<0 



ON THB FERTILIZATION 



taD of umber coloured Mri. and detached stamens filaments, anthers, 



\mp ot union formidable obstacle 



• Sn ii ^e * they encounter the much more formidable obstacle pre- 

 **ZZ L "g of ostiolar bracts. Here they set to work mueh more methodically, 



*" u - 11L* rfcht and left at random, but confining their operations to the centre 



; tSu/Tr ugh 'which they eventually succeed in tunnelling an evenly cylindrical 

 iTel lo «it looL-ly filled with soft brown debris and strugghng insects (Plate IV fig. 24). 



It is. ot course, difficult to deter„,ine the precise length of the intervals elapsing between 



„ ie J .K^rance of the receptacular fluid and the emergence of the male insects into the 

 i* and torn the latter to their exit from the perforated ostiolum. They probably vary 



con- dernbly in different instances, especially the latter, which must necessarily be directly 



.ted to the number of male insects present. The following are the only data regarding 

 tbi point whi h are attainabl 



I A mature receptacle was taken in the morning. At noon male insects were 



beginning to emerge from the ostiolum in large numbers, and at 5 p.m. 

 females were beginning to appear 



i A large mature receptacle was taken in the morning. A few male insects emer- 



(1 from ho ostiolun during the course of the day and on the following 



morning and were then followed by multitudes of females during the course 



of the forenoon. Here the emergence of the females was apparently delayed, 



due to the defective number of males present to clear the way for them. 



A largo mature receptacle was taken in the morning. At 4 p.m. one male had 



actually emerged and others could be seen struggling deep down in the 



ostiolar tunnel. By 7 a.m. of the following morning numerous males had 

 emerged, and females were emerging and flying off in a continuous stream, 

 and by 9 a.m. emergence had ceased. 

 4. A mature receptacle was laid open by transverse division, and was found to contain 



a lar^e number of free female insects and a comparatively small number of 

 males, who were already hard at work demolishing the male flowers and 

 1 ginning to attack the ostiolar plug. The ostiolar half of the receptacle 

 was put aside under a bell glass with the open surface of the receptacular 

 cavity upwards. Twenty-four hours later perforation of the ostiolar plug 

 had be Q completed, and a considerable number of male insects had 

 emerged from the orifice and were lying about beneath the under surface 

 of the specimen. As there was no evidence to show that any had escaped 

 over the cut edges of the receptacle, the purely reflex character of the 

 proeess of tunnelling was strikingly demonstrated. 



Having effected their exit, such of the male insects as escape immediate seizure 

 th predatory ants which are usually on the wait for them fall down from the receptacle and 

 ^ soon die. Under normal circumstances the winded female insects besrin to emerge 



o'" vv ~ » 



fi in the receptacles shortly after the completion of the ostiolar tunnel, appearing for a 

 time in company with the males, and, after these have all emerged, continuing to issue forth 



alone for a considerable time. But all receptacles are not alike in their insect content 



In normal cases the male insects, although by no means so abundant as the females, are 

 vet present in sufficient numbers to secure rapid and thorough perforation of the ostiolar 

 plug; m others they are still present, but in undulv small numbers, so that there is delay 



