632 
SPECIAL ARTICLES 
DECIDUOUS ROOTLETS OF DESERT PLANTS 
A stupy of the roots of many perennials 
and of a few annuals growing under arid con- 
ditions in the Tucson region, shows that there 
are two sorts of rootlets which, to a degree, 
have similar functions, but which usually have 
an unlike fate. Whether analogous plants of 
the more moist regions have similar rootlets 
is unknown to the writer. A characterization 
of the rootlets in question can be made by 
describing those of a typical shrub, such as 
Franseria deltoidea, and by referring to anal- 
ogous rootlets of some of the annuals. 
In Franseria the tips of the main roots end 
in a brush of slender rootlets which bear root 
hairs. Most of these terminal rootlets are 
formed during the moist seasons, generally 
summer, and die during the following dry 
season, but some survive to extend the root- 
system. Such rootlets are found on the most 
deeply penetrating, as well as the most shal- 
lowly placed roots. But on the more super- 
ficial roots only there is, in addition, an- 
other type of rootlet. These are filamentous, 
2 cm., more or less, in length, in groups of 
about one half dozen, which occur from 2 to 
4 cm. apart. They arise from older roots 
only, and hence are adventitious. Since these 
rootlets are formed during the moister seasons 
each year, and die during the succeeding dry 
season, and never, or almost never, persist to 
form permanent roots, I have referred to them 
as deciduous. 
The deciduous rootlets, as before suggested, 
are found on roots which are placed near the 
surface of the soil. From this fact, and from 
the known variation in water content of the 
superficial soil horizon, it is assumed that the 
deciduous rootlets are perhaps the first absorb- 
ing organs, after the beginning of the rainy 
season, to function. Also owing to the fact 
that the superficial soil layers are the first to 
become desiccated, it is assumed that the 
deciduous rootlets cease activities before the 
1<¢The Root Habits of Desert Plants,’’ W. A. 
Cannon, Carnegie Institution of Washington, Pub- 
lication No. 131, 1911. 
SCIENCE 
[N.S. Vou. XXXV. No. 903 
other type referred to. It is probable, there- 
fore, that the deciduous rootlets are of great 
importance in providing water absorption sur- 
face during the time of maximum, or opti- 
mum, water supply, and that the second type 
of rootlets in deeper soil, lingering longer, 
serve to provide the plant with water during a 
longer period, probably until the next season 
of precipitation. 
Undoubtedly an important factor in the 
physiology of water absorption, and in the 
water relation, generally speaking, is that of 
the distance of water transport. Where the 
distance from the point of absorption to the 
place of evaporation is great it is probable 
that a given amount of water is less efficient 
in a longer than in a shorter transport. We 
have, in the formation of the deciduous root- 
lets, the interesting condition that the water 
absorption surface is enormously increased 
without at the same time increasing the dis- 
tance of water transport. The importance of 
this in the physiology of the plant will appear 
at a glance. 
Nearly all perennials which have been ex- 
amined are provided with deciduous rootlets. 
It is not supposed, however, that these root- 
lets are essentially different from the other 
type, although they have a different origin, and 
although they appear to have a somewhat dif- 
ferent function. From a few instances it has 
been seen that permanent roots may be derived 
from groups of rootlets which probably were 
indistinguishable from the deciduous rootlets, 
but which were so fortunately placed that 
survival was possible. 
An analogous condition, but naturally dif- 
fering in many ways, is to be found in the 
root-systems of several desert annuals. In 
such forms, in addition to the usual and, of 
course, ephemeral absorbing roots, there are 
at the base of many laterals rudimentary 
rootlets. These may remain rudimentary, 
never developing, or if the moisture condi- 
tions favor, they may develop. Where such 
rootlets are not found, it seems from certain 
experiments that they can not easily be in- 
duced. Probably the greatest advantage to 
an annual bearing rudimental rootlet is that 
