122 W. H. GILBURT ON THE 



radially — i.e., a line drawn through the centre of each would extend 

 from centre to circumference of capitulum. The two carpellary 

 leaves preserve their individuality for a time, until sufficient height 

 has been attained to form the bifid portion of the pistil, then 

 growth commences round the whole cavity of the floret, enclosing a 

 small hollow space at the base, thus forming the future ovary — a 

 condition shown in two stages in Figures 5, 6, and 7, Figure 7 

 being a transverse section of fig. 6. 



Growth now proceeds rapidly above the ovary, so that the cavity 

 which was first at its upper part soon occupies its centre. 



While this is taking place, we observe at the base of the now elon- 

 gated pistil — a growth of tissue which, when fully developed, appears 

 as a cushion or collar, encircling the style — a cellular and generally 

 glandular body, secreting nectar, and known as the disc (Fig. 8). 

 Whether this is its function in the species under consideration, I 

 have not been able satisfactorily to determine. 



There is one fact concerning this part of the flower which, so far 

 as I know, has not hitherto been observed — at least, it is not 

 referred to in any papers, nor shown in any figures that I have 

 seen — i.e., the presence in comparatively large numbers of perfectly 

 developed stomata. Whether they are accompanied by the usual 

 intercellular spaces I am not able at present to say, the cell-walls of 

 the disc being so extremely thin that it is difficult to decide whether 

 two cells are in contact or not. The stomata however are so 

 numerous, that on a disc — when only -^ of an inch in diameter — 

 I have counted as many as 54. 



We come now to the development of the ovule, concerning which 

 in some particulars authorities are not agreed. 



In the Composites we always find an unilocular ovary, and the 

 ovule erect and anatropous, one question in dispute being whether 

 the ovule is a development of the floral axis and also its termina- 

 tion, or a lateral development upon it; some authorities, amongst 

 them Sachs, holding the latter view. Kohne appears to regard it 

 as a development from one of the carpels, but from which he was 

 not able to say. It is difficult, however, to see how either can be 

 supported in the species we are dealing with, as the first indication 

 we have of the ovule is an uprising of the floor of the ovary, the 

 centre of which is slightly conical, as shown in Fig. 10. True, in 

 the earliest stages the base of the funiculus is somewhat inclined to 

 the carpel which is nearest the centre of the capitulum ; but 

 this fact seems to favour the hypothesis of the axial character of 



