PREFACE. V 



liower witli a regular corolla, whereupon we look for 

 farther information in 534, where, as the leaves are in 

 whorls of four, etc., we make certain that it is called Ga- 

 Irum trifiduni^ L. 



Returning to § 77, to learn how to ascertain the names 

 of the three plants with one set of perianth, w^e are di- 

 rected to 537. JS'oticing that one of the three plants un- 

 der consideration has three to five flowers in a caljx-like, 

 funnel-form involucre, and that these flowers have a bell- 

 shaped perianth each, we find in 538 that it is Oxyhaphus 

 nyctagineus^ Nees. Of the two remaining plants wdth a 

 simple perianth, we may attack the grass-like one first. 

 Its flowers consist of an ovarj, three stamens, and two 

 pales (the perianth of grasses). We must start from 

 § 537, from w^iere, as there is no involucre to several 

 flowers, we proceed to 539. From here we are directed 

 to 564, where we are informed that we have a true grass 

 before us, and are required to consult § 565. Here, as 

 each spikelet contains a solitary, perfect flower, we are 

 advised to ask for further information in 566. Two 

 glumes being present, we proceed to 568, and, the flower 

 having two pales, to 570. The spikelets being not im- 

 bedded in hollows of the stem, we consult 571. As our 

 grass, with the spikelets of one sort, has four spikes placed 

 digitately on the summit of the culm, we proceed to 573, 

 where we ascertain that it is Cynodon Daetylon^ Pers. 

 Resuming § 539 (from where we must start in behalf of 

 the rest of our ])lants, as they have all incomplete flow- 

 ers), we inquire regarding the last plant with perianth- 

 bearing flowers. x\s it has a very conspicuous perianth, 

 we proceed to 564. As it is no grass, we must consult 

 639. The number of stamens not exceeding 12, we pro- 

 ceed to 640, and from there, the perianth being irregular, 



