85 



Text-Book of Grasses (1914), Ijut a fuller and more adequate 

 treatment was urgently needed. Professor Hitchcock's sound 

 and sane conservatism has not permitted him to depart too widely 

 from the fundamental doctrines of modern agrostology yet the 

 present work is by no means a mere compilation of existing views, 

 but marks in several respects a distinct advance over our previous 

 knowledge. 



The attention of the reviewer naturally was first drawn to that 

 stone of stumbling and rock of offense, the correct position of the 

 tribe Oryzeae. The Gordian knot has been neither untied nor 

 cut. After following Hackel and Scribner by placing the tribe 

 in the sub-family Panicatae in the Text-Book of Grasses, Pro- 

 fessor Hitchcock has now returned to the view taken in his re- 

 vision of the Gramineae for the Seventh Edition of Gray's 

 IManual (1908), and included the tribe again among the Poatae. 

 Evidently therefore the laterally-compressed spikelets now appear 

 to him a character of greater significance than the articulation 

 of the rachis hclotv the glumes. As a matter of fact, the tribe pre- 

 sents an impasse that can never be satisfactorily evaded as long 

 as the two sub-families are delimited as at present. Undoubtedly 

 there will always be good grounds for maintaining these two 

 series for the majority of the genera; but there is a progressive 

 obliteration of sharply-opposed characters as we descend toward 

 the median line, until we reach a debatable ground in which the 

 two sets seem to be inextricably blended. Perhaps the most note- 

 worthy advance in taxonomy afforded by the present volume is in 

 the new sequence of tribes. The arrangement that has been uni- 

 formly followed hitherto has been strikingly illogical, in that it 

 involved a progression from the most highly-developed to the 

 most primitive forms. The bamboos, as showing the least differ- 

 entiation in floral structure, should evidently begin the sequence, 

 and the allies of Tripsaciim should close it as the most complex 

 Wg accordingly find in the present work that the Poatae stand 

 first, with the tribes in the following order : Bamboseae, Festu 

 ceae, Hordeae, Aveneae, Agrostideae, Nazieae, Chlorideae, Pha^ 

 larideae. Oryzeae and Zizanieae (the latter tribe cut off from 

 Oryzeae on the basis of the unisexual spikelets, leaving only 



