Journal of Applied Microscopy. 591 



supply a few cards for the one need, a few different ones for another special 

 requirement. 



This then led to the final step in the reorganization of the work. In the 

 spring of 1 898 a complete printing office was established in a house close to the 

 offices* and special equipment was secured so as to permit the printers to inter- 

 change certain symbols without removing the frame from the press. Thirty-five 

 cards are printed at a time on large sheets, which are afterwards cut up and 

 passed through the punching machine. They then come into the store-room 

 and when a sufficient number have accumulated, pass on to the distributing room. 

 Here the sorting is done by younggirls by means of a multiple check system, which 

 removes all chance of error. 



Now that the system is in perfect working order, it is interesting to note 

 how the single intellectual operation of assigning a given number to an article in 

 a scientific journal causes the card representing that article to fall into the same 

 identical place in card catalogues scattered all over the civilized world, from Chris- 

 tiania to Cape Town, and from Honolulu to Vienna. Indeed, no manuscript is pre- 

 pared for the type-setters. Each paper is studied carefully, many are read entirely, 

 and a simple note is slipped into the book indicating the contents, the new 

 species, etc., and showing by simple signs the various uses that are to be made 

 of a given title. The type-setters are chosen from different nationalities, so that 

 all the modern languages can be dealt with. According to linguistic aptitude, 

 the type-setters are given the publications themselves and compose the references 

 according to certain definite rules of citation. In certain simple cases the 

 specialist charged with the indexing would write merely a number on the cover 

 of the pamphlet, 57.90 (74.7) for instance, and as a result, the announcement 

 would be sent out that a new genus of bees had been described from New York 

 State. This information would be given on a card satisfying all the rules of 

 citation ; it would also appear in its proper place among the Apidae, and 

 yet all the subsequent operations would have been carried out by persons 

 entirely ignorant of the very existence of the group Apida;. 



The principle of classification is that in every case two aspects come into 

 play and are given a varying value in the classification, according to the use to 

 which a given card is to be put. Thus, in the case given above, the card would 

 be printed for one subscriber in the order Apidae — New York, for another in 

 the order New York — -Apidae. This interplay of two or more aspects gives the 

 whole system an adaptability which had long been sought but never before 

 attained. At present there are 170,000 such groups in the scheme of classifica- 

 tion, not to mention the use that is made of the alphabetical arrangement of 

 genera within a given family. At first sight this statement would seem to fore- 

 bode a terrible complication in the symbols, but this is far from being the case. 

 Indeed, the users of the bibliography would be themselves the first to be 

 astounded by this bit of statistics. Properly speaking, the symbols play the 

 same part that do the page numbers of a book, the alphabetical key representing 

 the index, and the methodical key the table of contents. 



*They have since been brought under one roof in the new and more commodious quarters of 

 the Concilium. 



