August 21, 1908] 



SCIENCE 



229 



so great that accuracy even of observation 

 will be impossible, but the failure to 

 reckon with all the factors of the prob- 

 lem will make conclusions of little signifi- 

 cance. 



The interrelation of factors may be of 

 the most complicated nature; for instance, 

 a parasite which of itself might be wholly 

 inefficient due to its slow rate of reproduc- 

 tion as compared with that of its host, 

 might be rendered very efficient by the 

 cooperation of a contributing factor which 

 could only delay the rate of increase. 



It will be thus readily seen that the 

 efficiency of all these factors working 

 together is neither the sum nor the average 

 of the potential efficiency of each, though 

 much nearer the latter than the former. 

 Many writers have assumed that by adding 

 a new parasite, its efficiency was simply 

 added to that of others previously existing. 

 This supposition is certainly far from the 

 theoretical conception of the interrelations 

 of species as presented above, and has not 

 been borne out in actual experience. 



RELATION OF LIPB CYCLE OF HOST 



Thus far the insect whose control is 

 sought is conceived of as existing in but 

 one condition. The growth and trans- 

 formation of insects add still further com- 

 plications to the subject. The checks are 

 not simultaneous in their action, but at 

 each stage in the progress of its develop- 

 ment the insect lives in a different environ- 

 ment. The parasites, for instance, that 

 affect the egg will find the next generation 

 of eggs perhaps more profoundly influ- 

 enced by the checks that have operated 

 during the remainder of the life of the 

 insect than anything they have accom- 

 plished, and so perhaps with the checks 

 operating at any stage. A serious attack 

 of one parasite during early larval life 

 might result in protecting the insect from 

 still more efficient destroyers in the late 



larval stage and really cause more to come 

 to maturity. 



SUGGESTIONS FOR LABORATORY STUDY 



We can eliminate most causes of death 

 under artificial breeding conditions and 

 often produce one hundred per cent, of 

 survival. When this can be done we ai •" 

 in a position to begin the experiment of 

 testing first one at a time each cause of 

 death, then to study their interrelations or 

 the simultaneous or alternating effects of 

 two of these factors, in the case of parasites 

 studying as thoroughly in detail also their 

 environment. Until considerable work of 

 this kind is done the basis for our theories 

 will not have been well enough established 

 to deserve a place as science. 



ECONOMIC RELATIONS 



The power of an insect to do damage is 

 due as a rule to the number present during 

 their chief feeding period, and may be 

 quite independent of the numbers that 

 finally come to maturity, and is absolutely 

 independent of the ratio between birth 

 and death rates. A temporary disturbance 

 of this rate produces increase or decrease 

 and may place an insect suddenly in the 

 destructive class or remove it, but while an 

 insect maintains itself in injurious num- 

 bers the ratio is as low as though the insect 

 were rare. 



In the case of most of our injurious in- 

 sects the natural increase is more than a 

 hundred fold, so that less than one per 

 cent, is in these cases the established aver- 

 age rate of survival. This is true even of 

 such recently introduced pests as the 

 gjT)sy and brown tail moths, and the boll 

 weevil, everywhere, except when the condi- 

 tions are temporarily disturbed by efforts 

 at control and along the border of the in- 

 fested area where the insects are invading 

 new territory.^ 



'This invasion of new territory probably in- 

 volves but a narrow strip. In the case of the 



