6 ON THE JESTIVATION OF CRITHMUM MARITIMUM. 
Systematie botanists who have dealt with the whole vegetable king- 
dom have been as unsuccessful in finding distinctive characters for the 
two Orders as our local botanists, and the confusion that is thus caused 
in general systematic works is embarrassing. After the grave errors 
introduced into the ordinal characters of Hederaces by Don, and co- 
pied by Bartling and De Candolle, had been corrected by Brown and 
Bennett in * Plantze Javanicæ, it became evident that Umbelliferze and 
Hederacez were not so distinct as had been supposed by botanists 
labouring under the belief that Don's errors were the result of true 
observations. Lindley, in his * Vegetable Kingdom,’ makes the fol- 
lowing distinctions between the two Orders:— ` 
- * Umbellifere. Fruit didymous, with a double epigynous disk. 
** Hederacee. Fruit not didymous, without a double epigynous disk, 
3- or more celled. Pentamerous flowers. Corolla valvate. Leaves 
alternating, without stipules." 
The principal distinctive characters here relied upon (the didymous 
or non-didymous fruit and the presence or absence of the double epi- 
gynous disk) do not stand the test of practical application. About 
one-half of all known Hederacez have a didymous fruit, and in many 
Umbelliferz the disk is not double if the styles are closely united, there 
being in that instance only one disk, as is the case in most Hederaceze. 
The pentamerous flowers are general in both Orders, tetramerous 
ones forming the exceptions. Alternate leaves are also a feature com- 
mon to both Orders, opposite ones being again the exception. Sti- 
pules cannot be denied to many Hederaceze, being, for instance, highly 
developed in Tetrapanaz papyrifera, C. Koch, the rice-paper plant. 
Unless some additional characters besides those derivable from the 
vstivation of the corolla can be found, it will scarcely be possible to 
make Hederaceze more than a suborder of Umbelliferze; The general 
name of Umbellifere might be retained for the whole Order, whilst 
that of Apiacez (following Lindley) might be adopted for one suborder 
and Hederaceæ for the other. The two suborders would oceupy the 
same relative position as do Clematidew and Anemoneæ in Ranuncula- 
‘cee, and Papilionaceze, Cresalpineze, and Mimoseæ in Leguminosre, all of 
which are distinguished by the sestivation. However, when the wile 
genera belonging here shall have been carefully examined, it may be- 
come necessary to establish even more suborders, In Trachymene 
cerulea (Didiscus, Mook.), the petals are vexillary in bud, exactly as 
-— 
