THE ORCHIS FAMILY. 159 



tive position of anther and stigma, and endeavor to make 

 out for yourself how this arrangement prevents the appli- 

 cation of its own pollen to the stigma of a given flower, 

 and at the same time favors cross-fertilization. 1 



RELATIONSHIP. 



This large family of plants includes about three thousand 

 species, widely distributed in both hemispheres, and show- 

 ing the highest specialization of the flower yet attained in 

 the vegetable kingdom. Many of the most conspicuous and 

 curious kinds are tropical epiphytes, and are frequently 

 cultivated in conservatories. As Miiller points out, the 

 family is remarkable for the great differences of habit 

 exhibited by the different species, the extraordinary modi- 

 fications of its flowers, and the great number of seeds 

 produced in a single fruit. The differences of habit, some 

 being epiphytic, others saprophytic, and so on, indicate 

 great capacity of the vegetative organs for variation, and 

 the modifications of the flowers are manifestly correlated 

 with the visits of insects. Cross-fertilization is the rule, 

 but here again " orchids show the greatest possible differ- 

 ences, all of which, however, are linked together by inter- 

 mediate conditions. We find in this order, cleistogamic 

 flowers and o'pen flowers ; flowers regularly or occasionally 

 self-fertilized ; others never self-fertilized, though quite 

 fertile to their own pollen if it be applied artificially ; 

 flowers absolutely sterile to their own pollen, though fer- 

 tile not only to the pollen of their own species but even to 

 that of other species of their own genus ; finally, species 

 in which pollinia and stigma of the same individual act 

 as fatal poisons to one another." 2 



1 Cf. Gray, How Plants Behave. 

 2 Miiller, I.e., pp.527, 528. 



