395 



The strict relation that exists between Palms and Grasses has been already 

 adverted to in speaking of the former order : hence Nees considers Grasses to 

 be a sort of Palms of a lower grade. In reality, the habit of the Calamus and 

 Bambusa genera is nearly alike ; the inflorescence of Grasses may be consi- 

 dered to be the same as that of Palms, the floral envelopes of the latter taken 

 away, and only their bracteae remaining ; and, finally, their leaves are formed 

 upon exactly the same plan, with this difference only, that those of Grasses 

 are undivided. With Cyperacea;, however, it is that Grasses are most properly 

 to be compared : while a manifest tendency, at least to the degree of verticilla- 

 tion requisite to constitute a calyx, evidently takes place in the paleae of 

 Grasses, Cypcracea: are destitute of all trace of such a tendency, unless the 

 opposite connate glumes of the pistilliferous flowers of Carex, or the hypogynous 

 scales of certain Schasnus' and others, be considered an approach to the produc- 

 tion of a perianthium. For this reason, Grasses are to be considered plants in 

 a higher state of evolution than Cyperaceae. Independently of this difference, 

 the orders are readily known by the stems of Grasses being round, those of 

 Cyperaceae angular ; the leaves of Grasses having a ligula at the apex of their 

 sheath, which is split, while the sheath of Cyperaceae is not split, and is desti- 

 tute of this ligula ; and, finally, the embryo of the two is at variance both in 

 structure and position. With Asphodeleae their relationship consists in nothing 

 more than the tendency to branch which is observable in part of that order. 



Geography. As nothing can be uninteresting which is connected with the 

 habits of a tribe of such vast importance to man, I extract the following account 

 of the geographical distribution of Grasses by Schouw, from Professor Jame- 

 sons PhilosopliicalJournal for April 1825 : 



" The family is very numerous : Persoon's Synopsis contains 812 species, 

 l-26th part of all the plants therein enumerated. In the system of Roemer 

 and Schultes there are 1800 ; and, since this work, were it brought to a con- 

 clusion, would probably contain 40,000 in all, it may be assumed that the 

 Grasses form a 22d part. It is more than probable, however, that in future the 

 Grasses will increase in a larger ratio than the other phanerogamic plants, and 

 that perhaps the just proportion will be as 1 to 20, or as 1 to 16. Greater still 

 will be their proportion to vegetation in general, when the number of individuals 

 is taken into account ; for, in this respect, the greater number, nay, perhaps 

 the whole, of the other classes are inferior. 



" With regard to locality in such a large family, very little can be advanced. 

 Among the Grasses there are both land and water, but no marine, plants. They 

 occur in every soil, in society with others, and alone ; the last to such a degree 

 as entirely to occupy considerable districts. Sand appears to be less favoura- 

 ble to this class ; but even this has species nearly peculiar to itself. 



" The diffusion of this family has almost no other limits than those of the 

 whole vegetable kingdom. Grasses occur under the equator ; and Agrostis algida 

 was one of the few plants which Phipps met with on Spitzbergen. On the 

 mountains of the South of Europe, Poa disticha and other Grasses ascend al- 

 most to the snow-line ; and, on the Andes, this is also the case with Poa malu- 

 lensis and dactyloides, Deyeuxia rigida and Festuca dasyantha. 



" The distribution is of greater importance. As to the chief groups and spe- 

 cies, their distribution will not attain a real interest until we shall be in posses- 

 sion of a perfect natural classification ; for in this respect we are still, in my 

 opinion, far behind. The division of Beauvois appears to me too artificial, and 

 in that of Brown the groups Paniceae and Poaceae are too large. The best, 

 perhaps, is that of Kunth, according to which the Grasses are arranged under 

 ten groups. In respect to latitude, the relation of the Grasses, in the system 

 of Roemer and Schultes, in the hot and temperate zone, is the following : 



