228 THE BIOLOGY OF FLOWERING PLANTS 



Loranthus celastroides merely falls on to lower branches, the 

 seed being squeezed out : wider dispersal depends on birds. 

 Peirce also considers the explosive fruit inefficient in 

 securing fixation. It should be borne in mind that the 

 " fruit " of the Loranthaceae is partly composed of the 

 hollowed axis of the flower. 



Throughout the family the nutritive relations of host and 

 parasite are, on the whole, the same as for Viscum. Only the 

 West Australian Nuytsia florihiinda and the small American 

 and Australian genus Gaiadendron grow independently as 

 trees. The rest of the family is parasitic. So far as is 

 known the majority are not very highly specialised as regards 

 their hosts, but in this respect only Viscum album has been 

 thoroughly studied, and the results make caution necessary 

 in drawing general conclusions. Weir (1918) has shown 

 that each species of the American genus Razoumowskia has 

 a limited range of host conifers. The leaves are usually well 

 developed, but in some species of Viscum, Phoradendron, 

 and other genera they are reduced ; in these, however, the 

 stems take on the work of assimilation and show ridges, or 

 flattening such as is common in other xerophytic plants. 

 Interesting is Viscum Crassulce, parasitic on the succulent 

 Crassulas and Euphorbias of arid regions of South Africa ; 

 like its hosts it is a succulent, with short internodes and very 

 thick orbicular leaves. In Arceuthohium minutissimwn, a 

 Himalayan species growing on Pinus excelsa, only the flowers 

 come above the bark of the host ; its tissues are otherwise 

 internal, and presumably it is completely parasitic. 



Of interest are the transpiration relations of such plants. 

 Our mistletoe, which is evergreen, grows on both evergreen 

 and deciduous trees ; Loranthus europceus is deciduous. 

 The leaves of the mistletoe are leathery and strongly 

 cutinised, and we have seen that xerophytic features are 

 shown by other species. Kamerling (19 14) has shown for 

 several tropical species, e.g. Loranthus pentandnis on Mangi- 

 fera indica in Java and Loranthus dichrous on Psidium guajava 

 in Brazil, that the transpiration of the parasite is more rapid 

 than that of the host, whether equal areas or weights of leaves 



