80 A NATUnALISrS WAKDEBINOS 



fiowera, while the longitiidi|lal section on the opposite page 

 shows the complicate system of galleries — some of them 

 papillated— inhabited by the anta. 



Obserring the ants often employetl in carrying out whitish 

 particles, I at first conjectured that the irritation of their 

 digging out a dwelling must have induced the swelling of the 

 bulb J and, curious to see the iTiodm operandi of its commence- 

 ment, I decided to raise a few of them from seed. This turned 

 my attention to their flowers and fruit. The flowers are pro- 

 duced iu deep spine-protected pits on the axis surmounting 



YOPKB PtAST OP MTBMEOOniA TrBEEOfiA. 



the bulb, and are remarkable for the extreme rapidity with 

 which the cycle of their functional changes are performed. 

 The pellncid white flower appears, and is followed by an 

 orange, watery fruit, whose seeds ripen and often germinate in 

 the little pits where they grow, all within the space of thirty- 

 six hours. 



Some years later Dr. Burck, of the Buitenzorg Gardens, most 

 kindly showed me specimens and microscopic slides illustrating 

 some interesting observations * he had made on these flowers : 

 that the corolla segments rarely ojjen (though a slight touch 



* These have since been publiKlicd \n ihe * Anujitlca du JaHiti Botanique du 

 BiiitenKorg/ vol. iv., p, 16, 



