PECAN ROSETTE. 35~ 
tubers give rise to diseased progeny. Furthermore, the fact that 
many of these diseases are transmissible by insects shows them to be 
entirely different in nature from those chloroses due to soil or cli- 
matic conditions. In some of these diseases where the insect rela- 
tions have been most carefully worked out it is indicated that a 
definite incubation period must also elapse after feeding upon a 
diseased plant before an insect becomes infective. 
Except in the most general way, the infectious chloroses exhibit no 
pattern as related to the leaf. The factors controlling morpho- 
genesis of the leaf in many cases seem to have run riot. In the 
mosaic type the spots, irregular in themselves, are also irregularly 
distributed over the surface of the leaf. In many cases, it is true, 
they avoid the larger veins, but in other instances they are distributed 
irrespective of venation. 
There is nothing to show a difference in kind between those mosaics 
transmissible by expressed plant juices and those transmitted only 
by grafting, such as the infectious mosaic of Abutilon. The causal 
contagium in the latter case may be compared to a more obligate 
parasite which is greatly restricted in the conditions necessary to its 
life activities and reproduction. 
The fact that later investigation has shown some of the filterable 
contagium diseases to be due to microorganisms is at least presump- 
tive evidence in favor of the organism theory of infectious chloroses. 
Furthermore, Allard has shown that with a fine enough clay cup it 
is possible entirely to filter out the infective principle from the 
expressed juices of tobacco plants affected with mosaic. 
The crucial difference, however, lies in the power of self-reproduc- 
tion possessed by the contagia of all the infectious chloroses. To 
give a concrete example, starting with a mosaic-diseased tobacco 
plant and in each case using a drop of the expressed plant juice di- 
luted 1 to 1,000 in water, it is possible to cause the disease in an in- 
definite series of plants successively inoculated one from the other at 
proper intervals, and the juice from the last of the series will possess 
as high a power of infection as that from the first. In a case like 
this it might be mathematically shown that if reproduction had not 
taken place a drop from the last plant of the series must contain less 
than one molecule of the originally injected material. No such power 
of self -reproduction has ever been demonstrated for a definite chemical 
compound. The true enzyms are formed by living organisms in re- 
sponse to certain stimuli and have never been shown to be capable 
of self-reproduction. 
Assuming the contagia of infectious chloroses to be nonliving sub- 
stances, their effects may be compared up to a certain point with 
various phenomena of bacterial toxins and immunization. However, 
the fact of reproduction in these contagia is in no way elucidated by 
