No. 482] CATALOGVIM: MI SI-IM Si'l-ri .][ i:\S 81 



spomls to that which I have (h'sionatiM i the ' 1 ii incut Cala- 

 logue,' lacking the method of cross iiKh'xiiii: liy ( Npaii mciits, 

 marginal tabs, and colored cards (when <h'^ii;il.le The ;i ,") 

 inch cards used by Wray are too small, while ilie wiitin-- of two 

 sets for each specimen nearly doubles the clerical work involved 

 in the use of an Accession, I )e|)ai-tnieni . and Kefcrcnce catalogue 

 as noted in the following ])ages, since by the latter method a large 

 number of specimens are usually transcrihed on a sintjle dc})art- 

 ment and reference card. Space for certain valnahle data is like- 

 wise omitted by Wray. something unavoidable however with the 

 small card. 



From the first it seemed evident that the card catalogue arranged 

 in unit cabinet sections would furnish the most satisfactory solu- 

 tion of the problem. The value of such a system had long ago 

 been recognized in connection with library and general l)usiness 

 methods, where it rapidly displaced lh(> bulky vohimes formerly 



tages resulting from the tis(> (.f the card system are obvious since 

 (1) the retjtiired <lala are prcM-iUcd in a c<.inpact and ea>ily acces- 

 sible form; .2. the capacity i> unlimited. um'Icss ,vc.mb caii be 



4 X (), and s, inches, and although other vizc-, could l)e made 

 and UM.,I. it is well t.> adopt one of thv.v. inasnuich as the regtilar 



machinery Uscd is particniarly adapted for the three sizes. ' 'Vhv 



