352 Dr. Everett on a new Proportion-Table. 



so placed as to have their columns parallel. In rapid working, 

 this adjustment can be made with fair accuracy at a cursory 

 glance ; but in order to enable the operator to make it with the 

 greatest possible nicety, both the pieces are ruled with lines at 

 right angles to the columns, and these lines are placed about a 

 hair's breadth wider apart on one piece than on the other, so that, 

 on the principle of the vernier, one line can always be found 

 which will extend continuously across both. 



The arrangement described will be best understood by inspect- 

 ing the specimen which accompanies this paper*. The material 

 which has been selected for the pieces is Bristol-board, on which 

 the numbers and divisions are printed from copper plates. They 

 are distinguished by the names A and B in the sense already 

 explained. The card B contains twenty columns, each 8 inches 

 long, which if arranged in one line would form a logarithmic 

 scale in duplicate of the length of 13 feet 4 inches, precisely 

 similar to one of the scales on the common sliding-rule greatly 

 magnified, and having a proportionately greater number of divi- 

 sions. Card A has ten columns, each 16 inches long, and con- 

 tains precisely the same matter as B, only differently arranged, as 

 already explained. In using the cards, B is laid uppermost, in; 

 such a position that its columns alternate with those of A, the 

 latter being visible through the openings or slits which are cut 

 between the former. The divisions on A are on the right side of 

 its columns, and those on B are on the left, extending to the edge 

 of the slits, so that the divisions in those columns which work 

 together are in close juxtaposition. In adjusting the cards, it 

 will be found that a movement left or right to the extent of one 

 column-breadth is equivalent to a movement of 8 inches up or 

 down. 



The largeness of the scale is such that, without undue crowd- 

 ing, the space from 1 to I'l is divided into a hundred parts; so 

 that every number from 1000 up to 1100 has a division of its 

 own, the last and nearest of these divisions being about ^ °f an 

 inch apart. On the best sliding-rules of the ordinary construc- 

 tion, this interval, which is here divided into 100, is only divided 

 into 5. From 1100 to 2100 there is a division for every even 

 number, from this point to 4800 a division for every fifth num- 

 ber, and thence to the end of the scale a division for every three- 

 figure number. 



Again, whereas in the ordinary rule only the first figure is 

 printed, the first three figures are here printed as far as 479, and 

 thence to the end of the scale the first two figures, 



Some persons can never use the common sliding-rule success- 



* [The specimen in question was a copy of the " Universal Proportion- 

 Table" published by Messrs. Longmans. — Eds. Phil. Mag.'] 



