ti6 James Small. 
Biology. The occasional use of the involucre in the wind- 
dispersal of the whole fruit-head was noted by Hildebrand (V, 30). 
The protective function and the hygroscopic movements of the 
bracts are mentioned by Haberlandt (III, 15, p. 552) and Yapp 
(IV, 98), and form the subjects of short papers by Rathay (12) and 
Gressner (6). The latter also mentions the hairs which seal the 
closed involucre in the bud. These had been noted previously by 
Hildebrand ( loc. cit.). 
Biometrics. As the result of extensive statistical observations 
Shull (13-14) concluded that the number of involucral bracts shows 
a seasonal variation due to nutrition differences. The number 
showed a continuous change, beginning low, increasing for 2-6 days 
to a maximum and then decreasing for the rest of the flowering 
period. Shull maintains that the modes of his curves have no 
connection with the Fibonacci series. 
Harris (7) deals with the correlation between the number of 
involucral bracts and the number of fruits developed. 
Mutations. Sudden variations in the involucre are recorded 
in a few cases, e.g., partial loss of prickles in Xanthium Woottoni 
(IV, 87, pp. 139 and 152), and a peculiar spreading and variation in 
number of the bracts in Vernonia rubricaulis (15). Various 
abnormalities, chiefly apostasis, are mentioned by Cramer (IV 
21) and Masters (IV, 52). 
B. Origin and Development of the Involucre. 
The capitulum is a racemose inflorescence, usually with a more, 
or less flattened receptacle on which the florets are crowded. It is 
frequently considered to be a spike of sessile or nearly sessile 
flowers with the peduncle shortened. There is, however, another 
possibility which is usually overlooked—the capitulum may arise by 
the abortion of the pedicels in a racemose umbel. The two 
processes are very similar, but in the former the pedicels abort first 
and then the peduncle, while in the latter the stages are reversed. 
Although the reduction is similar the products would tend to present 
certain differences. 
In a spike the flowers are usually spaced well apart and the 
flowers are frequently large. Any reduction in the length of the 
peduncle would naturally be gradual, and the bracts of the flowers 
would tend to be reduced by the same causes (decrease in food 
supply, illumination, etc.), and, therefore, in the same measure as 
the flowers. The receptacle would rarely, if ever, become flat, and 
