145 



sight by the extraordinary lengthening of the secundary 

 peduncles and the circumstance of their originating, with only 

 a few exceptions, from the ligulate florets. The secundary 

 headlets consequently protrude a long way or droop far 

 downwards, the peduncle varying between 1 and 7i c.M. 



I hâve chosen to examine one of them in détail (comp. 

 fig. 10) : peduncle 2>\ c.M., at the base enclosed by the pri- 

 mary ligulate flower. Higher up, about halfway, a semi- 

 amplexicaulous bract springs from it, close under the 

 headlet a second, a little smaller one. It is not probable 

 that thèse bracts are to be compared with carpels : firstly on 

 account of the ordinary absence of the pistil in the ligulate 

 flowers of Kudbeckia, but secondly especially on account of 

 their non- apposition. They might rather be compared with 

 proliferous flowers. 



The headlet itself consists of a few whorls of ligulate 

 flowers and a compact mass of tubular ones, which are 

 closed. The other ligulate florets are constructed on the 

 same pattern, but as has been already said vary the length 

 of their peduncles from 1 to 1\ c.M. 



Most of them spring from a small ligulate corolla, two 

 only from tubular florets towards the centre. 



The number of cases observed of médian floral prolifl- 

 cation in the Compositae isnotgreat; Mas ter s mentions 

 as such Hieracium, Cirsium, Bidens, Calendula, Spilanthes, 

 Carthamus und Goreopsis. Penzig cites the said gênera 

 and has not much to add, e.c. Tragopôgon (II p. 101). 



On account of this circumstance I thought it worth 

 while to augment the séries with Rudbeckia. 



2. Raspberries on a bifuegate thalamus. 



In the summer of 1906 I was regaled with raspberries 

 which were extraordinarily broad and on eating proved to 



