831 



the raphe lateral, as in Ranunciilacea3 and Cucurbitacea?, and subse- 

 quently pressed downwards as in the former case. 



3. He maintains that a single pendulous ovule with the raphe next 

 the placenta is the only true pendulous ovule, with the exception of 

 pendulous campylotropal and amphitropal ovules with the foramen 

 (and subsequently the radicle of the embryo) turned away from the 

 placenta. 



4. He conceives that one or two erect ovules with the raphe tunied 

 away or obliquely away from the placenta result from pendulous 

 ovules pressed upwards by the elongation upwards of the cavity of 

 the ovarium ; and adduces in support of this opinion the pendulous 

 ovules of Geissoloraa contrasted with the erect ovules of Penaea, the 

 erect ovules of Calytrix compared with the pendulous ovules of the 

 neighbouring families, and the pendulous ovules of Calycereae com- 

 pared with the erect ovules of Compositge, provided further observa- 

 tion should substantiate his belief that in the last-named family the 

 raphe is really turned away from the placenta. Such ovules he would 

 term spurie erecta, in contradistinction to the opposite case, to which 

 Sprengel has applied the term spurie pendula. 



5. He considers that a single ovule erect with the raphe lateral is 

 a horizontal ovule spontaneously growing or pressed upwards by the 

 corresponding development of the ovary; in proof of which he cites the 

 fact that Trianthema micrantha has two seeds in a horizontal posi- 

 tion, with the radicle lateral, while T. decandra has two erect seeds 

 one above the other, with the radicle also in both cases lateral. 



6. He considers one or two erect ovules with the raphe next the 

 placenta (which seems general in Endogenous plants, and is frequent 

 in all the divisions of Exogenous) as for the most part truly erect ; 

 although this position may sometimes be derived from horizontal 

 ovules pressed upwards or spontaneously growing erect, the funiculus 

 becoming at the same time twisted so as to bring the raphe into rela- 

 tion with the placenta. 



Mr. Clarke then proceeds to illustrate the importance of these cha- 

 racters in a systematic point of view, as regards different families 

 usually regarded as nearly related. He states that Thymeleae differ 

 from Laurinese in having the raphe next the placenta, and that the 

 same difference of relation occurs in Sanguisorbeae and Amygdaleae. 

 In all the Urtical Orders with pendulous ovules the radicle is next the 

 placenta, or if campylotropal the direction of the curvature is equiva- 

 lent, and the radicle of the embryo is turned away from the placenta ; 

 while in the Chenopodal Orders with pendulous ovules the radicle is 



