212 
CONTRIBUTIONS TO BOTANY. 
the tissues of the fleshy coating, from the base to the apex, when 
it suddenly breaks through the endoderm, and arrives at a distinct 
foramen in the chalazal extremity of the nut, which I have called 
a diapyle, through which it passes, and soon becomes lost in the 
thickened areole or chalaza of the inner integument. 
On the other hand, we may conclude either that the nut ori- 
ginates in a development of the secundine, and that the tegmen 
i-esults from the tercine or membrane of the nucleus (which is a 
very rational inference), or that the nut is of independent origin 
(which is still within the range of probability). Indeed there is 
nothing to prevent the possibility or probability of the formation 
of an intraneous coat between the usual tunics in the growing 
seed, subsequently to the closing of the micropyle of the ovule. 
We frequently meet with a copious deposit of loose cellular tissue 
between the testa and tegmen, or even between the latter and 
the albumen ; and this deposition may become consolidated into 
a distinct membrane, or hardened by solid deposits : wherever 
there exists a budding-point, this may at any time take place. 
We have proof of the actual formation of such a tunic between 
the primine and secundine in the Gnetacece^ •, and this, if applied 
to the case of Magnolia, points to a far more probable cause for 
the appearance of the hard tunic in the seed, than tlie impro- 
bable suggestion of Schleideu and Gray, to which I have just 
referred. 
Schleiden, although entertaining so many ideal fancies re- 
specting developments, gives nevertheless a somewhat true ac- 
count of the present state of our knowledge on this subject : he 
says (/. c. p. 427) — “ A greater confusion than that which pre- 
vails in the theory of the seed-coats is scarcely conceivable ; the 
most heterogeneous things are throwm together under one name ; 
thoroughly identical ones are placed in difierent kinds of organs ; 
and there is nothing for it, if we would not make greater con- 
fusion, but to cut the thread and begin over again : the epidermis 
of the seed is sometimes described as testa, sometimes as arillus ; 
seed -membranes are introduced where no true integuments 
exist.” 
My object in bringing together the facts related in the fore- 
going pages is to show that we may always proceed with some 
degree of certainty, if w'e never lose sight of positive, invariable 
landmarks. Beyond doubt the surest course is to trace the de- 
velopments from the eaidiest growth of the ovule to the final 
perfection of the seed ; but where this cannot be done, notwith- 
standing the variable texture and condition of the tunics deve- 
* According to the interesting details given by Griffith, Lindley’s ‘ Ve- 
getable Kingdom,’ p. 233, where analytical figures demonstrate the fact. 
