yellow ochre colour with a very pale border, and produce a 

 pretty effect. 



Fig. 1. represents the inside of a corolla, showing- the very 

 peculiar stamens ; 2. is the pistil, seated in the five-lobed cup; 

 3. is a piece of the old wood and flowers, natural size ; 4. is a 

 reduced view of the entire plant. 



The Bignoniaceous order, although we know so little of 

 it in this country, has, like almost every other old group of 

 plants, grown up so as to have lost all resemblance to its former 

 self, since it was first suggested. What was once scarcely 

 more than a genus of 18 or 20 species has become a large 

 natural order, in which nearly 400 species are arranged under 

 46 genera. Of these more than three-fourths are American, 

 and none actually European ; Asia contains thirty or forty in 

 the tropical districts, and Africa rather fewer. In a recent 

 attempt at reducing this mass into order, M. DeCandolle has 

 divided it into two parts, the Bignonieae and the Crescentinoe. 

 " Under the first of these tribes," he says, " I unite all the 

 Bignoniaceae whose fruit splits into valves, and whose seeds 

 are winged ; under the second are placed those whose fruit 

 does not open, and whose seeds are wingless. These charac- 

 ters are important and natural. In fact, in the whole vege- 

 table kingdom winged seeds are never found in a seed-vessel 

 that does not burst. And this is a fresh instance of the neces- 

 sity of those correspondences in organs of which we find so 

 many instances in the animal kingdom. Wings, which are 

 intended to assist the dispersion of seeds through the air, 

 cannot exist in fruits which do not naturally burst, but which 

 remain constantly closed, allowing their seeds to escape by the 

 decay of the tissue, or even to germinate in the seed-vessel 

 itself. This division of the Bignoniaceous order into two 

 tribes, founded upon both anatomical and physiological cha- 

 racters, appears then to be perfectly natural; it is indeed not im- 

 probable that the Crescentinae may be regarded as a family, 



when they shall be better known. At present this division 



consists of but 21 species. Its fruit is fleshy, leathery or 

 woody, indehiscent, and the species are very rarely climbers ; 

 for out of the eight genera that compose the tribe, there is 

 but one that has the latter habit. Perhaps also the Crescen- 

 tinae differ from Bignonieae by their fleshy, not leafy, coty- 

 ledons ; at least this character is indicated by the younger 

 Gaertner in Crescentia cucurbitina, the only species of the 

 tribe the seeds of which are well known." 



