I ! !•; N I-: !•■ J t • I A 1 . 1 x s 1-: c rs . 1 95 



the \car. It therefore ser\ es as a constant food-supply for 

 any possiljle insect tliat feeds on it, and the parasite is thus 

 enabled to develop without any serious check till it balances its 

 host. Under these conditions there would be ample scope for 

 the tendency to early maturity in certain individuals to make 

 itself felt in <^radually reducing the time necessary for the 

 completion of any given brood, and thus increasing the number 

 of broods per vear. Perfect control suggests an association 

 sufficientl}- long to enal)le the parasite to increase in fecundity 

 until it eciuals its host. 1\) my mind, this is precisely what has 

 taken ])lace under natural conditions in the case of Novius car- 

 dinal is. If this is correct, then it would seem that where control 

 is imperfect the association of species is so recent that the 

 parasite has not had time to come to perfection, and that by 

 ascertaining the determining conditions we should be able to 

 select and breed to type, and thereby hasten the establishment 

 of a balance between species — i.e., both host and parasite will 

 remain at the minimum. 



In the case of non-fluctuating s])ecies of insect pests (cer- 

 tain scale insects, for example). I believe that, given the men 

 and the equipment to cope with the work, control through the 

 perfection of natural enemies can be achieved. Host species 

 that from causes other than insect parasitism are subject to 

 violent fluctuations would be more difficult to deal with. 



From the practical standpoint the great dif^culty will be 

 to enable the desirable strains to become dominant. Where 

 the ordinary type of the same species is widely distributed, 

 there would be a tendency for the selected strains to deteriorate, 

 and hence they would have to be reinforced until they became 

 dominant. In the case of Aitlis fa-data, an area could have 

 been selected where it was not represented, and the desirable 

 strain liberated there, and reinforced until it gradually en- 

 croached on the weaker type in other areas. In fact, it w^ould 

 have been an advantage to destroy the natural type as the 

 selected strain progressed. 



In certain hosts (black scale, Saissetia olew and Hessian 

 fly, Mayetiola destructor, for example) the problem is compli- 

 cated through there being a number of species of parasites. If 

 the different parasites were perfectly supplementary, they 

 would no doubt be an advantage ; but it is conceivable that they 

 hinder each other through competition. In other words, they 

 are supplementary up to a certain point, and then become com- 

 petitive. This is an important consideration where natural 

 enemies are being sought for species that have by some means 

 been transferred unhampered to new surroundings. The ideal 

 would be to make exhaustive investigations in the original home 

 of the host and determine the parasite essential for satisfactory 

 control (Novius cardinalis, for example), and transfer it to the 

 new home of the host. 



