78 CROCION ACHLYDOPHYLLUM 
varying with age were gathered and among them were some plant- 
lets the roots of which seemed so old and large as to excite suspicion 
that they were not real seedlings of the previous season’s germina- 
tion. On tracing the course of the roots it was found that they were 
those of a central mature plant towards which they radiated inward 
from all directions. In the majority of cases the plantlets seemed 
to take origin from roots that had decayed or were decaying and 
had been severed from the parent plant. Once or twice only was 
it found that the root that bore a plantlet still had a pronouced 
communication with the parent, though the region around the 
origin of the plantlet showed a rather unhealthy condition. This 
phenomenon, however conclusive called for actual experimenta- 
tion in order to substantiate former observation in the field. 
A number of roots, preference given to those that were fully 
mature and in normal state of health, taken at random and gathered 
from many parents not in close proximity were planted in our labora- . 
tory in boxes. The greatest precautions were used in taking soil 
specially sifted and selected so as to exclude seeds of violets. 
Placed in the light, the ground was kept under conditions of moisture 
and light as near as possible to those of the natural habits of plants. 
From time to time the roots were examined to see, whether they 
would show such characteristically localized regions of decay, 
as seen in those in their natural conditions and surroundings 
prior to giving rise to young plants. In about three weeks de- 
generation occurred in the cut ends of the majority. It is from this 
partially decayed part of the root that a bud soon appeared made 
up of scales and followed by one or two well developed leaves. 
This showed conclusively that the violet reproduced new plants 
vegetatively from old roots. , 
Among the many parts of seedlings and mature plants 
brought up in parraffin were included besides the mature healthy 
roots, roots showing different stages of decay with some bearing 
plantlets for comparative study of the fibrovascular arrangement 
and also to investigate the region of plantlet origin. The fibro- 
vascular arrangement was alike in roots showing decay and in 
those perfectly normal (Fig. 13-14). The plantlets originate from 
the region surrounding and including the pericycle. The anatomy 
of these propagated plants was not further studied after deter- 
mining that they were more or less like seedlings themselves 
in structure. Fig. 5 shows two old roots severed from the parent, 
