290 BOTANICAL GAZETTE [October 



condition. This shows that it is largely, if not entirely, the inner 

 layer of cells of the coat, with perhaps the acellular layer, that is 

 responsible for the failure to grow. 



The methods given by Tunmann (31) in his late work were 

 followed in the microchemical tests. The following reactions 

 justify the conclusions given in the description of fig. 1. After 

 treating sections of the coat 48 hours with cuprammonia and then 

 staining with ruthenium red, a, c, e, and part of b became red. 

 Treatment for 48 hours with cuprammonia followed by 2 per cent 

 ammonia dissolved all the section except d and part of b. Sudan 

 III stained only the inner wall of the inner layer of cells. Chlorzinc 

 iodide stained a violet and e blue. Chromic acid in 10 minutes 

 dissolves everything except d y but after 16 hours d still remains 

 intact. One half-hour's heating in 5 per cent hydrochloric acid 

 dissolves a considerable part of e, especially at the hilum end and 

 the small adjacent end of the coat. One-half hour's heating in 

 2 . 5 per cent sulphuric acid followed by treatment with IKI stains 

 e blue; IKI alone gives no reaction in any part. Treatment with 

 copper sulphate for 48 hours, followed by 2 per cent ammonia, 

 dissolved the greater part of c; 2 per cent ammonia alone brings 

 about no visible changes. Sections in water for 10 days showed 

 the contents and most of the cross-walls in c dissolved, and the 

 plano-convex cells at the large end swollen to about twice their 

 former size. 



These tests indicate that there is no cellulose in the coat. The 

 blue coloration of e with chlorzinc iodide is probably due to the 

 presence of hemicellulose rather than cellulose, as indicated by its 

 insolubility in cuprammonia, its ease of dissolution by 5 per cent 

 hydrochloric acid, and its ready transformation to amyloid by 

 2 . 5 per cent sulphuric acid. 



It is evident that the coat is largely composed of pectic sub- 



admixture of hemicel 



in e. The dissolution or marked 



and cross-walls of c, and the plano-convex cells at the large end of 

 the embryo, indicate that these are made of the type of pectic 

 substances included under the terms pectic, parapectic, and para- 

 pectosic acids. Slight chemical transformation of these substances 



