M. Dumas on the Laiv of Substitutions. 501 



Pratt's Philidor. Lond. 1825, 8vo. (I pass over many Ma- 

 gazine articles, and brief notices of the knight-problem, 

 in numerous works on chess ; as well in various collections 

 of a general character, as Guyot's Recreations, &c.). 



Ciccolini. Del Cavallo degli Scacchi. Paris, 1836, 4to, 

 pp. 70, and 20 large plates. (This volume treats not only 

 on the problem of the knight's move over the common 

 chess-board of 64 squares ; but also on the larger field of 

 100 squares; as well as the circular board of 64.). 



La Corso del Cavallo per tutti gli Scacchi della Scacchiere. 

 Bologna, 1766, 4to. 



Collini. Solution du probleme du Cavalier, &c. Manheim, 

 1773, 8vo. pp. 62. 



Essai sur les problemes de situation. Rouen, 1783, Svo. 

 pp. 74. 



Der Roesselsprung, &c. bei Eduard Billig. 1831, 24mo. 

 pp. 64. 



Warnsdorf, H. C. von. Des Roesselsprunges, &c. Schmal- 

 kalden. 1823.- 4to, pp. 68. 



Lettre adressee aux auteurs du Journal Encyclopedique sur 

 un Probleme de I'Echiquier. Prague, 1773. 



Dollinger, 24 verschiedene Arten den Springer, &c. Wien, 

 1806, 8vo. 



Netto das Schachspiel : — And numerous other German writers. 



LXXVI. Memoir on the Law of Substitutions, and the Theory 

 of Mechanical Types. By M. Dumas. 



[Continued from p. 447.J 



Mechanical Types. 



A FTER having verified in a manner which satisfied my own 

 -'-^ conviction, the existence of certain chemical types, I tried 

 the general application of this theory of types to all the 

 known series produced by substitution, and last year at the 

 School of Medicine I made this system of ideas the basis 

 of my lectures. 



But always constant to the experimental progress of the 

 science, and wishing never to swerve from it, I asked myself 

 whether it was necessary to class together bodies having the 

 same formula, produced by substitution, but essentially dif- 

 ferent in their most prominent chemical properties ? 



I said, the bodies produced by substitution are divided into 

 two different classes: some evidently belong to the same 



