including a new Arrangemei^ af PhantsrogamoyfS Plants. 453 



of the ovary of Labiatse, and the relation of the unequal stigmas 

 of Lippia to the cells leaves no doubt but that here the ovary 

 consists of a single carpel always anterior. (PL XV. figs. 1, 3, 

 3, 4 & 5.) 



MiMOSE^. Having observed carpels posterior in three species 

 of Acacia in which the inflorescence was dense, it was thought 

 possible that this might arise from the flowers growing three in 

 an axil ; this however is no longer a question, as carpels posterior 

 occur where the axil is always one-flowered ; but the posterior 

 position does not predominate, as in Cerasus Laurocerasus ; and 

 in three species having a globular inflorescence the carpel is 

 always anterior or lateral, which may be explained by the carpels 

 posterior being confined to the upper half of the spikes or almost 

 exclusively so. 



In two instances of dicarpous ovaries occurring in Phaseolus 

 vulgaris the ovary was syncarpous with parietal placentae (the 

 carpels in their upper part being separate and divei'ging in a 

 manner somewhat analogous to those of a Reseda, but to a 

 greater extent), thus showing an approach in Papilionacese to 

 Moringese and Polygalacese, as probably does Ormosia with two 

 unequal stigmas. But in an instance of three carpels occurring 

 in an Acacia, they were quite disunited. 



Proteace^. In Proteacese with capitate flowers, such as 

 Mimetes and Leucospermum, the carpel is always anterior, and its 

 apparently lateral position in Banksia seems owing to the axils 

 being two-flowered. And in Grevillea, where incomplete resu- 

 pination sometimes takes place, it is always anterior if the irre- 

 gularity of the flower is attended to, being alternate with the two 

 larger and longer sepals. But in Anadenia, where the flower is 

 regular, the carpel, although always anterior in the lower half of 

 the racemes, varies in its position towards the summit, and in rare 

 instances is perhaps posterior. The peduncles are however long 

 and the axils two-flowered, and carpels posterior are not so fre- 

 quent as in Acacia, being also confmed to the upper part of the 

 racemes. In Anadenia instances of two carpels occasionally pre- 

 sent themselves which are quite disunited, from which it may be 

 expected that the gland on the posterior side of the ovary in 

 Grevillea, &c., is a rudimentary carpel. 



Looking at the habit of Proteacese they might be expected to 

 be derived from Daphnacese, but the flower more nearly ap- 

 proaches that of Petiveria in having the cai-pel alternate with the 

 two anterior sepals, whereas in Daphnacese it is opposite a posterior 

 or lateral one ; and if Anadenia may be taken into comparison 

 they would agree in being apocarpous, i. e. supposing Petive- 

 riacese to be, like Phytolaccacese, apocarpous. 



Callitrichace.e. Callitriche is placed near Selaginese for 



