198 BOTANICAL dAZETTK. 



A little inconsistency or uncertainty is also to be noticed in the mat- 

 ter of the nomenclature of orders. We have Polemonlacea' and Gen- 

 fianece, RutacecB and Cactcir, Srn>]>hul(iriarew and ScwpJiularinen', etc. 

 In the existing confusion in such names we should have some stand- 

 ard and a work from the hands of such eminent botanists should be 

 consistent with itself. This is not meant as a censorious criticism 

 upon a work meriting so much praise, but as a plea for l)otanists 

 who, like the writer, are often bewildered in this maze of termina- 

 tions.— J. M. C. 

 t 



Revision of the North American Liliace.e, by Sereno Watson. — 

 Botanists will be delighted at the appearance of this paper, for if 

 any branch of plufuogamic botany needs revision it is the Monocoty- 

 ledons, We have so many works starling out with noble promise, 

 but stopping before tiie end is reached and hence the Monocotyle- 

 dons have been neglected. The work now done helps to till this gap, 

 and botanists should be congratulated that it has fallen into such 

 competent hands. The order Liliaccce. as now accepted, presents 

 such a diversity of characters that it is very difficult to group the gen- 

 era according to their affinities. The genera have here been grouped 

 in a manner that is entirely new, and, with a few exceptions, it works 

 like a charm. Baccate fruit as distinguished from capsular is consid- 

 ered a subordinate character, and a division into three suborders or 

 series has been made, based upon other characters. The first series is 

 distinguished by -^scarious Horal bracts, persistent nerved perianth, 

 perig^^ious stamens with introrse anthers, an undivided persistent 

 style, and a loculicidal fruit (if capsular.) Both of the other divis- 

 ions have the stamens hypogynous or nearly so, with more or less ex- 

 trorse anthers, and the tioral bracts are more or less foliaceous or are 

 wanting. Both also always have distinct perianth — segments and un- 

 jointed pedicels. But one has a nerveless deciduous perianth, the 

 styles (when present) more or less united, and the fruit a loculicidal 

 capsule or berry. The other division has distinct styles and a septi- 

 cidal capsule conjoined with a persistent nerved perianth." The sub- 

 division into tribes is based on the characters of the inflorescence 

 mainly. The first series contains tribes and 25 genera; the second, 

 3 tribes and 13 genera; the third, 4 tribes and 12 genera ; making in 

 all 10 tribes and 50 genera. In glancing through the paper we made 

 the following notes: AUinin. is the largest genus, containing 40 spe- 

 cies, 18 of which are new. For Hrsprrosr(yrdiu))i, (?) niafitrmum. of 

 Torrey is made a new genus, MallJii by name. Wood's Brevoortia is 



