180 JOURNAL, BOMBAY NATURAL mST. SOCIETY, Vol XXVIU 



mined by actual measurement) waxy-white flowers. Fig. 1 shows the 

 inflorescence (Infl.) inside the partiaUy opened bulb of specimen No 1. 

 Fi- 2 shows the scape (Sc.) (in specimen No. 1) which assumed quite a 

 curious shape. It deviated so much from the type, that, it became solid, 

 fleshy and stunted (its length being only 6-25 cm. whereas a normal scape 

 is 30-60 cm in length) and it bore the inflorescence along the whole of 

 its left side instead of bearing it on its top. The whole inflorescence thus 

 developed was wrapped up by a membranous covering (Memb.) with 

 prominent parallel nerves, A part of this membrane was found adnate to 

 the left side of the scape. This covering seemed to be nothing but a 

 modified form of a spathe. In the fourth specimen, which has not been 

 depicted here, 1 noticed two waxy ovate-lanceolate fleshy structures, 

 diflering in shape from all the other scale-leaves (Si.), adpressed to the 

 tiny inflorescence (Infl.) inside the bulb. Fig. 3 shows the two kinds of 

 inflorescence (Infl.) met with in specimen No. 2, in which a group of 

 flowers or fascicle (b.) arose directly from the stem below and a small 

 umbel (a), partially hidden by ' b. ' was borne by the irregularly z.gzag 

 solid scape (Sc) slightly twisted to the left. Except the basal part of a 

 withered normal scape (Sc\ seen also in Figs. 1 — 3) no trace of a fresh 

 scape is seen in Fig. 4, all the flowers (Fl.) having taken their origin 

 directly from the stem (St.). Dissections of the flowers (Fl.) from each 

 of the above specimens revealed the fact that, although etiolation had 

 taken place, owing to the partial exclusion of light, still, the perianth and 

 the sporopliylls were developed quite up to their normal size and shape. 

 Except in specimen No. 1 (in which the anthers appeared to have dehisced)' 

 the anthers in all the other specimens were found to contain scanty pollen 

 grains. Tue ovaries (' B ' and ' D ' in Fig. 5) were provided with either a 

 long-styled (D) or a trifid sub-sessile (B) stigma (intermediate forms being 

 noticeable in some of the flowers) and generally three compressed ovate 

 ascending and minutely pitted ovules (* F ' and ' E ' in Fig. 5) in each cell. 

 As the ovules, particularly those in specimen No. 1 were quite tough 

 (unlike functionless ovules which are easily compressible) and as in some 

 of the anthers the pollen-sacs were almost empty, it seemed probable that 

 at least some of the flowers were self -fertilised. Here, it is obvious, that, 

 no cross-fertilisation could have taken place at all. 



Histological difl'erences between a normal (Sc.^) and an abortive scape 

 (Sc. in Fig. 3 ) were no less marked. The following were the main points- 

 worth noticing : — 



1. The epidermis (Ep.) of the abortive scape (Fig. 7) was thicker than 

 that of the normal scape (Fig. 6) and was provided with comparatively 

 larger cells, which were not of uniform size and shape throughout. 

 Whereas the cuticle (Cut.) was uniformly thickened in the normal scape, in 

 the abortive scape it was distinctly thicker on the outer or ' dorsal ' surface 

 than on the inner or ' ventral ' surface. Stomata (Stom.) were often present 

 in the epidermis (very clearly seen in longitudinal sections of the epidermal 

 region) of the abortive scape, whereas no stomata were generally found in 

 that of the normal scape. 



3. The ring of sclerenchymatous cells (Scl. ) in the abortive scape 

 (Fig. 7) in which the vascular bundles (V. b.) lie scattered, was composed 

 of cells having walls thicker than of those in the normal scape (Fig. 6). 



3. The vascular bundles (V. b.) in the abortive scape (Fig. 7) were 

 numerically less than those in the normal scape (Fig. 6), but proportionately 

 greater for the area supported by them. 



4. Two distinct groups of large vascular bundles (V.^ b.^ ) were found 

 developed in the ventral area near the centre of the solid abortive scape 

 (Fig. 7), whereas in the normal scape (Fig. 6) the central portion was hollow. 



