90 TAXONOMY. 



known to him, he observed six or seven species to have six 

 anthers, and the other two or tliree species to have only four, 

 he placed the genus in the Sixth Class, but referred, in what 

 he called the Key to his Classes, or in the Preliminary Gene- 

 ral Index, when speaking of the fourth class, to the Genus 

 Convallaria in the sixth. By this means the investigation 

 was completely facilitated. But it has happened, that later 

 discoveries have exhibited a multitude of species, in which 

 the subordinate numerical proportion has become the prevail- 

 ing one, in so much, that w^e must now transfer many plants 

 from the class in which Linna?us placed them into another. 

 This has happened with Verhena^ which Linntcus placed in 

 the second class, but which later botanists have justly placed 

 in the fourteenth. Thus also Bocrhavia, which Linnaeus 

 placed in the first class, is now transferred, on account of the 

 greater number of its species which have two anthers, into 

 the second. With respect to Rhexia, it is doubtful whether 

 it belongs to the eighth or tenth class, because we find nearly 

 as many species with eight as with ten anthers. 



In the second place, Linnaeus was accustomed to observe, 

 among several species of the same genus, which of them was 

 most common, and was produced in greatest abundance ; 

 that, according to these circumstances, he might detennine 

 its place in the system. Thus he decided with respect to 

 Lythrum^ the most common species of which, namely Ly- 

 thrum salicaria, has from twelve to fifteen anthers, whereas, 

 in other cases, the number is far smaller. But even here 

 Linnaeus was not consistent ; at least the most common species 

 of Euonymus is almost always observed to have four an- 

 thers, although the genus stands in the fifth class. Defective 

 observation, also, often led him into mistakes, which have 

 been propagated to our time. Thus Ruppia stands in the 

 fourth Linnaean Class, although no such numerical proportion 

 can be discovered in it. We thus also find Calla in the 

 seventh class, and yet the number of the filaments is wholly 

 indeterminate and fluctuating. We shall speak of Eupliorb'ia 

 upon another occasion, when we shall shew that it holds a 

 very unsuitable place in the eleventh class. 



