U. S. DEPARTMENT OF AGRICULTURE. 
not fit under different conditions. It is well known that the length 
of vine of any variety is greater in the South than in the North. The 
length and size of the vine, as well as the size of the leaf, are greater 
on heavy soils than on light sandy soils. In the description of the 
Nancy Hall variety the stem is classed as short (under 4 feet long), 
while in practically every region where this variety is grown to any 
great extent the stem grows to be more than 4 feet in length. In the 
vicinity of Washington, D. C, on a sandy-loam soil the stems grow 
4 to 8 feet long. The vine of the Shanghai variety is also classed as 
short, but under the conditions mentioned in connection with the 
Nancy Hall it grows to be 6 to 9 feet in length. The stems of the 
Ticotea and Florida varieties are also classed as short, while under 
fairly favorable conditions they grow to a length of 6 to 9 or 10 feet. 
The other characters used in the key, with the possible exception 
of the "Arrangement of hair on the upper surface of the leaf, if any," 
and " Distinctness of wood elements in the root," are very important 
and reliable. All characters should be considered in connection with 
the mature plant, as many do not develop on new growth and some 
disappear as the plant nears maturity. The shape of the leaves on 
new growth is often very different from the shape of fully developed 
leaves. The surface of young leaA r es may be covered with hairs, while 
older leaves may be entirely smooth. The use of letters and numbers 
rather than the terms for which they stand makes the key rather 
awkward. 
Another attempt at classification of sweet-potato varieties is given 
in Bulletin 33, " Estacion Experimental Agronomica," Santiago de 
las Vegas, Cuba, entitled "Las Variedades Cubanas de Boniato." 
This work deals to a certain degree with American varieties, much of 
the material having been obtained in this country. In this system 
of classification four groups — white, yellow, purple, and reddish — 
are made, and these are further divided according to the color of the 
stems, leaves, and veins on the leaves. This system of classification 
is not sound, as the color of the roots varies greatly with difference 
in soil and climatic conditions. 
PROPOSED SYSTEM OF CLASSIFICATION. 
The work on which the proposed system of classification is based 
was begun by the Bureau of Plant Industry in 1901 and has been 
continued up ta the present time. During the season of 1901 the 
varietal work was carried on at College Park, Md., in cooperation 
with the Maryland Agricultural Experiment Station. At the end of 
the season, the work and the varietal collection were transferred to 
the Bureau of Plant Industry. From 1902 to 1920 the collection has 
been grown every year on the Potomac Flats, near Washington. The 
