3i6 AFFINITIES [ch. 



suggest that this envelope and possibly the * spathes ' also 

 may be, not foliar organs at all, but structures more closely 

 comparable with such outgrowths from the floral axis as the 

 membranous cup which surrounds the essential organs in the 

 male and female flower of the Poplar, and, more remotely, with 

 the arillus of the seed of the Yew-tree. The Potamogetonaceae 

 are characterised throughout by the absence of a perianth; if 

 Naias be descended from the Pondweed stock, any * perianth ' 

 which it possesses must have been acquired de novo and hence 

 it is highly improbable that any such organ which it might form 

 would be morphologically a normal perianth^. In Althenia^ the 

 ' perigonium ' of the male flowers and the scarious ' bracts ' 

 associated with the female flowers, and, in Zannichellia^ the 

 membranous cup surrounding the female flowers im.c. in Fig. 

 45, p. 70) may also be mere cupules of no phylogenetic import- 

 ance, but in the case of the female flowers of these genera, the 

 possibility that we are dealing with spathes is not excluded. The 

 variable occurrence of the floral envelopes in different sections 

 of the genus Naias^ harmonises well with the theory that they 

 are recently acquired organs of no historical significance. On 

 this view we are absolved from making the forlorn attempt to 

 recognise in this genus the counterparts of all the organs which 

 characterise the typical Angiospermic flower. 



The Lemnaceae have long been regarded as connected with the 

 Arum family. MorethaneightyyearsagoSchleiden^ propounded 

 the view that Pistia and Lemna both belong to the Aroideae and 

 are related to one another. He showed that in Pistia the axis is 

 abbreviated instead of being elongated as in most Aroids, and 

 he regarded the River Lettuce as forming, in this respect, a 

 transition to the Duckweeds^. Certain dissimilarities between 

 Lemna and Pistia have, however, been emphasised by Koch ^. 



The aquatic family Pontederiaceae (Farinosae) is somewhat 



1 This follows from the ' Law of Loss ' which will be discussed in 

 Chapter xxviii. 



2Schleiden, M.J. (1838I). 



3 See also Arber, A. (1919^); and p. 74. * Koch, K. (1852). 



