NECTAR AND ITS USES. 331 



solely at night,. In this case the plant would be under the necessity of 

 secreting nectar only during the night. 



As I have stated before, the extrafloral nectar of the cotton plant is 

 far more abundant during the night and in the early morning than at 

 any other time, and this is true whether we consider the iuvolucral or 

 foliar glands. At first, from the visits of ants to glands in which I could 

 detect no nectar, and from the fact that when the largest drops of nec- 

 tar were seen early in the morning, the leaves were covered with dew ; 

 I was led, after satisfying myself that these drops were not confluent 

 dew-drops, to conclude that a thin film of sacccharine matter covers the 

 glands at all times when they are in a t .althy condition and of suffi- 

 cient age, and that this is hygroscopic, absorbing so much watery vapor 

 from the damp night air as to present the phenomenon mentioned. But 

 I was led to doubt this conclusion by noticing that the secretion of the 

 iuvolucral glands lasts only during the blooming period of the iiower 

 about which they are placed, and I could see no reason why their nectar 

 should be hygroscopic for so brief a time. This led me to examine 

 glands in damp weather, before, during, and after a rain ; but no drops 

 of nectar were found, though drops of rain-water were occasionally found 

 hanging from the border of the glands. So the hygroscopic theory 

 would not do. On the contrary, I found that in the early morning after 

 a cloudy or rainy day, there was comparatively little nectar in the 

 glands, which seems to show that the secretion during the night is the 

 result of the solar impulse of the preceding day. I could then scarcely 

 avoid the conclusion that this nectar was originally developed by nat- 

 ural selection, that it might attract some nectar-loving animal to protect 

 the plant from the depredations of some leaf and flower eating creature 

 whose visits were chiefly made at night ; and such I believe to be the 

 case, both attackers and defenders having been ants in all probability. 



But, it may be urged,. you have said that this nectar is, at the pres- 

 ent time, an important factor in securing the well-being of the plant, 

 since it attracts ants and wasps which are among the most powerful of 

 the natural enemies of its great spoliators, the boll- worm, and cotton 

 caterpillar; why can it not have been developed to secure protection 

 against them or some similar insects? The fact that it is protective to 

 the plant in this way is undeniable; but from what we know of the 

 economy of nature it seems improbable that a nocturnal secretion of 

 nectar should have been secured as a means of protection against larvae 

 which feed for the most part by day; while its very abundance at 

 night was certain to greatly increase their irumber on plants where this 

 peculiar secretion chanced to be most marked during the process of 

 selection, by attracting to those plants a greater number of the moths 

 whose offspring the larvae are. 



On the other hand, it may be urged that inasmuch as this nectar is 

 now so attractive to the moths of Aletia and HeliotMs, it probably does 

 more harm to the plant in attracting them where they may lay their 



