326 REPORT UPON COTTON INSECTS. 



oviposits on it, and this in but small numbers, its larvae feeding on the 

 green seeds. The same complication, therefore, exists here as in the case 

 of the cotton plant ; but in this case the attack appears to be limited to 

 the early fruiting period, and a body-guard of ants is maintained during 

 this period. 



Coming, now, to leaves, we may briefly refer to the sweet fluid known 

 as honey-dew, which is sometimes found on the foliage of plants. In 

 many cases this will be found not to originate in the leaves, but to drip 

 from the anal tubes of aphides, or plant-lice; and with this we have 

 nothing to do, since it is not a production of the plant. But in some 

 cases this substance is an excretion from the leaves, apparently due 

 either to the climatic conditions obtaining at the time of its production 

 or to a diseased state of the plant. It is not, so far as I know, produced 

 by structures, such as glands, in any case. Though bees and ants col- 

 lect this substance with avidity, it does not appear that they render the 

 plant any service while doing so.* 



Small glands are found at the tips of the serrations on the leaves of 

 many plants, and some of these produce a plentiful supply of nectar ; 

 some of them being frequently visited by insects, and others scarcely at 

 all. Like the last, this nectar is believed by Darwin to be merely ex- 

 cretory, and as going to show that such is the case we may mention the 

 fact that the leaves of peaches, nectarines, and apricots which may be 

 glandular in some, and not glandular in others of the offspring of a 

 single parent if glandular, are less liable to the attacks of mildew than 

 if they bear no glands.t 



Leigh ton found that 



On the upper edge of the vertical phyllodia of Acacia magnified, subtending the 

 showy spikes of yellow flowers, which proceed from their axils, appeared a pellucid 

 drop of liquid, varying in size from that of a large piu's head to that of a grain of 

 mustard-seed. This to the taste was sweet and sugary. The flowers themselves had 

 no odor, except toward nightfall, when they gave out a weak disagreeable smell, only 

 perceptible on close contact. In wiping off the sugary secretion it was observed that 

 it proceeded from a small sunken linear-oblong orifice or slit, surrounded by a swollen 

 margin. * * * The secretion takes place only during the period that the plant is 

 in blossom. So soon as the flowers fade and begin to fall, the secretion ceases and 

 disappears. It would seem then to be in some way or other connected with the fer- 

 tilization of the flower ; and as, when the secretion becomes excessive, it falls and 

 blotches the lateral expansion of the phyllodium, it is probably to attract insects to 

 efl'ect this, * * * it seems almost evident that it would require an insect of some 

 considerable size and of some peculiar structure to remove and apply the pollen, the 

 secretion not being in the blossom itself, but at a short distance from it, 011 the phyl- 

 lodium.t 



This case appears quite similar to that of the cotton flower previously 

 given, and I cannot avoid the conclusion that the real object of the nec- 



* Darwin, Cross and Self Fertilization, 1877, page 402, mentions undoubted cases of 

 the occurrence of this excretion, besides giving references to other writings bearing 

 on this point. 



t For references on this subject see Darwin, Animals and Plants under Domestica- 

 tion, Orange .Tudd edition, 1868, i, 413: ii, 280. 



J Annals of Natural History, third series, xvi, 1865, page 12. 



