March, 1922] BENEDICT — NEW VARIETIES OF NEPHROLEPIS 
The reversions of elegantissima-compacta described above all belong to 
the category of stable forms, presenting distinct and definite changes in 
leaf form and division. The kind of reversion mentioned earlier in the 
paper as reduced stabiHty of division (''fixed instability") has also occurred, 
and one case has been under continuous observation and cultivation for 
several years. 
A plant of typical elegantissima-compacta, obtained directly from its 
place of introduction, the Pierson establishment, was one of the original 
group of bostoniensis varieties to be received in 191 5, and in fact was desig- 
nated "no. I " in my accession series. With the other forms then at hand, 
it was planted out in open soil in a bench for reproduction by runners. 
Among its numerous progeny in that first cultivation was one small plant 
with leaves almost entirely once pinnate. In the spring of 1916 (March) 
this was planted in a large square pan for further propagation with the idea 
that it might prove to be a stable once pinnate reversion. At this time it 
consisted of six leaves (PI. IX, fig. i, shows a similar plant), entirely once 
pinnate with the exception of one which was shallowly lobed on some pinnae. 
By August the pan had become full of small plants which were separately 
potted and grown for eleven months. Although the stock crown had been 
almost completely once pinnate, these runner plants were in general like 
typical elegantissima-compacta in leaf division, with scattered once pinnate 
leaves mixed in. In July, 191 7, the stock consisted of fourteen plants in 
four- and six-inch pots, but not in the best condition, owing to crowding. 
Three sorts of plants could be distinguished among them. One plant with 
fifteen or more leaves was almost entirely once pinnate, only one or two 
leaves showing any double division. There is little doubt that this repre- 
sented the original once pinnate crown first detected, but this cannot be 
absolutely assured. Nine plants were mainly twice pinnate as in typical 
elegantissima-compacta, but each had one or more once pinnate leaves. 
Four plants had leaves practically without any reversion. 
The cultural conditions had been the same for all, so that the modifica- 
tions observed must have taken place as a result of internal changes. Ex- 
amples of the three sorts were kept for further observation, but their sub- 
sequent behavior was the same. The plant with mainly once pinnate 
leaves continued once pinnate in its original crown, but the secondary 
crowns produced in the same pot by runners to the number of eight had 
leaves mainly of the twice-divided type of elegantissima-compacta. In other 
words, the original once pinnate ciown, produced as a runner variation, 
retained its leaf characteristics as its own new leaves developed, but gave 
rise only to new plants with more divided leaves, either like typical elegan- 
tissima-compacta or, in some cases, with the less stable amount of division, 
but with no real modification of leaf form (PI. X, figs. 6-12). 
This behavior is worthy of emphasis because it has been found to be of 
frequent occurrence in the study of reversion among other division forms. 
