10 PROCKKDINGS OP THE ANATOMICAL AND ANTHROPOLOGICAL 



by which the particular form could at once be known. Often it was 

 difficult to distinguish between arches and loops, and these doubtful 

 patterns were written A 1 or L a , according as the tendency of the 

 pattern was more to the arch or loop type respectively. 



So also there were patterns written L w and W. 



Again, complex patterns may be written thus : L vy , where the r 

 indicates that the loop is an invaded one, and the y denotes that the 

 core of the loop is eyeletted. 



In the secondary classification it was often necessary to count the 

 number of ridges in a loop. This was done beginning at the innermost 

 ridge and counting to the outermost of a loop. When the centre 

 ridges in a loop were uneven the middle one was taken. If the 

 core was a staple the limb of the staple farthest from the delta was 

 selected. 



With regard to the outermost ridges there are two types : one 

 where a single ridge divides to form the boundary lines of the pattern, 

 and the other where these boundary lines are formed by two ridges 

 diverging. 



By this system of classification a very accurate description can 

 be got, and a finger-print directory may be formed by means of which 

 the peculiarities in the patterns of each set of digits are recorded. 

 Gallon shows how such a directory can be formed. 



Since Gallon's book on "Finger Prints" was published in ISJI-J 

 a criminal investigation system has been instituted in Britain, which 

 includes the taking of finger-prints of all criminals according to the 

 methods recommended by Galton. The Government adopted this 

 because they found that 



1. Finger-prints cannot be altered. 



2. Finger-prints are not liable to be obscured except in persons 

 engaged in particular kinds of manual work. 



3. Finger-prints can easily be recorded. 



4. Finger-prints can be used for purposes of proof in a court of 

 law. 



In the Report of the Commissioner of Police of London for 1903, 



