe gives the range 

 between the averaged hottest and coldest months of the 

 year about 33° Fahr., while for Australia the range is only 

 about 19° i.e., but little more than one half. We thus have: 



Ranges in 



Temperature. 



Suicide Frequency. 



Europe 



33° Fahr. 



517 



Australia 



19° „ 



253 



That is to say, the variation in the suicide frequency on 

 the whole corresponds very closely to the range in temper- 

 ature, being strongly marked where the temperature 

 differences are strongly marked. It is evident from this 

 that large temperature fluctuations tend to bring about 

 large changes in the frequency of suicide. 



7. Mode of suicide.— In a relatively small population the 

 number of suicides for individual years by any particular 

 mode of self-destruction is naturally variable, nevertheless 

 there is greater uniformity than might have been antici- 

 pated a priori. The statistics have been computed for the 

 years 1907 to 1910 inclusive. 



These results shew that the mode of suicide is very 

 regular. The relative frequency of any particular mode is 





