800 THE AMERICAN NATURALIST. [Vor. XXXIX. | 
noted that the vessels are very small and there is a relative pre- 
ponderance of the mechanical tissue. Assuming that the wil- 
lows originated under the influence of a relatively warm climate 
as we have reason to believe must have been the case on the 
basis of evidence brought forward by Wieland (:03), it is evi- 
dent that Sax uva-ursi in common with other boreal and alpine 
forms, must represent a degenerate type which has been re- 
duced under the influence of a diminishing temperature. But 
such reduction has clearly involved a loss of the originally erect 
position, and in consequence of the diminished necessity for 
mechanical support, it might be inferred that there would be a 
more or less pronounced tendency to reversion in such wise that 
the transition zone would tend to regain its primitive character 
to some extent, and that it would once more exhibit those devel- 
opmental features of structure which have essentially disap- 
peared from the great majority of species as now known. It 
was therefore felt that this plant might offer exceptional oppor- 
tunities for a solution, in part at least, of the question as to the 
actual genesis of the vessels and the wood cells, and their 
mutual relations. The stem selected for study was of a some- 
what sinuous form and in cross section ovate, the diameters 
being 1.20 and 1.80 cm., with the pith eccentrically placed to- 
ward the narrower or under side, the major axis of the cross 
section being oriented vertically. Longitudinal sections were 
cut from this stem at a slight angle with its principal axis, and 
in such number as to embrace not only the entire pith, but a 
considerable portion of the secondary xylem lying on opposite 
sides. It will thus be observed that while some of the sections 
were strictly radial their character gradually changed as the 
relation of the plane of section to the radius changed, until 
they became strictly tangential. It thus became possible to 
study the elements of the transition zone from these two points 
x en and in a progressive series, But owing to the slight 
M plane of section to the longitudinal axis of the 
, section would exhibit at one end the radial aspect 
= won iens ie the tangential aspect, while between the 
Se a graduated series of changes showing all 
re radial extent of the transition zone. This 
