148 
AMERICAN JOURNAL OF BOTANY 
IVol. 9, 
State of unstable division like that of Piersoni but even more pronounced. 
It was the first of all the Boston fern sports to be reported, but seems always 
to have been so unstable and so much less attractive than Piersoni that it 
never gained much general popularity. In well developed plants about 
half the leaves show the double pin nation, but this is never so completely 
developed as in good Piersoni leaves. The best leaves are partly once 
pinnate, and most of the divided pinnae are incompletely divided. At the 
present day, twenty years after its appearance, there has come down one 
strain attributable to Anna Foster which shows only an occasional twice 
pinnate leaf. Such a plant presumably represents reversion from the more 
completely divided original type. 
Harrisi and Roosevelti. These two varieties, possessing almost identical 
leaf forms, may be dealt with together as far as possible reversions are con- 
cerned. No reversions are known. The conditions in this respect are 
almost exactly like those discussed in connection with the possible complete 
reversion of Piersoni to a once-pinnate form. Florists who grow the rufifled 
forms almost always grow hostoniensis as well, since the cultural conditions 
required are the same. If a complete reversion should occur, it would pass 
unnoticed or be interpreted as a chance hostoniensis plant accidentally 
associated. 
Scotti. Exactly the same conditions prevail with Scotti as with Roosevelti 
and Harrisi. No reversions have been detected, not even occasional 
fluctuating leaves. 
Gretnai. The fish-tail form, Gretnai, shows fluctuations in the degree 
in which its pinnae and leaf tip fork, but the character is always present 
to some extent. No complete reversion to the normal Boston-fern type 
has been noted. 
Giatrasi. This fern represents one of the most distinct of all mutations 
from hostoniensis shown in the primary sports. Piersoni, while differently 
divided, has the same habit and size, and produces continually some similar 
leaves. Giatrasi, on the contrary, differs so markedly in size, habit of 
growth, and leaf characters that its separation as a "species" could easily 
be justified. Its leaves attain a length of no more than one third that of 
hostoniensis; the petioles are darker colored and wiry, those of hostoniensis 
being greenish and herbaceous; the pinnae are shorter in proportion, blunter, 
and wavier, the midribs are often sinuous, and the plant is notably slower 
in growth. 
Giatrasi (PI. VIII, fig. 2) does not produce any single fluctuating leaves, 
but it has produced as a runner sport (at the Giatras establishment) a form 
which may well be considered a reversion. This new form was introduced 
to the trade under the name of the ''New York'' fern (PI. VIII, fig. i). In 
characters it is intermediate between the Boston fern and Giatrasi, possessing 
in considerable degree all the characters of the latter fern except that it is 
considerably larger although never equalling hostoniensis (PI. VIII, fig. 3) in 
