THE LEAF. 



flowers, these being more or less imperfect in their 

 organization. The main feature of the phenomenon 

 consisted in the presence of transitional structures 

 between bulbil scale-leaves and perianth-leaves, and 

 even stamens, for in some cases both of these two last- 

 mentioned organs had dark-coloured swollen bases 

 which are obviously of scale-leaf nature. It would 

 appear that the axis of the bulbil elongates somewhat, 

 the lowermost scales remain unaltered, while the 

 uppermost become changed wholly or partially into 

 floral organs. The ovary in such flowers was either 

 absent or excessively undeveloped. The plant was 

 otherwise uninjured and normal (PL XIII, fig. 1). 



Now the bulbils of Liliacese have hitherto been 

 regarded as, from the morphological point of view, 

 reduced vegetative shoots, owing to their resemblance 

 to miniature bulbs, and to the fact that, after falling 

 to earth, they grow out into a vegetative shoot. But 

 a little consideration will probably show that they 

 can hardly be such, and this for more than one reason. 

 Firstly, the production of lateral vegetative branches* 

 is unknown among the herbaceous members (the vast 

 majority) of the order ; hence it is highly unlikely 

 that such branches, in whatever form, should suddenly 

 appear in this isolated way among a few scattered 

 representatives of the order. Secondly, bulbils in this 

 order of plants do not occur as a rule in the position 

 of vegetative branches, but practically always in that 

 of flowers, viz., at the base of the inflorescence or the 

 highest part of the stem, a short way below the latter 

 in the axils of bracts or leaves transitional between 

 bracts and foliage-leaves, as in Lilium bulbiferum and L. 

 tigrinum; never, or very rarely, in the axils of the typi- 

 cal vegetative leaves. t In Allium sps. they constantly 

 and often normally replace flowers. In the yellow 

 star of Bethlehem (Gagea) they are known to do the 



* Miss Green, in her paper, draws attention to these points. 



f The fact that bulbils occasionally occur in this position can be under- 

 stood when the further fact is recognized that flowers have been known to 

 occur in the very same position. 



VOL. I. 11 



