178 BOTANICAL GAZETTE ‘[sepremam 
length on the side opposite the roots. In those sterile spec 
mens which do not form a runner the corresponding opening is 
found on the same side of the bulb at the base of the foliage — 
leaf, about 5™" from the roots upwards. In both cases the 
opening leads to a hollow space, near the end of which the new 
bulb is formed. In the sterile individuals mostly two runnes 
are formed by each bulb, the one being about twice as long 
the other and also thicker. The longer runner reaches an aver 
age length of about 12, with a diameter of 2-3, the lon 
est I found being 17™ in length. In flowering specimens | 
found the runner 3~5™™ thick and 8™ long, but I do not know 
the definite length it may attain. Where the new bulb is formed 
close to the old one, one finds instead of the smaller runner oflf 
a little bud. The smaller runner, as well as the little bud met 
tioned, may be found sometimes at the right, sometimes at the 
left side of the main bulb. | 
Without entering into morphological considerations = 
the parts of stem and leaves which take part in the formation : 
these structures, we will consider only their physiological ie 
ior. At the end of April the elongation of the runner st 
grains are found in the whole extent of the runne 
quantity in the new bulb. During May the development 
new bulb is accomplished, and at the end of the month t 
age leaves, the old bulb, and the runner die off, so *# 
beginning of June the new bulb is isolated. It remains 
mant state until the fall of the same year, when the @ 
described cycle recommences. 
The difference between those specime 
new bulb from the old one by a long runner, #” ' 
form it close to the old one, lies only in the relative ¢ 
the stalk uniting the stout stem-portions of the e 
Those stem-portions in which the leaves are inserted 
which the roots start are rich in vascular bundles, au 
nants may be preserved several years. Sometimes 4 
