410 Dr. Schunck on Rubian audits Products of Decowposition. 



that they are not parallel, and then applying the corollary above- 

 mentioned, showing that through a point in one of them a straight 

 line would pass so as to be parallel to the other, and therefore 

 the two given straight lines would meet, which is a contradiction. 



In the last two instances it might perhaps have been assumed, 

 that if the given straight lines were not parallel, then through a 

 point in either, a straight line might pass so as to be parallel to 

 the other, in the sense of the assumed definition. This would 

 have shortened the arguments, but is not so strictly logical. We 

 are entitled to assume an example of anything defined, but are 

 scarcely entitled to assume anything more than that one ex- 

 ample, or some other exhibiting the assumed property under the 

 same relation to other things. 



The equality of the perpendiculars can be deduced from the 

 same hypothesis, but this scarcely belongs to the same group of 

 relations, inasmuch as it depends on and establishes the relations 

 of straight lines as to length. 



The following points will perhaps be found to be established 

 by the considerations which have been offered : — 



1. That definitions are the primary hypotheses on which all 

 conclusions rest, and ought to be chosen so that these conclusions 

 follow readily and clearly from them. 



2. That in employing definitions or converses of theorems as 

 criteria, the source of certainty is the reduction to a contradic- 

 tion of some statement opposite in character to the criterion. 



3. That all argument must necessarily, as to its nature, de- 

 pend on the character of the hypothesis, and that logic fails only 

 when this necessity is not attended to. 



LXVII. On Rubian and its Products of Decomposition'^, 

 By Edward Schunck, F.R,S, 



Part Il.f 



Action of Alkalies and Alkaline Earths on Rubian. 



IN the first part of this paper I have described, in general terms, 

 the effect produced on rubian by alkalies. It is only the fixed 

 alkalies that are capable of effecting the decomposition of rubian. 

 Ammonia produces no perceptible change in a watery solution 

 of this substance, except that of altering its colour from yellow 

 to blood-red. This blood-red. colour remains unchanged even 

 after long-continued boiling, and the solution still contains 

 rubian, for on supersaturating the ammonia with acid, the solu- 

 tion again becomes yellow, and not the slightest precipitate is 

 produced. The fixed alkalies, on the other hand, act very 



♦ From the Philosophical Transactions for Parti, 1853; having been 

 read June 10, 1862. 



t Part I. was given in the 3rd volume of this Journal. 



