780 REPORT— 1886. 



opinion that there was no sufficient reason for the present great excess cost, and that 

 much more tonnage of new ships ought to be produced for the money expended. 



He pointed out that it had been decided, after great inquiry and deUberation, 

 that at least 20,000 tons of new war shipping ought to be added yearly to our nav y 

 under normal circumstances to keep up its strength properly, that in 1869 to 1872, 

 under the administration of Mr. Childers, carrying out the recommendations of 

 Mr. Seely's Committee of 1868, 22,387 tons (of which 17,114 tons were iron- 

 clad) had been yearly added at a much less cost than a smaller tonnage before or 

 since ; and that he (the author) never could discover sufficient reason why these 

 data resulting in the decision as to 20,000 tons yearly, were departed from, and only 

 about 17,000 to 18,000 tons yearly added. 



The consequence was panics arose, and our navy had to be suddenly increased 

 at an enormously excessive cost — indeed, at any cost almost that those who had war- 

 ships chose to ask. 



He concluded by stating if he had carte, hlanche he believed he could again bring 

 about similar results to those of 1869 to 1872. 



WEDNESDAY, SEPTEMBER 8. 

 The following Papers were read : — 

 1. Proportional Mortality. By Baldwin Latham, M.Inst. C.E., F.G.S. 



The author of this paper describes a method of very correctly arriving at the 

 proportional death-rate of a district with great facility by the use of ' constant ' or 

 basic numbers. These ' constant ' numbers were obtained at census periods by 

 ascertaining the average death-rate at such periods, and then adding a quantity 

 equal to ten times the proportion of births to one death found in the same period. 

 As one example he gives the case of the combined disti'ict of Birmingham and 

 Aston, where the average death-rate at the census periods of 1861, 1871, and 

 1881 was 22-.31, and the average proportion of births to deaths in that period, 

 when multiplied by 10, was 17-69, giving, when added together, a constant number 

 of 40; so that if ten times the proportion of births to deaths was deducted 

 from 40 it would show the proportional death-rate in any period as compared 

 with the average at the last three census periods. There might be cases in 

 which a modification in procedure would be necessary in arriving at the propor- 

 tional death-rate, as in those places and years when the deaths exceed the births, 

 in order to arrive at a result in such a period, which now fortunately seldom occurs 

 in this country, but which was common years ago. When the births and deaths 

 were equal it was equivalent to deducting 10 from the constant number. If there- 

 fore when the deaths exceeded the births 10 was deducted from the constant num- 

 ber, and the remainder multiplied by the ratio by which the deaths exceed the 

 births, it would give the proportional mortality. 



A second and more correct method was described in this paper, in which the 

 constant number was exactly equal to twice the average death-rate at the census 

 period, the proportional mortality being arrived at by deducting a quantity equal 

 to the proportion of births to one death when multiplied by a number obtained 

 by dividing the average death-rate by the average proportion of births to deaths. 

 Under this system, in Birmingham and Aston districts, the constant number became 



44-62, and the multiple 177^0 = 12'6; so that the proportion of births to one death 



multiplied by 12'6, and deducted from 44-62, would give the proportional mortality 

 in this district at any time. 



