148 DEVELOPMENT OF THE OVULUM 
It is at once distinguished from the teguments by its never 
forming a perfectly closed sac. 
Mr. Lindley has supposed that in Scepacee it is acces- 
sory to fecundation, and that it subsequently becomes the 
fleshy part.of the seed. I must however say, that this view 
which is erroneous, is founded on the description of the 
plant by Roxburgh. 
M. Mirbel however, describes the arillus of Nymphza 
alba, as becoming subsequently a perfectly closed sac, al- 
though from his drawing it would appear obviously to be an 
extension of the placenta. See his work Pl. 6.—Fig. 15, 
p. 60. 
Arillus often eatable. . . 
Inner tegument, Tegmen, Secundine. 
I have just alluded to the opinion of Mirbel as to the 
almost universal existence of this covering, that in the 
instances cited by authors to the contrary it had become 
united to the primine. 
This author is of opinion that it never exists as a covering 
to the seed, but that it sooner or later becomes intimately 
united with the outer membrane, to such a degree that its 
existence is not demonstrable.  . 
For my own part, observations have led me to suppose that 
this tegument is in some cases ab origine single. Hence the 
difficulty of ascertaining whether it is testa or tegmen be- 
comes materially increased ; because the tegmen being first 
formed, it would seem to follow, that in all cases in which 
such singleness can be demonstrated, the coat must be tegmen. 
Hence too this coat may be vascular, contrary to the usual 
opinion. 
On the subject of this coat Mr. Brown iiobis remarks, 
"the inner membrane of the ovulum, however in general 
appears to be of greater importance, as connected with 
fecundation, than as affording protection to the nucleus at a 
more advanced period. For in many cases, before impreg- 
