156 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9, 
form, and then to return in the next generation either to the original typical 
form or to one with reduced stability of division. The point is, there are 
cases of apparent reversion in which the new {i.e., reverted) form seems 
incapable of reproducing its new type. Its runner progeny do not "breed 
true" but return in different degrees to the form from which the reversion 
originally sprang. The possibility suggests itself that such unstable varia- 
tions may correspond in kind to the variations noted as probably due to 
seasonal changes. It may be that as reverted leaves appear to be more, 
frequent during winter, long continued observation might also show that the 
temporarily reverted runners, the sort incapable of reproducing their kind, 
may also have a seasonal frequency. 
2. Discontinuous variation in a regressive direction affecting the leaf 
division has been found in a number of forms: in elegantissima, to robusta, 
a Piersoni-Yike form; in elegantissima-compacta, in two steps through Wana- 
maker and to a once-pinnate form; also in elegantissima-compacta, in one 
step, to "Dwarf Boston," once pinnate; in superhissima, to viridissima. 
3. Discontinuous variation, reversion in size: in elegantissima-compacta, 
to Wanamaker, and to a Piersoni type; in viridissima, to a taller form, inter- 
mediate in size between viridissima and hostoniensis. 
4. No evidences of reversion to a fertile condition have been noted. 
The characteristics and behavior of reversions from tertiary and higher 
degree progressive sports of hostoniensis will be left to a future paper for 
description. 
Brooklyn Botanic Garden 
EXPLANATION OF PLATES 
Plate V 
Fluctuating reversion: pinnae showing range of fluctuation in three forms. 
Figs. 1-5. Pinnae from one leaf of Piersoni, ranging from typical twice pinnate divi- 
sion (fig. i) through intermediate forms (figs, 2-4) to once pinnate type (fig. 5) 
Fig. 6. Simple pinna from another leaf of the same plant of Piersoni as shown in 
figures 1-5. 
Figs. 7-10. Pinnae from one leaf of "Anna Foster.'' 
Figs, ii, 12. Pinnae of Anjia Foster, taken from another plant. 
Figs. 13-17. Pinnae of fertile Pi^r^owi-like reversion, var. /erfe'm, showing same sort 
>i fluctuation as in true Piersoni. 
Plate VI 
Regressive mutations: leaves of twice pinnate reversions from difl"erent progressive 
forms. 
Fig. I. From Smithi, a four-times-pinnate form. 
Fig. 2, From elegantissima-compacta, itself twice pinnate, pinnatifid. 
Fig. 3, Also from elegantissima-compacta, but a distinct form. 
Fig. 4. From superbissima, a reversion in size (see also Plate VIII). 
Fig. 5. Robusta, introduced by Robert Craig, Philadelphia; a reversion reported 
from elegantissima. 
Fig. 6. Var. fertilis, possibly reversion from Piersoni. 
