36 BULLETIN 1038, U. S. DEPARTMENT OF AGRICULTURE. 
any such comparison. Diphtheria toxin, for instance, if injected into 
a susceptible host may cause all essential signs of the disease, but it 
takes living bacteria to produce more of the toxin so that the disease 
may be transmitted down the line through a successive series of in- 
dividuals. 
On the other hand, the " Contagium vivum fluidum" theory of 
Beijerinck (17) in some modification is at least worthy of serious 
consideration as a working hypothesis. It is not impossible that de- 
composition processes may be propagated by methods entirely unlike 
reproduction by division as known in living organisms and so as to 
imitate self-reproduction. If some such course of events could be 
demonstrated it would perhaps explain the known facts, including 
the gradual progress of the disease from the point of entry, equally 
as well as the organism theory. However, though it is well to keep 
an open mind on all questions still in the realm of theory, no such 
course of events is yet known to chemistry. On the other hand, most 
of the known facts concerning the so-called filterable virus diseases, 
so it seems to the writer, conform to the known results of invasion by 
parasitic organisms. 
SUMMARY. 
As a class, the chloroses clue to soil or atmospheric conditions are 
rather general effects which are more or less comparable to starva- 
tion, overfeeding, or direct poisoning. There are not the profound 
or strictly localized derangements in both metabolic and anatomical 
development, and it is a question in regard to many of these phe- 
nomena whether, in the restricted sense, they should be regarded as 
diseases at all. 
In the specific chlorotic diseases of an infectious nature funda- 
mental derangements in both physiological and structural develop- 
ment are simultaneously brought about. Although all effects of 
parasite upon host when reduced to their ultimate reactions are 
probably to be explained in terms of physics and chemistry, it is 
in the regulatory effect and in the extreme complexity that these 
and many other infectious diseases differ from the direct effects of 
nonliving substances. 
The histological and cytological evidence suggests that pecan 
rosette in its specific sequence of signs and in the complexity of 
the structural and physiological derangements bears far more simi- 
larity to the known infectious chloroses than to those caused 
by soil or climatic conditions. Whether in this particular disease 
the factors responsible for alterations in the normal structure and 
metabolism must be introduced into the plant from without, or 
whether they originate within the plant itself, is a question yet to 
be answered ; but whatever the ultimate solution of the problem the 
cause will undoubtedly not be found in any simple soil or water 
relation. 
