438 BIOLOGIC RELATIONS BETWEEN PLANTS AND ANTS. 



changes that occur in its tissues up to the time when the young tuber- 

 cle has ah external opening giving access to an internal gallery. 



The hrst cavity or gallery in the young tubercle is not hollowed out 

 by the ants. It does not start from a lesion of the peripheral tissue, 

 due to insects, but is the result of an internal differentiation. A trans- 

 verse section of a young tubercle shows a parenchymatous homogene- 

 ous mass, inclosing in its center a libero-ligneous fascicle, and limited 

 at the periphery by an epidermic layer. The growth and subdivision of 

 the parenchymatous cells produces a thickening of the tubercle, at 

 whose periphery is formed a generating layer of cork. In proportion 

 as the tubercle grows, it forms, at the expense of some of the parenchy- 

 matous cells, new libero-ligneous fascicles, arranged parallel to the sur- 

 face of the tubercle, and soon becoming connected by transverse anas- 

 tomoses, not only with each other but with the fascicles of the stem and 

 of the root. The formation of these peripheral fascicles indicates the 

 beginning of the first gallery. A generating zone, parallel to the sur- 

 face of the tubercle and situated deeply within its mass, begins to dif- 

 ferentiate. All the parenchymatous material that it incloses in its 

 interior then begius to dry up. This dessication brings about a rupture 

 of this material, and there is thus formed the beginning of a central 

 cavity circumscribed by a layer of cork, the result of a differentiation 

 of the generating sheet. This cavity, cylindrical in its general form, 

 extends in two directions. Above, it ends in a vault near the insertion 

 of the stem, properly so called ; below, it approaches the periphery. At 

 last it constitutes a gallery nearly in the axis of the plant, lined with a 

 layer of cork and inclosing a flocculent matter, the remains of the primi- 

 tive parenchymatous tissue. The gallery is separated from the exterior 

 by a thin peripheral layer of cork which soon tears and permits com- 

 munication from without. 



The essential point is that the generating layer results from an inter- 

 nal differentiation, and not as a consequence of the sting of an insect. 

 This generating layer, which toward the interior of the tubercle pro- 

 duces cork, forms externally parenchyma, which contributes to the 

 increase in thickness of the tubercle. In proportion as the tubercle 

 thickens the number of galleries that traverse it increases. The new 

 galleries are formed by a process identical with that which gave rise to 

 the first gallery. 



Although the first gallery is spontaneously formed in the young 

 tubercle, it is possible that the hypocotylous axis did not thicken 

 enough to produce it without the stimulus of an ant, by a bite perhaps 

 imperceptible. But the lower part of that axis commences to thicken as 

 early as the first stage of germination (fig. 2, PI. XX). Why should not 

 that spontaneous thickening continue? We may add that the young 

 tubercle is provided with chlorophyll, and can assimilate even at the 

 time when the cotyledons are yet inclosed in their seminal envelope. 

 According, then, to these observations, the sting or bite of an ant 



