Representation of Propositions and Reasonings. 



17 



board. Cut out all the subdivisions by following the lines of 

 the different figures, after the fashion of the children's maps 

 which are put together in pieces. The corresponding step to 

 shading out any compartment would then be the simple re- 

 moval of the piece in question. We begin with all the pieces 

 arranged together, and then pick out and remove those which 

 represent the non-existent classes. When every one of the 

 given premises has thus had its turn, the pieces left behind 

 will indicate all the remaining combinations of terms which 

 are consistent with the data. I have sometimes found it con- 

 venient, where the saving of a little time was an object, to use 

 a contrivance of this kind. There is no reason to give a 

 drawing of it, since any one of the figures we have hitherto 

 employed may really be regarded as such a drawing. 



Again, corresponding to Prof. Jevons's logical machine, the 

 following contrivance may be described. I prefer to call it 

 merely a logical- diagram machine, for the reasons already 

 given ; but I suppose that it would do very completely all 

 that can be rationally expected of any logical machine. Cer- 

 tainly, as regards portability, nothing has been proposed to 

 equal it, so far as I know ; for though needlessly large as 

 made by me, it is only between five and six inches square and 

 three inches deep. It is intended to work for four terms ; and 

 the following figures will serve to show its construction : — 



]. II. 









r-i n n n L 











c 

 c 

 c 



W/ 



Z/> 





■S.Y 



X 









d \ 













n 

















D 

















: 

















: 

















i — 











ULJU 



l_l 









The first figure represents the upper surface of the instru- 

 ment. It shows the diagram of four elipses, the small irregular 

 compartment at the top of them being a representative part 

 of the outside of all the four class-figures ; that is, this com- 



Phil. Mag. S. 5. Yol. 10, No. 59. July 1880, 



