is frequently the case, with dialysis or separation of the petals.[225] Many of the cultivated Gloxinias also show erect, regular, five stamened flowers, but these are probably cases of irregular peloria. [Illustration: FIG. 120.--Regular peloria, _Eccremocarpus scaber_.] A solitary flower of _Pedicularis sylvatica_ was found by the Marquis of Stafford near Dunrobin Castle in Sutherlandshire, in which the usual ringent form of the corolla was replaced by the form called salver-shaped. There were six stamens, four long and two short. Sir W. Hooker and Mr. Borrer are stated to have found a similar flower in the same locality in 1809.[226] The passage of ligulate to tubular corollas among _Compositae_ is not of such common occurrence as is the converse change. I owe to Mr. Berkeley the communication of a capitulum of a species of _Bidens_, in which there was a transition from the form of ligulate corollas to those that were deeply divided into three, four, or five oblong lobes. These then were instances of regular peloria. [Illustration: FIG. 121.--Flower of _Cattleya marginata_. Lip replaced by a flat petal.] In _Orchidaceae_ a similar change is not by any means infrequent; in a few, indeed, a regular flower is the normal character, as in _Dendrobium normale_, _Oncidium heteranthum_, _Thelymitra_, etc. Fig. 121, reduced from a cut in the 'Gardeners' Chronicle,' 1854, p. 804, represents an instance of this kind in _Cattleya marginata_. From the same journal the following account of a case of peloria in _Phalaenopsis Schilleriana_ is also cited as a good illustration of this peculiar change. The terminal flower differed entirely from all the others; instead of the peculiar labellum there were three petals all exactly alike, and three sepals also exactly alike; the petals resembled those of the other flowers of the spike, and the upper sepal also; but the two lower sepals had no spots, and were not reflexed as in the ordinary way: thus, these six parts of the flower were all in one plane, and being close together at their edges, made almost a full round flower; the column and pollen-glands were unaffected. Professor Reichenbach also exhibited at the Amsterdam Botanical Congress, of 1865, a flower of _Selenipedium caudatum_ with a flat lip. M. Gris[227] has placed on record some interesting cases of peloria of this kind in _Zingiber zerumbet_; in the more complete forms the androecium or staminal series was composed of six distinct pieces, the three inner of which were fertile, while in the ordinary flower the androecium is composed of two pieces, "a lip" and a fertile stamen. "Is it not a matter of regret," says M. Gris, "to be obliged to call the latter the normal flower?"