130 A. H. Graves, 



horizontal direction, around the rhachis, closely following its sur- 

 face, until finalh' they nearl}- meet, as in PL IX, fig. 49. 



The pistils, occurring between the thecae in a group of four, form 

 a diamond-shaped pattern arranged in a transverse position on the 

 rhachis. Each pistil develops, in the mature flower, into a cyl- 

 indrical structure tipped with a sessile peltate stigma and con- 

 taining a single ovule. 



As regards the number of the pistils, I have never found it to 

 vary ; but Ascherson (1889), who divides Ruppia maritima^ his only 

 species of the genus, into three subspecies, sa3's in his generic 

 characterization, "4 (selten bis 10)". Roze (1894, p. 479), indeed, 

 makes the number of pistils the main specific difference between 

 R. maritima and R. rostellata, alleging that the former has eight 

 pistils and the latter four. In this vicinity, however, Ruppia maritima 

 seems to have always four. 



Eichler (1875, p. 89), Irmisch (1851), and other older authorities 

 do not hesitate to allude to the subfloral leaves of Ruppia (the pair 

 at the base of the inflorescence) as the spathe leaves. This homol- 

 ogy is not used, however, by more recent authorities. For reasons 

 already adduced, I have applied to them the name of subfloral 

 leaves. As to the spathe, however, it seems quite probable that 

 a rudiment of it is represented in the floral scale leaf. As has been 

 shown (pp. 99-102), its manner of origin, development orientation, 

 etc. all connect it ultimately with the flower, and although its mor- 

 phological origin is clearl}^ identical with that of a vegetative scale 

 leaf, its relations to the spadix indicate a spathe-like nature. Since 

 the flower of Ruppia represents a much reduced type, we should 

 expect to find such reduced structures here.^ 



Floral Development 



With these remarks on the general morphology of the mature 

 flowers, I shall describe briefly the developmental stages of the 

 young flower, from the time when it is first recognizable until about 

 the period when the archesporial cells first appear. From that 



^ I have already indicated (p. 101) that Ii-miscli overlooked this spathe- 

 like floral scale leaf, and I had been unable to find any reference to it 

 in the literature. At the last moment, however, I find that Grriffith (1851 

 I, pis. 257, 258 and 259 ; II, pp. 196 and 198) figiu-ed and described this 

 structure, considering it a true spathe, and explaining its origin and 

 development essentially as I liave done. 



