THE CRTTCIFEILE. 153 



each case there are four sepals and four petals. There 

 is no cohesion in any of these flowers, unless you ex- 

 cept the spuriously syncarpous pistil (Fig. 230). They 

 are also without adhesion. I do not know how suc- 

 cessful you may be in observing the embryo, but, with 

 a good magnifying-glass, you should be able to see 

 that the radicle is folded upon the cotyledons, some- 

 times against their edges, sometimes against the back 

 of one, but always folded. Now, these invariable 

 features are the ordinal characters of the cruclferse. 

 You may identify any one of the sixteen hundred 

 known species by the three features first named, and, 

 when you have done this, you may safely infer the 

 existence of all the others. You are enabled to 

 do this because botanists have carefully studied and 

 analyzed these plants, and in every case, along with 

 a cruciferous corolla, tetradynamous stamens, and 

 bractless inflorescence, the other features have inva- 

 riably been found. 



I wish to say a word about the Importance of the 

 characters by which you determine whether a plant is 

 or is not a crucifer. Some of you may think it strange 

 that such features as the length of stamens and the 

 absence of bracts should be named in describing an 

 order of plants. These points of structure would not 

 be looked upon as ordinal characters but for one cir- 

 cumstance, to be carefully borne in mind. It is their 

 constancy, which here gives them value. They take 

 rank from their permanence. Permanent or constant 

 characters, no matter how trivial otherwise consid- 

 ered, are of high value in classification. 



Order II. of Chart I. exhibits the characters of 

 the cruciferse as here described. 



