302 BOTANICAL GAZETTE [APRIL 
slightly differing in tracheid width, often mingle; thus in a second 
manner is shown that there is some problem as to the unexpected 
width of spring tracheids in Abies balsamea. Juniperus virginiana 
(32 uw) often mingles with the two species of Picea in moist ground 
and does not differ much from them in tracheid width. But Pinus 
Strobus (41.5 mu), often replacing these three species, and Pinus * 
Banksiana (43 p»), often occurring with Picea alba, though agreeing 
fairly with one another, have much wider tracheids than these three 
species. Larix americana (39.5 u) often largely replaces Abies 
balsamea (40 p) and Picea alba (33.5 uw) on sphagnum bogs. 
On the warm temperate Pacific coast, Pinus insignis (40 w) and 
Cupressus macrocarpa (39m) are often grown together as dune 
fixers, and nearly agree in tracheid width. 
Still closer agreement characterizes Abies concolor (39 u) and 
Sequoia gigantea (39 w), which are often associated in the warm to 
cool temperate Pacific forests. Again, the equivalent species of 
Pinus have relatively wide spring tracheids, for P. Jeffreyt (47 u) 
often is associated with Abies concolor (39 4), whose demands for 
moisture are about equal to those of P. Lambertiana (45 u) (Mayr). 
In the cool temperate Pacific region, the deciduous Larix occi- - 
dentalis (42 ») has much wider spring tracheids than its frequent 
companion Pinus Murrayana (34 u); while Abies nobilis (41 u) and 
Abies amabilis (40 4), which occur together, differ only slightly 
width of spring tracheids. 
In the Pacific so-called alpine region, there is a remarkable 
approximate agreement in the width of spring tracheids in three 
genera: Abies lasiocarpa (33.5 u), Picea Breweriana (33 ), Pinus 
Balfouriana (32 u), Pinus aristata (33.5 u), and Pinus albicaulis 
(35 #). All these contrast with the alpine larch, Larix Lyallis 
(43 mu), whose wide tracheids correspond with the deciduous habit. 
American deciduous species of Quercus 
In a previous paper (1) I have shown that the width of the 
spring vessels in Quercus is decided partly by the habit of the specte> 
and that these vessels are wider in deciduous than in evergreen 
species, even though the latter may grow in moist Florida, as 1S the 
case with Q. virginiana. The subjoined statistics suggest strongly 
