18 BULLETIN" 697, U. S. DEPARTMENT OF AGRICULTURE. 
of lemon. This condition may be due, in part at least, to the more 
recent introduction of the Marsh variety and the comparatively 
small amount of propagation of it, so that only a few bud generations 
have been available for observation. Recent commercial propaga- 
tions of this variety have been made, as a rule, with the use of buds 
selected from trees of the Marsh strain and for the most part by the 
use of fruit-bearing buds, which explains the presence of only a small 
number of trees of the undesirable strains. 
The trees of the Unproductive strain, and the fruits borne by them, 
are similar to those of the Marsh strain, except that the trees usually 
bear relatively fewer first-grade valuable fruits and show a marked 
alternate-season bearing habit. 
INFREQUENT FRUIT VARIATIONS. 
Fruits possessing many minor variations of interest and of some 
importance have been borne by the performance-record trees in 
addition to the important strains already described. These varia- 
tions have not been found as frequently as has been the case in the 
other citrus varieties in the performance-record plats. Fruits are 
occasionally found that, from outside appearance, more nearly 
resemble oranges than grapefruits. Some of these variable fruits 
resemble in flesh and juice characteristics combinations of some of 
the qualities of oranges, tangerines, and grapefruits. 
Another variation occasionally met with is the ribbed fruit, an 
illustration of which is shown in Plate IX. In this case the fruit is 
usually bell shaped and the peel has regular longitudinal ridges and 
depressions, giving it a ribbed or fluted appearance. The rind 
is usually thick, the rag is rather coarse and bitter, the juice is 
abundant but of inferior flavor and quality, and the fruits are com- 
mercially seedless. These fruits are probably worthless for com- 
mercial purposes. 
Fruits having raised or sunken sections, or both, are of rather 
frequent occurrence, particularly on trees of the Seedy and Rough 
strains. 
Several fruits having well-developed navels have been found 
among the fruits on the performance-record trees. In some cases 
they were absolutely seedless, while in other instances the fruits 
contained several seeds. So far, no branch or tree has been found 
bearing only navel grapefruits. No attempt has been made as yet 
to isolate a strain of navel grapefruits through bud selection, although 
it seems probable that this can be done. The presence of a navel 
in the seedless grapefruit, other things being equal, would likely be a 
disadvantage except as a distinctive mark for the variety. 
Occasional limbs bearing absolutely seedless fruits have been 
observed in the performance-record trees. Other limbs have been 
