iQiSl BROWN— PINUS STROBUS 225 



Changes in temperature are less effective there because the thick 

 corky layers tend to equalize the conditions which prevail at different 

 times during the growing season. Cool autumn nights, for example, 

 would chiU the cambium in the upper parts of the tree much earlier 

 than below. Temperature changes become operative first where 

 the primary cortex still persists, that is, where the bark is yet 

 smooth. This without doubt explains the disparity of growth as 

 we find it in white pine. Growth is first retarded above, but may 

 go on vigorously below for a much longer period. 



The exact time of growth cessation apparently varies widely 

 in different species, in different localities, and in different sites. 

 While wide variations occur, still certain generalizations apply. At 

 the outset the term "growth" is a misnomer. As already noted, 

 phloem formation, at least in conifers, does not cease with xylem 

 formation, but continues uninterruptedly untU late in the fall. 

 It is necessary, therefore, to discuss xylem and phloem separately 

 in their relation to cessation of growth. 



A comparison of cuttings from tree I for the years 191 2 and 19 13 

 will give an idea of the seasonal termination of xylem formation. 

 One discrepancy was noted at the start. In spite of the fact that 

 the final cuttings in 1912 were a week later than those in the follow- 

 ing year, growth was apparently more vigorous at the later date 

 in 191 2 in all four cuttings. This is to be explained in two ways. It 

 was due either to seasonal variations or to the fact that the vigor 

 of the tree had been materially lessened the second year through 

 the many cuttings taken from it. An examination of the meteoro- 

 logical data for the two seasons has added no convincing evidence, 

 inasmuch as comparative figures of growth for the two years were 

 not at hand for a sufficiently large number of individuals, and 

 general assumptions were therefore out of the question. Possibly 

 both factors were in force. 



To give the exact time or a definite place in the tree for the 

 termination of xylem formation is quite impossiblej as the data on 

 tree I indicate (table F). In 191 2 growth was still in evidence on 

 September 26 in all four cuttings, as transitional forms of tracheids 

 could be noted in every case (fig. 6) . Growth, however, was going on 

 at this date very sluggishly. Often only one flattened transitional 



