ULMAOEM. 171 



filaments short and straight. Female calyx gamophyllous, 

 cupnliform. Ovule campylotropous. Fruit dry induviate and in- 

 dehiscent. Seeds without albumen, embryo recurved. — Odorous 

 herbs, annual and erect or evergreen and climbing, with aqueous 

 juice, leaves opposite (at least in the lower part of the stem), 

 scabrous, palminerved, often lobate, stipules persistent, inflorescence 

 in cymes. — 3 genera. 



A family thus constituted, " by concatenation," has manifold 

 affinities. Lindley has placed the Ulmece in his Alliance of 

 Rhamnahs, between the Aquilariece, which belong to the Thymelacece, 

 and the Chailletice, which we include in the EuphorliaceoB. We are 

 not unmindful of their affinity to the Urticece. Now, although 

 the Artocarpece and the Morece have been more or less widely 

 separated from the Urticece on account of the constitution of their 

 dicarpellar gynsecium with ovules oftener descending and ana- 

 tropous, and although we ourselves have formerly admitted this 

 separation, the study of a large number of genera of the small 

 group of Conocephalece, in which the ovule is more or less distinctly 

 ascending and orthotropous, we are convinced that this disjunction 

 cannot be maintained and that the opinion of Mr. Bentham, restor- 

 ing to the older Family of Urticece its unity and integrity, ought 

 now to be fully adopted. We perhaps go still further than he does 

 in leaving the Ulmem and Artocarpece in one and the same natural 

 group;- but the Celtidece, on the one hand, could not be disjoined 

 from the Ehns, the fruit of Planer eoe being intermediate between 

 the drupes of Celtis and the samarse of Ulmus ] and, on the other 

 hand, they could not be separated from Morece and Artocarpece by 

 any truly absolute character, neither by the nature of their juice, 

 nor by the characters of their stipules, stamens, gynsecium and 

 fruit ; and if the differences which have been put forward for this 

 purpose really exist, there is not one which, in every ease, can be 

 considered constant. On the other hand, the family we are now 

 studying approaches, as we shall presently see, very near to the 

 Castaneacece and, consequently, to the Hamamelidece and Platanece. 



TJsEs. — The milky or opaline juice found in a great number of 

 Artocarpece give them very characteristic properties,' analogous to 



1 Endl. JEnehmd. 168.— Lindl. Veg. Kingd. ii. 320. — Eobenth. Syn. PI. Dia^hor. 196, 

 270 ; Pi. d. 301.— GuiB. Drog. Simpl. 6i. 6. 1108. 



