BIOLOGIC RELATIONS BETWEEN PLANTS AND ANTS. 437 



the plant dies ; in the contrary case the wound made by the insect causes 

 a considerable development of cellular tissue, the tubercle enlarges, the 

 stem develops. (Fig. 2, Nos. 6, 7, PI! XX.) Soon the ants find a suffi- 

 cient space in which to found a colony, and they excavate within the 

 tubercles galleries in all directions. If this view is correct, these plants 

 could not live nor develop without ants. These insects must contribute 

 to the formation of the organ that is to be the water reservoir of the 

 plant. But on the other hand the ants could not live and reproduce 

 their kind if they had not at their disposal plants in which they could 

 construct such a living home. 



According to Beccari the tubercles of the Myrmecodias and Hydno- 

 phytums are products primitively foreign to the plant. They are devel- 

 oped in the same way that galls or cecidia are — that is to say, they are 

 produced upon vegetable organs in consequence of irritation caused by 

 various insects. There is a striking analogy of form, and, indeed, of 

 internal structure, between these tubercles and a certain gall formed by 

 a curculio of the genus Gcntorynchus on the root of the garden cabbage. 

 The larva of this insect feeds exclusively on the cortical portion of the 

 root. As fast as this food is consumed a new generating layer prolif- 

 erates and replaces the destroyed tissue. The life of the insect is per- 

 fectly compatible with that of the plant. It injures no essential organ, 

 and the losses to which the plant is subjected are compensated for by 

 the hypertrophy of the tissues under the irritation caused by the insect. 



This analogy between the tubercles of the Myrmecodia and the gall 

 formed on the cabbage by the Gcntorynchus may suggest the hypothesis 

 that these tubercles are organs whose development must have been 

 caused by a parasitic lesion made, perhaps, by ants, which are known 

 to attack for food vegetable tubercles — potatoes, for example — as we 

 have ourselves seen. This lesion might be at first merely compatible 

 with the life of the plant, then useful to it, by the adaptation of the 

 injured and hypertrophied organ to an organ for the storage of water. 

 Fixed by selection, this character, at first accidental, would finally 

 become hereditary. In order to know what credit to give this hypoth- 

 esis it is necessary to study in detail the existing relations between the 

 ants and the Myrmecodias. There is no doubt that ants, even in our 

 own countries, sometimes establish themselves within certain galls that 

 have been abandoned by the insect that produced them. Such is the 

 case with a gall formed upon the Gynara cardunculus by a curculionid 

 larva (Larinusf). 



Is there between the Myrmecodias and the Hydnophy turns on the one 

 side, and the ants on the other, a reciprocal exchange of services, 

 mutualism, symbiosis in the strict sense of the word, or, indeed, can not 

 those plants do without the ants; are not the insects merely commen- 

 sal? The interesting researches of Treub upon a Javanese species of 

 Myrmecodia allow us to partially answer these questions. 



We must first study the structure of the young plant, then the 



