238 THE BIOLOGY OF FLOWERING PLANTS 



opportunities for attachment in twining plants so cha- 

 racteristic of the bindweed family. The opportimity of 

 the Rhinanthoideae Ues simply in the abundance of roots of 

 other plants in the soil of meadows and pastures, and the 

 frequent contact with them which must be made. Given 

 the start of root parasitism, it is not difficult to picture 

 the evolution of the more advanced and specialised types, 

 culminating in such forms as Rafflesia. The problem of 

 the origin of the mistletoes is more difficult. One might 

 think of them as originating from epiphytes. But in all 

 the multitude of the true epiphytes of the tropics there seems 

 to be no indication of any sort of parasitism The distri- 

 bution of the Loranthaceae to tree branches is now secured 

 by the slimy berry, but whether this appeared before or 

 after the assumption of the parasitic habit we do not know. 

 It is quite conceivable that the Loranthaceae, too, started 

 as root parasites and became branch parasites later, for two 

 genera of the Santalaceae are branch parasites though root 

 parasitism is dominant in the family ; it is also quite 

 possible that they were originally lianas, for some tropical 

 species, as we have seen, have tendril-like roots and ram- 

 bling stems. We have, in fact, no means of deciding how the 

 parasitic Loranthaceae took to their present mode of life. 



Some efforts have been made to determine experi- 

 mentally what conditions will permit of parasitic existence. 

 It is quite possible to grow many ordinary plants for a time 

 in wounds of others. Peirce (1904) planted germinating 

 peas in slits of a bean stem ; the root system grew down the 

 hollow internodes and drew a supply of water from the walls 

 of the cavities. The peas reached maturity and bore seeds. 

 Molliard (1913) grew cress in wounds in the hypocotyl of 

 the French bean, in a saturated atmosphere, and found 

 that sucker-like side roots were formed, 



MacDougal and Cannon (1910) made an extended study 

 of the growth of several plants used as cuttings in wounds 

 of various cactuses. In several cases the cutting survived 

 for upwards of two years and showed considerable growth. 

 Agave produced so vigorous a root system as to destroy much 



