Mr. J. Miers on the Bignoniacese. 155 



pellicle, as Prof. A. DeCandolle concluded (Prodr.ix. 142, note). 

 The many radiating lines that diverge from the coriaceous discoid 

 portion are formed by the plicature of the laminae at those parts, 

 evidently from the effect of pressure, for they contain no vessels 

 of any kind, and have no connexion with the raphe. The discoid 

 centre, where these laminae are not conjoined, is quite hollow 

 and opake, forming a distinct cell lined with a quantity of opake 

 white cellular tissue; and the next integument lies in the centre 

 of this space, filling about two-thirds of its diameter. This inter- 

 mediate integument is somewhat opake, transversely oval, with 

 a deep emarginature in its summit, reaching nearly to the cen- 

 tre ; and in the bottom of this sinus the chalaza is clearly seen : 

 at the base is another emarginature, of less depth, from the 

 bottom of which proceeds a tube of the length of the whole 

 tunic, which extends downwards to the basal hilum through the 

 coriaceous portion of the winged integument, to the sides of 

 which this tube is agglutinated by solid deposits ; and in this 

 manner it forms a semiseptum in the lower part of the coria- 

 ceous cell, which thus becomes bimarsupiate at its base. The 

 third integument that immediately invests the embryo is thinner 

 and of the same form as the intermediate coating, only that a 

 very short sac surrounding the radicle occupies the place of the 

 long tube. The embryo is of the same transversely oval form, 

 and consists of two very fiat foliaceous cotyledons, deeply cordate 

 at base, with a very short obtuse radicle in the sinus that occu- 

 pies half its length : in the summit is a similar emarginature 

 that reaches the centre, thus cleaving the cotyledons almost 

 in two, and leaving a very short space between the two emar- 

 ginatures, so that the embryo appears almost to consist of four 

 cotyledonary lobes. The semiseptum above mentioned coincides 

 with the line of the simple raphe that is imbedded in the ventral 

 discoid lamina of the outer winged tunic, but has no connexion 

 with it more than is due to the subsequent agglutination of the 

 parts. The raphe, starting from the hilum, follows the course 

 mentioned, imbedded in the outer lamina, until it arrives oppo- 

 site the apical sinus of the embryo, when it suddenly pierces 

 its way thi'ough the coriaceous deposits of the tunic, and imme- 

 diately communicates with the chalaza. 



It is thus seen that the embryo in the Bignoniece usually con- 

 sists of a very short radicle seated in the basal sinus of two very 

 deeply lobed cotyledons, which are again 2-lobed at their sum- 

 mit : this form is subject to several modifications, one of the 

 most remarkable of which is found in Argylia, where the embryo 

 appears cleft only at the base, being quite truncate at the sum- 

 mit; but this scissure extends from the base to the very apex. The 

 embryo is thus divided into four equal lobes, greatly flattened, 



11* 



