DR. M. T. MASTERS ON THE PASSIFLORACE E. 607 
stamens, and a one-celled ovary with three styles. To what cause this augmentation was 
due, whether to an absolute development of six tubercles instead of five, or to a sub- 
division of one into two, I had no opportunity of ascertaining. 
A. transition from the verticillate to the spiral arrangement was seen in a flower of 
P. alata; or, rather, it may be said that the distance between the sepals was greater than 
it usually is. "Thus, of the five sepals two were smaller, and on a lower level than the 
remaining three. The petals of this flower were six in number, four free, two united the 
one to the other. Considering the successive manner in which the sepals are developed - 
under ordinary circumstances, it may be held that the unusual interspace between them 
was occasioned by an excess of develop tin the receptacle of the flower, which lengthened 
more than usual during the development of the sepals. On the other hand, the sepals 
themselves, from their smaller size, seem to have experienced an arrest of development. 
The sixth petal, from its position between two other petals, and in front of a sepal, may be 
considered the sole member of a second row of petals. 
In Passiflora Raddiana (kermesina of gardens) I have now frequently seen all the 
flowers on the plant presenting adventitious growths between the stamens and the ovary. 
The growths in question were like the filaments of the ordinary corona, and apparently 
emerged from the stalk of the ovary immediately above the spot whence the filaments 
separate from it, and as it were in the axils of the filaments, superposed to them in fact. 
In the Kew Museum is preserved a similar specimen from the same plant; but in this 
instance some of the adventitious filaments bear an anther. Have we here a tendency to 
manifest the same polyandrous condition that is found in Smeathmania &e? Karsten 
founded a new genus (Poggendorffia) upon a similar adventitious production of thread-like 
growths above the ordinary stamens*. | 
In a garden-variety of Passion-flower, x P. hybrida floribunda, 1 have met with the 
same malformation; and in other flowers of the same variety I have observed the con- 
nective of the anther to be prolonged into a crest as in Violacee. We can hardly lay 
much stress on an aecidental ease like this as an evidence of affinity between the two 
orders; nevertheless it at least adds to the acknowledged marks of relationship between 
the two. . 
To Mr. Miers I am indebted for a drawing of the fruit of Passiflora quadrangularis, 
wherein small flower-buds were found springing from the placentas intermixed with the 
seeds. These flowers were monstrous in character. A similar production of flowers within 
the fruit oceurs occasionally in Cruciferst, and has been also previously recorded in 
Passiflora. 
But the most remarkable deviation is that described in the Transactions of this Society 
by Mr. S. J. A. Salter, in which the ovules of a species of Passiflora were found to 
contain pollen 1. 
* Karsten, Flor. Columb. Spec. Select. i. t. 15. f. 29. 
+ Masters, Vegetable Teratology, p. 137; Bernouilli, Bot. Zeit. 1869, p. 19, durchwachsene Frucht von Passiflora. 
+ Trans. Linn. Soc. xxiv. 143, t. 24. See also * Vegetable Teratology, p. 185, f. 99 ; and Seemann's Journ. Bot. 
1867, t. 72, for a similar instance in Rosa. 
