COMPOSITION OF CALIFORNIA LEMONS. 13 
natural products are under consideration. The samples varied 
decidedly from this average, some having a higher and some a lower 
specific gravity. It is necessary, therefore, to use mathematical for- 
mulas applicable to such cases, with the result that the chances are 
even that the true mean is not greater than 0.942 or less than 0.934, 
or, as it is expressed, 0.938 ±0.004. The same explanation applies 
to the other figures. The results reveal little difference in the com- 
position of these varieties of lemons. 
By applying other formulas it is possible to ascertain whether the 
differences shown are really significant, and, if so, to what extent. 
For instance, the odds are 78 to 1 that the difference between the 
specific gravity of Eureka and that of Lisbon lemons shown is sig- 
nificant. On the other hand, the odds are only even (1 to 1) that 
there is a significant difference between the specific gravity of Villa 
Francas and that of the Lisbons, and about 5 to 1 that the difference 
of 0.012 between the Eureka and the Villa Franca specific gravities 
is significant. It is probable also that no significant difference exists 
between the oil content of the Eureka and Lisbon varieties, nor be- 
tween that of the Lisbon and Villa Franca lemons. The odds, how- 
ever, are 18 to 1 that the difference between the oil content of the 
Eureka and that of Villa Franca lemons is significant. No significant 
difference is shown in the citric acid content of the varieties. 
As these averages are obtained from trees located in all parts of the 
lemon-growing area of California and from samples taken consistently 
throughout the year, there is little doubt that the data are repre- 
sentative of the actual composition of these varieties as grown in 
California. 
In this connection, it is interesting to consider some data obtained 
from analyzing 18 sets of samples of the Eureka and Lisbon varieties 
of different strains grown in two groves at Corona. These sets 
are derived from monthly samples taken over a period of two years. 
In considering them, the fact that all the Eureka trees were in one 
grove and all the Lisbons in another should be kept in mind, as this 
makes the data less desirable for comparative purposes than those 
from the field samples. The fact that these trees were chosen to 
illustrate differences between strains within their respective varieties 
rather than those between the two varieties also lessens their value 
for comparison. Interesting studies of the strains of these varieties 
have been reported by Shamel and his coworkers (1) (2). 
There are certain marked differences, however, that are not ap- 
parent between strains within the variety, but become apparent 
when the varieties are compared. To illustrate, the fruit of 10 
Eureka trees under observation had an average specific gravity 
varying from 0.925 to 0.989, the average being that shown in Table 
7. The fruit of the three Lisbon trees had specific gravities ranging 
