— 45 ~ 
denced by the statements to be found scattered through the 
literature on this subject ascribing a length of ten, fifteen, 
thirty-hve, and even more feet to its roots. The popular 
estimation of their length has been and is equally appreci¬ 
ative of their power to penetrate to considerable depths. 
1 he size attained by the roots has also been stated to be 
large, but the writer does not recall having seen any fioures 
given to convey a definite idea of the size actuallv 
attained under stated conditions of soil, age of plant, cultf- 
vation, etc., but rather that the root is a tap root, large and 
fieshy, resembling a carrot more or less,or is represented 
as forming a symmetrically formed but inverted cone, in 
which system the tap root is, as a matter of course, the 
longest and central portion or axis. 
.The size of alfalfa roots is not so great as the usual ad¬ 
jectives used in describing them would lead one to infer. 
It is a strong root, but is under one-half inch in diameter, 
rather than above it. This statement is true of the plants 
when grown in a deep, sandy loam, under favorable con¬ 
ditions as to lirigation and climate, including mild winters. 
Laiger roots have been observed by the writer, but there 
have been special conditions obtaining wherever this has 
case and these roots represented the size which 
the alfalfa root may attain, and not the average size which 
they actually do attain when growing in ordinary soil, and 
standing thick enough to produce, say 3 1-2 tons of hay per 
acre, with three cuttings annually. 1 l he largest root meas¬ 
ured by me, was 2.82 inches in diameter, being nearly circu¬ 
lar in section, though not quite ; its largest diameter was 
rather more than three inches. An examination proved 
that this was an anomalous root. For some reason, 
not discovered, the tap or central root was short, not ex-' 
ceeding 1 1-2 feet, at which point it divided, giving rise to 
several rather small branches which were not followed as 
they spread out, running several feet almost horizontally, 
this dividing could not be attributed to the roots having 
encountei ed a haidpan or other obstacle, for the soil at 
this depth was uniform in hardness above and below the 
point of spreading. 1 have seen several very large roots, 
but have found upon digging them out, that they were 
in every case short and at variance with what seems to 
be the normal type. 
1 he root system of this plant, growing in our soils, 
is exceedingly simple and is shown in the plates. The 
roots represented, are from three different counties, the 
soils varying from sandy loam to heavy clay. They 
show a marked permanency in type of development in a 
