406 Bertha Rees .- 



even after prolonged treatment with the acid, remained un- 

 swollen in spite of the fact that a considerable amount of the 

 outer coat had obviously been removed. The reason for this can 

 be seen by reference to the figures of A. lophantha. Fig. 80 a 

 shows the structure of the untreated seed coat. It was covered 

 externally by a distinct cuticle 0.015 mm. in thickness ; there 

 were two layers of palisade cells instead of one, as is more 

 usual. The cells were of varying lengths, those of the deeper 

 row formed an undulating surface, over which those of the 

 outer were moulded in such a way as to form a level surface 

 on the exterior of the seed. The walls of the outer palisade 

 cells were entirely cuticularised. and those of the inner also 

 for some distance from the outer end, the remainder of the 

 wall was of ordinary cellulose, and the cell contents were proto- 

 plasmic. The cuticularisation ended abruptly, and in the stained 

 sections its limit was marked by a sharp line running across 

 the cells. The lumina of the cuticularised portions of the cells 

 are coloured black in the figure as they were otherwise difficult 

 to define, but as far as could be seen they were quite empty. 

 Fig. 80 b is a section of a seed which had been treated with 

 acid for several hours, and which remained unswollen in water, 

 although a good portion of the testa had evidently been re- 

 moved. It can be seen that the outer palisade cells were almost 

 corroded away, but as the inner cells were still intact, it was 

 impossible for water to enter the seed. Fig 80 c is a section of 

 a seed which had swollen in water, and shows that as soon as 

 the walls of inner cells were corroded away as far as the 

 openings of the lumina, water could then run into the cell 

 cavity, pass through the inner cellulose wall and so enter the 

 seed. 



('anna indica. — This was the only hard seed I examined that 

 had not a definite cuticular membrane covering the surface. The 

 coat consisted of palisade cells, the walls of which were cuti- 

 cularised except at the inner ends. Running transversely near 

 the middle of the cells was a definite line which did not appear 

 to have any morphological existence, or even to mark the boun- 

 darj' between layers of varying cuticularisation, but which was 

 apparently the result of an optical effect. The lumina were 

 narrow, with two dilatations, a slight one at the outer end, and 



