uf the External Coatings of Seeds. 283 



upon any other i)rinci[)l(; than that I liavc above demonstrated? 

 The facts here shown are completely fatal to Dr. Gray's later 

 conclusions. The outer tlcshy seminal tunic in Euonymus cer- 

 tainly cannot be held to be a true arillus, as generally supposed, 

 but an arillinc ; in all resj)ects it is analogous to the fleshy coat 

 in Magnolia. If, therefore, the outer seed-coat of Euonymus be 

 acknowledged to be an arilline, or extraneous tunic distinct from 

 the growth of the real ovular integuments, so, pari passu, must 

 the scarlet coating of Magnolia be of the same nature : as the 

 previous condition of the ovule, and the subsequent structure of 

 the seed, are precisely similar in both cases, it follows as a neces- 

 sary rule, that the nature and origin of these seminal tunics 

 must come under the same category. 



In order to render the fact of this expansion of the placentary 

 sheath more palpable, I will proceed to demonstrate the manner 

 in which the raphe is developed under other circumstances, 

 which, though of ordinary occurrence, has been only cursorily 

 alluded to on a former occasion* — I mean, a branching raphe. 

 In this instance, the ovule before impregnation exhibits the 

 usual appearance seen in the first of the preceding figures, with 

 a simj)le lateral band enclosing the nourishing vessels; but 

 during its growth into a seed, either these, or other vessels 

 springing from the same source, spread themselves in dichoto- 

 mizing branches over the whole surface of the integument, 

 which is generally described as the testa ; but if such vessels 

 were really existing within the true testa, they would necessarily 

 have entered its tissues at the point b of the chalazal extremity, 

 whence s})readiug through its mesoderm, they would infallibly 

 show, in the ripe seed, the appearance in the adjoining figure 6, 



Fig. 6. Fig. 7. 



7, J> 



where a is the hilar point of attachment. On the contrary, I 

 have invariably found, that the starting-point of this distribution 

 of the vessels is at the opposite hilar or micropylar end of the 

 seed, as in figure 7, where a and b refer to similar points. The 

 development of this form of raphe is easily accounted for upon 

 the explanation 1 have given, and the following figures will ex- 

 * Linn. Trans. .\.\ii. p. 8S. 



