367 



out food iuduceiueut. A loiigitudinul section of the upper portion of 

 the pistil will show the style with the stijiiuatic tube, which at this time 

 conuMunicjitcs with tlie ovarian cells. Xow, Tr<'lease hsis sliown that the 

 stigmatic liipior is not uectarian, but that the slight amount of uectar 

 associated with the llower is secreted in pockets formed by the partitions 

 that separate the tliree cells of the pistil, and which open externally 

 near the style by a contracted pore from which the nectar is poured 

 through a capillary tube to the base of the pistil. The accompanying 

 illustiation (Fig. 65) renders this more intelligible, a being a longitud- 

 inal section through the center of aiX)istil, .showing the septal gland (^), 

 the duct (rf), and the outlet at the base; b a cross section of the pistil 

 about the middle, also showing the duct (<?), and gland {(i) ; c, a more 

 enlarged cross section of the nectar ai)i)aratus; e showing more fully 

 the structure of the septal gland, while h is a longitudinal section of 

 the top of the pistil, through the lobes, showing how the stigmatic tube 

 {s) connects with the ovarian cell (o c), o being the ovary, /the funicu- 

 lus, p the placenta, and/ y flbro- vascular tissue. 



These interesting facts, which I have fally veritied, show that nectar- 

 feeding insects seek it not about the stigma, but at the base of the 

 stamens or of the petals, whether within or without. In short, the 

 nectar in these Yucca flowers has no value in pollination, and Pronuba, 

 in collecting the pollen and transferring it to the stigma^ finds no food 

 compensation, a conclusion which is confirmed by a study of the minute 

 structure and internal anatomy of the moth, whi<'h indicate that the 

 tongue proper, though strongly developed, has to a great extent, if not 

 entirely, lost its function as a sucking organ, and that the alimentary 

 canal is i^ractically functionless, being aborted before reaching the anus. 

 This defunctionization, if I may use the term, of important structures 

 has not proceeded so far in Pronuba yuecaseUa as in P. maculata, which 

 pollinizes Yucca whipplei. Those not familiar with the structure of 

 Lepidoptera will hardly appreciate the modifications to which I shall 

 allude, however, without the preliminary statement that the tongue in 

 Lepidoptera consists of two distinct parts (maxillai) which are more or 

 less concave on their inner side and united at the borders of the con- 

 cave portion by certain locking arrangements to form between them 

 the sucking tube. Now, while in most cases the two parts may be re- 

 laxed and separated by force, in nature they are never so separated, 

 while the tij) of the tongue is more or less acuminate and the two parts 

 here very firmly united. 



In Pronuba yurcaseUa I had often noticed that the two parts became 

 separated, and in fact were almost always separated toward the tip, 

 thus suggesting the loss of function as a sucking organ, but otherwise 

 the tongue is strongly developed, and, with the exception of the weak- 

 ness of the locking arrangement, not particularly abnormal. In Pro- 

 nuba maculata, however (Fig. 00), the two parts of the tongue are but 

 very feebly united, and often more or less disconnected, and are actu- 



