ANTHRACNOSE OF CUCURBITS. 45 



melons unquestionably is another means of long-distance spread of 

 the disease provided infectious material reaches the fields. 



It seems fairly safe, however, to elimina'te all of the above factors 

 except overwintering as rather remote possibilities. As an explana- 

 tion of the appearance of the disease in new localities, all evidence 

 points toward introduction with the seed. Previous workers have 

 suggested this possibility, notably Sheldon (46; p. 127-137), Garman 

 22, p. 51), Eckardt (11), and Eriksson (16, p. 126-127). Eriksson 

 presents observational evidence which strongly indicates the intro- 

 duction of anthracnose into greenhouses in Sweden with cucumber 

 seed from England. 



FIELD OBSERVATIONS. 



The experimental field at Princeton in 1915 had previously been 

 in sod for seven years. Anthracnose appeared in this field as soon 

 as in any other fields of the region. So far as could be learned none 

 of the five experimental fields at Madison in 1916 had previously 

 grown host crops, yet the disease appeared simultaneously in all 

 of these fields and in practically no other fields or gardens of the 

 surrounding territory. In the watermelon industry of the South 

 crop rotation is necessitated by the ravages of the wilt, yet anthrac- 

 nose recurs annually. 



Instances could be multiplied, but sufficient evidence has been 

 adduced to show that the disease appears quite commonly in new 

 fields. In a case such as that presented by the 1916 Madison fields 

 there seems to be but one plausible explanation, namely, disease 

 introduction with the seed. 



The very manner in which the disease originated in these fields 

 suggests this hypothesis. The disease first appeared in the Prince- 

 ton field and in the Madison fields in what we have termed "original 

 centers." These were scattered here and there through a field and 

 were usually limited to a comparatively small number per acre 

 (figs. 13 and 14). With a disease as infectious as this it is not 

 always easy to differentiate between original and secondary centers, 

 but in fields under constant observation those centers appearing 

 first and about simultaneously may safely be termed " original." 



The nature of these initial infections furnishes another argument 

 in favor of their origin in situ. The first original center found in the 

 Princeton field consisted of two adjacent plants not yet large enough 

 to "run." These plants bore lesions on the first and second foliage 

 leaves. Since the rows had already been thinned, several plants had 

 been removed between and adjacent to the two above noted, and 

 the hypothetical originally diseased plant might already have been 

 removed. 



In 1916 the first center of anthracnose found in field 1 consisted 

 of two adjacent plants, one bearing one lesion on the fourth leaf, the 



