38 BULLETIN 859, U. S. DEPARTMENT OF AGRICULTURE. 
Examination of the fruit produced in this experiment showed that 
both normal and hollow fruits were to be found on every plat. Com- 
plete counts could not be made, owing to the destruction of the 
vines by a flood before the end of the season, but enough observa- 
tions were made to show that within the limits used varying quantities 
of fertilizer elements did not influence the production of hollow fruit. 
No positive results were obtained in this study showing the cause 
of puffiness in tomatoes, but the evidence indicated that the con- 
dition is not correlated with any considerable differences in the 
chemical composition of the mature fruit. The phenomenon is 
probably physiological in its nature, for the same varieties which 
show it in Florida are said not to do so, or only to a very slight extent, 
when grown in Michigan. A great difference that immediately 
occurs to one between conditions in the two places is that in Florida 
the crop is produced only through heavy annual applications of 
commercial fertilizers, which are not used in Michigan. Puffiness 
may therefore be dependent upon an unbalanced soil solution, but, 
if so, none of the variations in the fertilizers just enumerated sufficed 
to restore a proper condition. It is, of course, not inconceivable that 
puffiness is of a genetical nature and due to somatic variation. If so, it 
might, in conformity with the observed facts, be much more frequent 
in some varieties than in others, and the same plant might show both 
normal and "puffy" fruit. The whole subject is one which needs 
investigation. 
ADDITIONAL COPIES 
OF THIS PUBLICATION MAY BE PROCURED FROM 
THE SUPERINTENDENT OF DOCUMENTS 
GOVERNMENT PRINTING OFFICE 
WASHINGTON, D. C. 
AT 
10 CENTS PER COPY 
