33 



only to the fruits of other genera in the same family prevents 

 one from taking for the outer integument of the seed. 

 Here the nodus of the attachment of the calyx is separated 

 by a very distinct internodium from that nodus where the 

 bases of the four carpellary leaves are attached, and from 

 whence proceed the partial seminal pedicels through the 

 appropriate apertures of the carpels, which I call spermo- 

 pyle* into the cavity of those carpels. In the genus Cyno- 

 glossum two evident internodia may be observed, the one 

 between the attachment of the calyx and corolla and that of 

 the carpellary leaves, the other between these and the partial 

 seminal pedicels. t 



The latter internodium may be observed much more 

 conspicuously in the Echinospermum which M. Dvigoubski 

 in his Flora of Moscow distinguishes as a section of Myosotis 

 with prickly seeds. Here, as Schrader has admirably repre- 

 sented in the Echinospermum Lappula, the axis presents a 

 four -sided pyramid with excavated concave surfaces. At the 

 base of each of these longitudinal concavities is a bunch of 

 fibres forming the connection between the carpellary leaf and 

 the axis, and representing its petiole. At a considerable inter- 

 val higher up there proceeds from the same concavity another 

 bunch of fibres forming the partial seminal pedicel. And thus 

 it is evident that the partial seminal pedicel does not pro- 

 ceed from the margin of the carpellary leaf, and that the in- 

 ternodium which separates the two abovementioned branches 

 of fibres is that which authors generally call the placenta, 

 with this difference only, that it is not enclosed within the 

 cavity of the carpel, but is what I call the extracarpellary 

 spermophorum. Precisely the same structure is observable in 

 Echinospermum deflexum Lehm, and probably in all the 

 species of the genus. In the Labiatae, and in all those Bora- 

 ginese described as having nuces basi imperforatse, the struc- 

 ture of the carpellary leaf is very similar, as I have already 

 said, to Mirbel's campylotropous seeds, and on this account 



* From the Greek word (nrep/j-a seed, and ttvXt) a gate, an entrance cor- 

 responding to the other term I have adopted of blastopyle in lieu of the 

 micropyle of Turpin. 



t See de Asperifoliis Linnaei Commentatio auct. H. A. Schrader, fig. 

 d. c. — A third internodium in this family is formed by the continuation of 

 the axis which gradually runs into the stigma, without producing any more 

 lateral appendages. 



