BIOLOGIC RELATIONS BETWEEN PLANTS AND ANTS. 415 



accumulated by the auts. These are, in fact, to be considered as weeds 

 of cultivation and strangers to the lavender and cistus covered banks 

 of the garrigue, being plants sprung from seeds that the ants have 

 brought and abandoned for some unknown cause. The plants thus 

 transported belong to the following classes : Veronicas, fumitories, oats, 

 a nettle ( Urtica membranacea), bird chickweed, wild marigold (Calendula 

 arvensis), snapdragon (Antirrhinum majus), a flax (Linaria simplex), 

 watercress (Cardamine hirsuta), and goose foot. 



Quite frequently these plants are found along the sides of minature 

 gullies or crevices hollowed in the rock where their seeds have been 

 washed by the rain and there germinate. Thus these interlopers have 

 been drawn into competition with the primitive occupiers of the ground ; 

 they accompany the ants as the plants of our harvests accompany man. 

 The ants serve indirectly in their dissemination, using, indeed, part of 

 them for food, but yet assisting in the propagation of the species. As 

 soon as the harvested seeds are brought near to the nest, some hundreds 

 of workers are employed in separating them from the husks while others 

 store them away in the depths of the ant hill. The refuse is dragged 

 out of the nest, in the immediate vicinity of which are found heaps of 

 debris formed of bits of straw, pods, and empty capsules. 



The nest is simply hollowed out in the soil, but it seems that some- 

 times the ants know how to appropriate the work of certain beetles. 

 Moggridge has, in fact, seen in one of the nests a cavity covered over 

 by a spherical dome having walls of hardened earth closed at the bot- 

 tom, there being a large circular opening at the top and a smaller one 

 below. This appears to be a dome constructed by a beetle and used 

 by the auts for storage purposes. 



The floor of these grain cellars is well cemented. The rooms differ 

 in size, being on an average as large as a good-sized watch. Bach 

 of these rooms contains about ~> grams of seed. A nest made up of 

 from 80 to 100 rooms may contain a pound or more of seeds belonging 

 to different plants. The majority of these are, however, from culti- 

 vated grasses, especially Tragus racemosus. These are evidently pre- 

 ferred because of their richness in alimentary principles. 



Especially interesting are the means employed by the ants to pre- 

 vent the germination of seeds. In examining 21 nests Moggridge 

 found, among some thousands of seeds, but very few that had germi- 

 nated; these were, nevertheless, attacked by the ants, who attempted 

 to mutilate them in order to stop their germination. This arrest of 

 germination caused by the intervention of the ants is unquestionable, 

 but we have not yet discovered the exact means by which it is effected. 

 In isolated or abandoned portions of the nest the seeds sprout and 

 develop in the granaries, and if the ants are prevented from penetrat- 

 ing into one of these the seeds germinate there normally. It has been 

 surmised that the ants could prevent germination by closing, with a 

 gelatinous substance, the micropyle of the seed, through which it was 



