138 



their bases and just outside the circle 

 of filaments a depression containing in 

 some flowers one, in others two small 

 openings, of about the diameter of pins 

 (c', Fig. 2), these lead into one or two 

 slender tubes which run from this point 

 to the bulb already noticed on the side 

 of the pedicel, and are, indeed, the 



Fig. 2.— Cross-section of the upper part of the pedicel 

 and nectary before the cavity of the latter divides 

 into two, magnified 17 diameters. 



doors through which insects must reach 

 the nectar secreted in the lower part of 

 the tubes. 



To see the difference in the structure 

 of the upper, non-secreting, and the 

 lower, secreting, part of these tubes, it 

 will be necessary, with a very sharp ra- 

 zor, to cut off thin sections of the flower- 



Fig. 3.— Part of the wall of the nectary at the same 

 point, magnified 200 diameters to show the simi- 

 larity between the epidermal cells on the outside 

 (2) and those lining the cavity (1;. 



stalk and adhering spur at different 

 points, and to study these with the mi- 

 croscope. Such a section, taken very 

 near the upper end of the tube, and be- 

 fore it has divided, is represented in 

 Pig. 3, magnified 17 diameters, and a 



small part of the same section is shown 

 more highly magnified in -Fig. 4. A 

 glance at the latter shows that the cells 

 (1) which line the cavity of the spur, 

 differ very little in size or shape from 

 those (2) which cover the outside of 

 the spur. Fig. 5 represents a similar 



Fig. 4.— Cross-section of nectary and pedicel near the 

 lower end of the former, magnified 25 diameters. 



section cut just above the bulb, and a 

 few of the cells which line the larger 

 tube are shown more highly magnified 

 in Fig. 6. The epidermis or skin on the 

 outside of the spur does not differ mate- 

 rially from Fig. 4, 2, but it is seen 

 at once that the cells which line the 

 tubes at this place are very different 

 from those which line them further up. 

 Instead of being flat or slightly rounded, 



Fig. 5.— A portion of the same magnified 200 diame- 

 ters, to show the form of the secreting cells (1) ; 

 a— shows one of these cells magnified still more 

 strongly, its protoplasm having shrunk away from 

 the cell-wall. 



they are conical in outline, and their 

 ends project as so many little papillae 

 into the tube ; moreover, their walls 

 have curious thickenings at the apex 

 of the cone, and the protoplasm or jelly- 

 like substance that tills them is much 



