32 BULLETIN 1038, U. S. DEPARTMENT OE AGRICULTURE. 
differentiation and lack of tissue differentiation, hypotrophy and 
hypertrophy of cells may all occur together within an area only a 
few millimeters in diameter. There seem to be focal centers out from 
which these alterations spread, and the resulting abnormal develop- 
ment appears to be controlled from these centers rather than by the 
normal morphogenic forces. The factors controlling morphogenesis 
as related to cell differentiation, to development of tissues in dis- 
eased areas, and to general pattern of leaf are profoundly altered. 
No less far-reaching are the derangements in metabolism evidenced 
by the altered assimilation and translocation of starch. 
Pecan rosette frequently appears first on the tip leaves of a single 
branch and seems as likely to affect a part of the tree first as to occur 
at once over the whole top. 
Careful field observations have given no evidence of varietal dif- 
ferences in susceptibility or resistance to pecan rosette. This same 
statement would apply equally well to most infectious chloroses. 
Fertilizer and transplanting experiments and field observations 
indicate that rosette is affected, at least indirectly, by soil and cli- 
matic conditions, but similar relations are exhibited by chloroses of 
known infectious nature as well as by many diseases of bacterial or 
fungous origin. Furthermore, great reliance is not to be placed upon 
the results of plat soil experiments unless such results are unusually 
definite, or unless in a large number of similar tests the data all 
point in the same direction. 
In view of the refractory attitude exhibited toward grafting and 
cross-inoculating experiments by certain of the infectious chloroses it 
is considered that this type of experimentation has not yet given a 
conclusive answer to the question of possible infectivity in pecan 
rosette. In the early experiments a small portion of the diseased 
buds developed the disease, though in no larger percentage than in 
contiguous nursery trees worked with supposedly healthy buds in 
the ordinary commercial propagation. These experiments need repe- 
tition on a large scale under controlled conditions and with extreme 
care in selection both of diseased and healthy buds and stocks. More- 
over, the possibility of insect transmission has not been touched upon 
from the standpoint of experiment. 
The disease has not been definitely and experimentally caused by 
a set of known conditions. Though it is more prevalent and severe 
under certain environmental conditions, it occurs to some extent in 
practically every soil type where the tree has been observed, with 
the possible exception of swamp land. Under these conditions the 
tree makes very little growth and presents a starved and stunted 
appearance followed by a general form of chlorosis bearing no 
resemblance to rosette. Ultimate proof of the cause must account 
for all cases of the disease. 
