. Xmphicarpcea monoica. 287 



ful specimens of axillary runners have been those growing 

 anion- the dead leaves, which excluded nearly all the light 



All axillary shoots from the axils of compound lea\ 

 exhibited for a time geotropism. Plants which w< 

 inverted gave proof of this. Care- was taken that plants 

 selected for this experiment should have short axillary 

 shoots (not more than an inch and a half). Within twenty- 

 four hours, the portions of the specimen had adjusted 

 themselves to their altered relations; the shoots were 

 geotropic. 



Ordinarily this condition does not last long, for as growth 

 increases, a great curve is noticed near the tip, sometimes 

 extending for several inches. All that is needed is the 

 placing of a support near the apex. In a few hours twin- 

 ing begins, and consequently apogeotropic tendencies are 

 manifested. If no support is given, the shoots will twine 

 around any neighboring support, or indeed around some 

 adjacent individual of the same species. 



We may now recapitulate briefly. Axillary runners from 

 si mple leaves are geotropic and negatively Jicliotropic. In the 

 young condition, the axillary runners from compound leaves 

 are geotropic ; this characteristic is usually lost quite soon, 

 for they become apogeotropic and twine readily. The last 

 mentioned shoots usually appear to resemble the main stem 

 in their habits of growth. 



Occasionally exceptions to this last statement are seen in. 

 certain plants that, differing in no apparent particular from 

 others in their early life, exhibited later the peculiarities 

 which will be described below. The axillary shoots from 

 compound leaves, as well as the branches from these, were 

 purplish, rather than green, and leaves were very few ; 

 often entirely absent. Such shoots seemed always geotro- 

 pic. I have now in mind, one plant which grew in the 

 oreenhouse during the summer; it became ful lv five feet 

 high ; the main stem twined vigorously. The long purple 

 shoots were numerous, but they did not exhibit the slightest 

 inclination to twine, even around neighboring shoots ; they 



