280 ON THE EUKOPEAN SPECIES OF FESTUCA. 



diagnostic characters — ex. in F. arundinacea it is obtuse, though 

 the blade itself gradually narrows. Stomata are found only on the 

 upper surface of folded leaves, but on both surfaces of flat leaves, 

 though fewer in number on the under surface. 



Several pages are given to the consideration of the construction 

 and disposition of the tissues of the blade, and we find Schwendener 

 very frequently quoted, pp. 24-32. 



The author lays much stress on the fact that certain characters, 

 which are very variable as regards some organs, are on the other 

 hand constant as regards other organs ; for example, the hairiness 

 of the leaves is variable, while that of the ovary is constant. As 

 regards the spreading of the panicle during flowering time, the 

 close panicle is much rarer than is usually admitted. Grenier and 

 Godi-on describe F. tenuifolia as differing from F. ovina in having a 

 close panicle when in flower, whereas both have an open one at 

 that time. F. pulcheUa var. lAicata, the panicle of which remains 

 open after flowering, offers a remarkable exception to the usual 

 rule, that all Festuccc have a more or less close panicle, except 

 when in flower. 



Some authors describe the rachis of Festuca as universally 

 triangular, which is an error. F. montana and sylvatica offer 

 exceptions ; the panicle-branches of these species are also roundish, 

 not triangular. 



The presence or absence of an awn to the pale, and the form of 

 the latter, offer characters of little value for diagnosis. 



The author adopts the view of Turpin, Bentham, &c., as regards 

 the homology of the spikelet and flower, recognising the two outer 

 glumes as barren glumes, the outer pale as a fertile glume, and 

 the inner pale as the only true pale (proj)hyllumj. The lodicules 

 Prof. Hackel looks upon as representing a single organ (a second 

 prophyllum). My own observation and study in this direction 

 have led me to adopt a similar view, and to consider the true pale 

 as a single organ and the two lodicules as another single organ, the 

 former (the pale) being of the nature of a bract, the latter (the 

 lodicules) representing part of the true floral whorl. In Bromus 

 I have found all three stamens situate on the inner face of the 

 Iodides, two stamens on one Iodide, and one stamen on the other 

 Iodide. A paper by Prof. Hackel, alluded to by Mr. Bentham as 

 a very able one ( Journ. Linn. Soc, xix., p. 23j, will be found on 

 this subject m Engler's ' Botanische Jahrbiicher' (i., p. 336). 



Prof. Hackel finds little difference in the stamens, except in 

 size. The pollen grains are always similarly formed. 



The ovary offers valuable diagnostic characters, and the 

 Professor gives it as his opinion that on the whole the parts of 

 the flowers, with the exception of the ovary, are of much less value 

 for diaynosis than the leaves ; and he instances the florets of F. ovina 

 and F. rubra, two widely distinctive species, but which cannot be 

 separated by their florets. 



The caryopsis in two of Prof. Hackel' s sections, viz., Sect. 4, 

 VaricB, and Sect. 6, Montana, is stated to be universally free, or 

 attached only at its base to the pale. 



