320 Gwynne- Vaughan. — On Polystely 
Sometimes the transition takes place in P. involucrata 
much in the same way as in P. japonica, the endodermis at 
the seventh or eighth leaf-gap being completely invaginated, 
and cut off by the closing of the gap. As in P. japonica , the 
departure of the next leaf-trace establishes a horseshoe of 
vascular tissue. But here the gap between the arms of the 
horseshoe remains open until the departure of a still higher 
leaf-trace, which divides it into two groups, and then it pro- 
ceeds as above described. In either case the vascular groups 
are at first imperfect steles, but after some time they develop 
internal vascular bundles, and assume the form found in the 
mature stem. 
In P. Auricula the fact that even the earlier leaves are 
supplied with more than one leaf-trace from the stem serves 
to complicate matters considerably, but the median trace has 
such a paramount influence on the transitional phenomena 
that it alone is alluded to below. 
The epicotyledonary stele is maintained intact until the 
fourth or fifth leaf (cotyledons included) is reached. Until 
this point a very small gap only is left by the outgoing traces, 
but at the departure of the fourth or fifth leaf-trace a gap of 
considerable width is formed, around the margins of which the 
endodermis is invaginated into the medulla, just as in P. ja- 
ponica. Thus a horseshoe of vascular tissue is produced 
which remains open on one side until the next leaf-trace is 
given off on the opposite side of the stem, dividing it into two 
separate groups, just as in P. involucrata . These two groups 
are almost always of unequal size, and they appear to become 
perfect steles even before the next leaf-trace comes off from 
the largest of the two. So we see that P. Auricula becomes 
completely and typically polystelic at a very early stage. 
The lateral leaf-traces arise on either side of the median trace 
from the steles that have just been formed by the departure 
of the latter, or, in the case of the formation of the horseshoe, 
from its corners, whereby the gap between its two arms 
becomes much widened. 
Van Tieghem seems to have entirely overlooked the all- 
