ON PRECIOUS METALS IN USE AS MONET, 223 



method might be fairly accurate ; while the third method, relying on the exports 

 and imports of a single year (or a biennial period), would be fatally inaccurate. 

 Moreover, there exists in the third method a difficulty from which the first is 



fi-ee as to the incidence of the exports for any year on the different strata of the 



circulation. This difficulty is considered by Mr. Edgeworth in_ a Me7norandum on 

 Jevons' method which he has prepared for the British Association. 



(9) The careful computations made by Mr. Kimbal, Director of the Mint at 

 Washington (' Reports,' 1886 et seq.), inspire confidence. In the case of the United 

 States there is, Jirst, a pretty accurate initial datum for the year 1873— on the 

 occasion of the resumption of specie payments. Secondly, the statistics of exports 

 and imports are perhaps specially deserving of confidence (cf. above, Kemark 6). 

 Mr. Kimbal seems to regard them as satisfactory. Thirdly, some attention has been 

 paid to the amount of coin used annually in the arts, though hardly so much 

 attention even yet as to command complete confidence. 



(10) It has been suggested by Mr. Edgeworth, in the Memorandum above 

 referred to, that Jevons' method may admit of improvement. Fu-st, the ratio on 

 which the reasoning turns might be determined with greater precision by a more 

 careful and methodical sifting of the data. This correction by itself would not 

 come to much in view of the extreme incorrectness of the other datum (above. 

 Remark 6). But it is further submitted that in some countries the coinage for a 

 particular year or short period may have sufiered particularly little from exporta- 

 tion. In fact, something of the sort appears to have occurred in France according 

 to the statistics given by M. de Foville in the papers above referred to. 



To revert to the metaphor above employed, the monetary census is effected 

 according to Jevons' method : 1, by ascertaining from samples the proportion of, 

 say, three-year-old infants to the total population; 2, in the absence of accurate 

 data for the number of three-year-old infants now alive, the number of infants born 

 three years ago. Now suppose, as M. de Foville shows grounds for supposing, that 

 a particular generation escaped in an extraordinary degree the maladies to which 

 infancy is liable ; then it might be the best plan to operate, not on infants, as 

 usually taken for granted, but upon that favoured generation, already, perhaps, 

 adult or even superannuated, always supposing that the other datum, the ratio 

 determined by sampling, can be ascertained for the favoured generation with ade- 

 quate precision. But this appears to be doubtful, owing to the circumstance that the 

 generations appearing to enjoy extraordinary longevity consist each of comparatively 

 few individuals. Now a deserved suspicion attaches to the use of small numbers in 

 statistics. At any rate, it is not contended that the correction would have any 

 application to the United Kingdom. 



(11) In comparing the use of sovereigns in England and one-pound notes in Scot- 

 land it would be desirable if possible to make an allowance for the difference which 

 may exist in the use of half-sovereigns in the two countries ; and for other differences. 



(12) As a specimen of the results which may be attainable, we submit an 

 estimate obtained by Mr. R. H. Inglis Palgrave with respect to a particular town. 

 This town contains about 6,000 inhabitants. It is a railway junction, and also 

 stands upon a river by which seagoing vessels can reach the place. _ There is also 

 a considerable inland navigation connected with it. The surrounding district is 

 agricultural. The town may be regarded as above the average in prosperity for 

 the agricultural districts. In this town Mr. Palgrave estimates, as the result of 

 investigations set on foot by him, that there woidd be on an average, taking one 

 time with another, about 4,000^. in coin in the hands of the population ; that is to 

 say, about 15s. a head in gold and silver. This estimate does not include the 

 specie held by the banks. The proportion of gold to silver is about 4 : 5. Mr. 

 Palgrave is more certain about this proportion than about the absolute arnounts. 

 He thinks the latter may be exaggerated, as the country districts surrounding the 

 town are to a certain extent included in the supply of specie mentioned. If this 

 estimate were a fair average it is quite possible that, including the country popu- 

 lation referred to in the districts surrounding the town in question, the whole 

 specie in active circulation may not exceed 125., or even 10s., a head. Further 

 investigation on the point is highly desirable. 



