Oct. 22. X92t Hopkins Host-Selection Principle 219 



Furthermore, although not selecting chestnut when the other host was 



present, they produced a good colony when isolated on it; but in 19 18 



these adults again selected tulip in preference to chestnut. The same was 



true with the original gooseberry form which was transferred to chestnut 



(experiment XXVI ^), for in 19 18 it returned to gooseberry in preference 



to chestnut. 



SUMMARY OF RESULTS 



(i) In practically all the species experimented with the adults show a 

 marked predilection for the host in which they have fed as larvae, 

 provided they are not deterred by other factors, such as the unfavor- 

 able condition or the small quantity of the host. 



(2) There is considerable variation in the degree of preference for the 

 original host, as between different species. Thus — 



(a) Certain species are capable of living in only one genus or species of 

 plant, which consequently they select. 



(6) Certain species, chiefly those living in nature in several hosts, can 

 be forced to adopt a new host. 



(c) Certain species, chiefly those feeding in nature in a great variety 

 of plants, show little discrimination in the selection of hosts. 



(d) Certain species feeding in nature in a great variety of hosts often 

 show a preference for a few of these. 



(3) In forced transference of individual adults of a species to a new 

 host, a high mortality of the broods usually occurs, especially in the case 

 of eggs laid by beetles emerging from the original host, in which case 

 the mortality is often total. One-half to full-grown larvae, however, 

 usually can be successfully transferred to a new host and live and trans- 

 form to adults. 



(4) With some species that can be reared in a secondary (new) host, 

 by the larvae feeding one or part of one year, preference for that host is 

 shown by the resulting adults. 



(5) In general, the fewer the hosts in nature, the more marked the 

 predilection for a particular host, and vice versa. 



(6) Continued breeding in a given host intensifies the preference for 

 that host. 



(7) The condition of the host has a great influence on host selection, in 

 that every species prefers an optimum condition of the host which it 

 selects and will choose a new host in the optimum condition in preference 

 to an old host in which the conditions are unfavorable. 



(8) The quantity of wood at the disposal of the ovipositing adults 

 may influence the insects in their choice between different kinds of host 

 wood, in that, if there are many adults to a limited amount of the primary 

 host, some species will select a secondary host if such is available. If 

 this is done, however, the resulting brood is weakened. 



It is altogether possible that these experiments may indicate the origin 

 of certain closely related species or varieties of insects. For instance, a 

 species restricted to a very few plants may accidentally be forced to 



