300 PROCEEDINGS OP THE 



mildews are. A destructive parasitic fungus could not so readily 

 survive on living plants, if it confined all its operations to one, for 

 destroying the species it would itself perish. It seems, besides, to 

 be of advantage to such a parasite to have a secure retreat at one 

 stage of its existence, where, without proving very injurious to the 

 host plant, it may develop its cluster-cups ; and, if we assume 

 these to be the sexual generation, the importance of such a nesting- 

 place is apparent. Having obtained a certain stability of this 

 kind it can freely exert its destructive powers on plants other than 

 this host, without danger to its own existence as a species. Para- 

 sites, like Ustilago segetum, that usually only attack a limited 

 number among many plants of the same kind, or that are not very 

 destructive, may pass L their whole life on one individual without 

 bringing their existence as a species into peril. This moderation 

 is for them a path of safety. The more destructive parasites, 

 however, have an advantage in being heteroecismal, and by curbing 

 themselves at one time of their life may safely break through all 

 restraints at another. Parasites must have powers of destruction 

 co-ordinated with the security of their own existence as species, or 

 they must perish. Cluster-cups commonly attack temporary organs 

 — leaves generally — as if they meant no permanent harm to their 

 host-plants; while rusts and mildews are usually destructive of 

 the plants themselves on which they prey. For instance, Roestelia 

 cornuta does little harm to the rowan-tree, only attacking the 

 leaves ; but the Gymnosporangium affects the juniper differently, 

 as it preys on the permanent parts of the shrub, and year by year 

 silently but surely brings it to destruction. 



It suggests itself as a coincidence worthy of note, that while 

 among the Lepidoptera, which depend for food on living vegetation, 

 the caterpillar subsists on a plant usually different from that in 

 which the winged insect finds its food of nectar, among the 

 Coleoptera that depend on decaying matter, beetle and larva 

 alike generally feed on the same substance. The reason for this 

 arrangement seems clearly to be that it would prove awkward were 

 a voracious caterpillar forced to restrain its appetite for the plant 

 on which it preys, lest it should damage its chances of food when 

 it becomes a creature with wings. Thus, it is arranged that the 

 caterpillar shall fill its place in the world as a glutton, and that when 

 it reaches the perfection of winged existence it shall sip nectar 

 from plants that, as a caterpillar, it never attacked. 





