[ 200 ] 



XXIII.— DESCEIPTION OF AN APPAEATUS FOE DEMON- 

 STEATING SYSTEMS OF CLASSIFICATIONS, • ETC. 

 By ALFEED C. HADDON, M. A., Professor of Zoology, 

 Eoyal College of Science, Ireland. 



[Eead, January 21, 1884.] 



A METHOD of classification which would indicate the affinities of 

 organisms to one another has for a long time been th^ aim of 

 biologists, and they have more or less succeeded, by arranging the 

 names of the forms whose affinities they desired to illustrate in a 

 given manner, and connecting them by means of lines. 



The older zoologists, for instance, were content with arranging 

 the names in a continuous sequence, thus forming a linear series or 

 a classification in one dimension of space. 



It was soon found that such an arrangement by no means indi- 

 cated the real affinities, nay more, tended to retard a true conception 

 of the animal kingdom as a whole ; so recourse was had to arrang- 

 ing the names in a dendritic manner on a flat surface, the length of 

 the lines, their position, and their various ramifications illustrating 

 the supposed relative position of the animals themselves. This 

 constituted a classification in tivo dimensions of space. 



Some zoologists, however, have felt that even this method was 

 insufficient to express their views; but, so far as I am aware, this 

 want has not yet been supplied ; and it has occurred to me that 

 what was really required was a system which would more accu- 

 rately render in space that well-known figare of speech — the 

 Grenealogioal Tree — and that the facts of the case could be more 

 closely represented by a solid method of classification — in other 

 words, by a classification in three dimensions of space. 



Unfortunately such a system requires a special apparatus, and 

 is not adapted for ordinary representation on plane surfaces ; but 

 the apparatus I propose is very simple, merely consisting of a series 

 of glass plates placed vertically over one another, leaving a small 

 space between each, and a number of small movable blocks. 



The exact method for supporting the glass plates is of no con- 



