ORCHIDS AND EVOLUTION — COSTANTIN. 353 



failures they experienced. The yield of seeds is always irregular 

 and uncertain. Very often, without any apparent cause, successful 

 results are obtained in one greenhouse and complete failure in an- 

 other. Finally, although hybridizations had given first-class results 

 with Cypripediums, Cattleyas, and Laelias, there was a host of 

 genera famed for the splendor of their flowers whose seeds could not 

 be germinated. Such was the case with the Odontoglossums for a 

 long time. It was indispensable to find an explanation of these 

 anomalies. 



The riddle is solved to-day and in such a thorough fashion that 

 the consequent results for an industry of the first rank will be of 

 the highest importance. It was Imown fi-om the studies of Wahrlich, 

 published in 1886, that the roots of orchids invariably contain fungi. 

 The fact had been demonstrated long before by Schleiden von Res- 

 sek, Prillieux, and other botanists, but no one knew until Wahrlich 

 made his investigations that it was a condition existing in 500 repre- 

 sentatives taken at random in the orchid family, thus being a pe- 

 culiarity nearly universally characteristic of the group. It seemed 

 to me^ that important consequences must result from this verifica- 

 tion, and that the invasion of the roots of these plants by fungi 

 must have affected their evolution and been the cause of their pe- 

 culiarities of structure and modes of life. The statement, in par- 

 ticular, that in all the true holosaprophytes (plants without chloro- 

 phyll and bearing fungi in their roots) the seed was undifferentiated, 

 led me to affirm that the minuteness of orchid seeds, so marked 

 that they are described as scobiform, was one of the remote results 

 of the presence of mycorhiza and due in all probability to toxins, 

 which, acting at a distance, prevented the development and pro- 

 duction by the embyro, as in nearly all seeds of a radicle, a hypoco- 

 tyl, and cotyledons. 



This new point of view involved unexpected results: the seed 

 deprived of fungi must certainly be infected once it was placed in 

 the soil and it might be surmised that the failures of horticultur- 

 ists in their attempts to germinate seeds resulted from their ignor- 

 ance of the peculiarities which have just been set forth concerning 

 the biology of the plants. It was reserved for one of my students, 

 Mons. Noel Bernard, to gain the credit for the proof that these 

 deductions were quite correct. After having worked with the bird^s- 

 nest Neottia and tried vainly to germinate the microscopic seeds, 

 he had the good fortune upon a botanical excursion to find a cap- 

 sule of this plant which was bent toward the ground and had de- 

 hisced upon the soil. Its seeds had germinated. Upon studying 

 their structure he saw that they were infected by a fungus, in all 



1 Costantin, La nature tropicale (Blblioth6que scientifique internationale, p. 226), 1898. 

 44863°— SM 1913 23 



