174 Proceedings of the Royal Society of Victoria. 



last census. These returns fi'om the municipalities date back to 

 about September last, the figures being adjusted in the Govern- 

 n)ent Statist's office, and brought up to the end of the year. 



That the figures thus obtained are liable to considerable uncer- 

 tainty must, I think, be admitted. It may be interesting, there- 

 fore, at a time when the country is taking stock of its resources, 

 to test the correctness of these estimates by an altogether 

 independent method. 



Tlie birth rate in any community is a tolerably fixed quantity. 

 Taking the three years, 1890-92, as an example, it appears that 

 in Victoria the birth rate averaged 33-24 per 1000, with extremes 

 of 33 '60, and 32 "54. The rate does vary, of coyrse, but only to 

 a slight extent within any short period. The probability, there- 

 fore, is, that if the rate seems to vary greatly, within a very short 

 period, there is some error in the figures used, the most likely 

 source of error being in the population figure, the number of 

 births, being matter of almost exact record. It is this test of the 

 birth rate, or, rather, the variation in the births recorded, which 

 I propose to apply for arriving at an estimate of the population 

 of Melboui'ne in the present year. For this purpose I will take 

 the first half of several successive years for com[)arison. 



Tablk 1. 



Showing the Number of Births Eegistered in Melboiiruo aud Suburls 



for 1890-94. 



