TRANSACTIONS 0? SECTION fi. 



PU 



inlpressionS are also heeded to equalise the size of each division to as nearly as 

 possible 10 per cent, of the total number of records. 



In the first instance the records are classified into two divisions bj the ridge* 

 pattern on the thumb, according as this happens to be A, an arch or either form 

 of loop ; B, a whorl. 



Each of these two groups is broken up into four smaller divisions by the 

 pattern on the forefinger according as it is («) an arch ; (f>) a loop with the mouth 

 opening to the left ; (c) a loop with the openini,'- to the right ; (d) a whorl. 



Of the eight divisions thus obtained no further subdivision is necessary of six 

 groups— namely, of a, b, and d of the A division, and of a, b, and c of the B 

 division. 



In the A division, further subdivision of the r. group by the middle and ring 

 fingers ia necessary. _ This is done by separating the cases where there are loops on 

 each of the four digits from those in which there is any other combination of 

 patterns on the thumb, mid, and ring fiugers. Thus we obtain five groups in this 

 iirst division. 



Again, in the B division — namely, the cases in which there is a whorl on the 

 thumb — subdivision has to be resorted to in the d group when there is a whorl on 

 the forefinger as well as on the thumb. This is done by separating the cases where 

 the middle finger bears a whorl from those in which there is an arch or either form 

 of loop. Thus we obtain five groups in this second division. 



We have now got ten groups, which are of approximately equal size, except 

 the a group of the B division — that in which the thumb bears a whorl and the 

 forefinger an arch, which is considerably smaller than the others, but this cannot 

 be obviated. In any given number of individuals there will always be some 

 in whom one or more of the four fingers used in classification have been damaged 

 from some cause or other, so that the pattern of the ridges is undecipherable, or 

 one or more fingers of either hand have been lost partially or completely ; espe- 

 cially is this the case amongst the labouring classes in manufacturing districts or 

 towns. In adult criminals such cases amount to about 6-5 per cent. In any s\'stem 

 of classification it is necessary to provide for such cases either by making a separate 

 group of them or by placing them with some other group as may be found more con- 

 venient. If the latter course be adopted, and they are added to a of the B group, 

 it will be brought up to the level of several of the other groups. 



The following is the scheme of classification reduced to tabular form, and the 

 approximate percentage of the records in each of the ten divisions is indicated 

 by the figures given in the lowest line : — 



A. 



B. 



Right Thumb 

 Right Forefinsrei 



/\ /\ 



9-3 13-5 



\ 



WW 



121 



7-7 



O 

 110 



o 



With the above ten divisions of the finger impressions worked in conjunction 

 with and secondary to classification by measurements, sufficient power is available 

 to enable the records of a large number of criminals to be easily manipulated. For 

 example, if only four measurements be used in the tripartite classification, which 

 is universally followed, we have 81 divisions; by the employment of this decimal 

 division by finger-prints, a total of 810 divisions are obtained ; while if five 

 measurements be taken, as adopted in England, which gives 243 divisions, by the 



