se ae BOTANICAL GAZETTE [MARCH 
number of xylem rays. The secondary xylem forms between the 
primary xylem rays and under the primary phloem, and it follows 
that there are as many secondary xylem areas as there are primary 
xylem rays. In a young root where the secondary thickening has 
begun we find the primary and secondary xylem areas alternating 
with each other (figs. 9 and 10). It is usually not until the 
following year that the segments unite and complete the ring of 
cambium. 
An unsuccessful attempt was made to find out at just what 
period in the growing season secondary thickening began in the 
root. Roots were examined on May 11 and again on May 30 with 
this object in view. No secondary growth was in evidence in 
either case in the new tissues, even when, as at the last date, 
elongation had gone on to the extent of 5-6 inches. In every case, 
however, where the last formed growth of the preceding season 
was examined, secondary xylem was in evidence between the poles 
of the primary xylem, and evidence of a resting period was to be 
seen, so that it must be concluded that secondary growth occurs 
later in the growing season than May 30, probably during the second 
period of activity in the autumn. The cambial segments, however, 
apparently do not unite over the poles the first year, so that second- 
ary growth the first season is confined to as many separate areas 
as there are poles. 
The course of secondary thickening in the root, once started, 
is much more irregular than in the aerial parts. The annual rings 
are usually thickest on the lower side of the roots as they enter the 
root crown, but all regularity is lost a short distance from the bole. 
The rings may be narrow here and broad there, and apparently 
their position in the ground has no appreciable effect; geotropism 
is not a factor in annual ring formation. False and double annual 
rings are often present. As RUBNER (28) has pointed out, the 
cambium may be active on one side of a root and dormant on the 
other for several years without its vitality being impaired, and this 
is responsible, in part at least, for the irregularities in growth which 
arise. Furthermore, the tissues of exposed roots present the same 
characteristics as those in aerial parts, a peculiarity previously 
noted by Kny (20). In conclusion, it may be said that roots, 
