490 STUDIES IN SEEDS AND FRUITS 



NOTE 23 (p. 205). 

 The rupturing of the coats of the germinating seed of Entada scandens. 



IT cannot be said that the splitting takes place in the weakest point 

 in the coverings. As a matter of fact, it is in the thickest portion, 

 namely, at the hilar scar, that the rupture occurs. If we dry the shell- 

 like coverings of the germinating seed, we find that the thickness in and 

 about the scar is about 3 millimetres, whilst the thinnest portion at 

 the opposite border of the seed is only half as thick. Usually the shell 

 thins away as we leave the base of the seed where the rupture takes 

 place. In fact, there is all the appearance of a special provision for 

 adding to the thickness and strength of the coats at this point, the 

 micropylar opening being obliterated. Mere weakness in the cover- 

 ings at any particular place does not result in their rupture at that spot. 

 This is well shown in the fact that although all the seeds of this plant 

 were filed half through the shell at the sides and top to procure 

 germination, in no case did the splitting occur there when the process 

 began. I may add that the splitting of the coats in the swelling seed 

 is not necessarily followed by germination, as indicated by the protrusion 

 of the radicle, and that therefore the question of any pressure from this 

 cause does not arise. 



NOTE 24 (pp. 208, 219). 

 The Hura difficulty. 



THE difficulty with the seeds of Hura crepitans is that their 

 coverings, which attain their maximum development and weight long 

 before the kernel reaches its full size, lose weight as the kernel proceeds 

 with its growth, the albumen and the embryo which make up the 

 kernel growing well together as the seed matures. As the result of 

 a number of observations on the later stages of maturation, it was 

 ascertained that while the seed added but little to its size and weight, 

 a considerable change occurred in the relative proportion of the coats 

 and the kernel. 



When the pre-resting seed is fully mature, the coats have already 

 lost considerably in weight, whilst the kernel has just reached the 

 maximum of its growth. This is the condition that had to be repre- 

 sented in the table on p. 219, which contains the data for constructing 

 the shrinking and drying rdgime of these seeds as compared with other 

 plants ; and of course the full shrinkage of the seed-coats is not there 

 indicated. This is clearly brought out in the comparison of the two 

 regimes given below, the first for the mature pre-resting seed, where 



