203 
but one exception. In Myrtus communis no simple-pored libri- 
form cells were found, though it is possible that they exist, but in 
such small quantities as to have escaped detection. In this fam- 
ily there isa marked difference in the quantity of the simple-pored 
libriform in the different genera, and it is difficult to give a very 
definite order of arrangement of the different layers. In general, 
it may be said, whenever the libriform of any part of the year’s 
ring contains bordered pores, they are never wanting in the last 
layers of the year’s growth, so that they stand in direct commu- 
nication with the cambium layer in the early spring. 
A few facts of interest in this connection may be mentioned 
in regard to the size of the border and position of the pores in 
the nineteen families where only bordered pores occur. 
In the genera studied, all variations of size occur between the 
smallest and largest found. Among the smallest are those of 
Erica and Tabernemontana. The length of the border in Evica, 
as seen in profile, is about 1% micromillimeters, measured on the 
cross-section ; the longitudinal section averaged a trifle longer, be- 
tween 1% and 2mm. The largest were found in Syringa, Drimys, 
and a few others. One of Drimys was found to measure 9 mm., 
but this was an exception, as they average here from 6 to 7%. 
In Syringa they are between 6 and 7 mm. These larger ones 
compare very fairly with the size of the border in many of the 
Conifers. For example, those of Pinus and Larix measure from 
7 to 7% and 8 mm. 
Frequently variations were found in genera of the same 
family, as in case of Ericaceew, Magnoliacee and others, some- 
times in the same specimen, as in Quercus, where they vary from 
3% to § mm., according to their position in the year’s ring. One 
or two instances were found where the borders of the same cell 
differed in size. 
Of all the genera examined in these eighteen families, Erica 
and Zabernemontana are the most interesting, as they furnish the 
only exceptions thus far found, to the statement before made, 
that the simple-pored libriform cells are the only ones used as 
reservoirs for starch. Here the libriform cells contain very small 
starch grains and are furnished with bordered pores; the borders, 
however, are so extremely small, that in spite of their perfectly 
