AMERICAN ECONOMIC ENTOMOLOGY 223 



keeping the parasites in captivity and during transport by sea 

 cannot fail to be of great value. 



As it will be unnecessary to follow the authors through the 

 various stages of their investigations, we may turn at once to 

 the parasites finally selected as being those most likely to yield 

 satisfactory results. Two factors were of prime importance in 

 regard to this selection — first, that the species should be one 

 with a wide geographical range in order that it might be likely 

 to withstand the variety of physical conditions to be met with 

 in Hawaii ; and, secondly, that it should be endowed with great 

 rapidity of increase so that it might have a fair chance of coping 

 with an insect which had probably reached its average maximum. 



The choice fell on parasites which attack the eggs of the 

 hoppers ; but there were found to be two types of these egg- 

 parasites so eminently suited for the post that it became a 

 matter of difficulty to decide to which to award the preference. 



On the one hand are the tiny little hymenopterous egg- 

 parasites of the genera Anagrus and Paranagrus, collectively 

 known as myrmarids ; while on the other are those of the genus 

 Ootetrastichus. The special claims of the myrmarids rest on the 

 fact that in the climate of the Sandwich Islands these insects 

 complete their entire life-cycle in about three weeks, and 

 apparently breed nearly at the same rate throughout the year ; 

 while they are likewise mainly parthenogenetic, males being 

 only produced at long intervals. The testrastichid parasites, on 

 the contrary, are comparatively slow breeders, taking at least 

 twice as long as the myrmarids to complete their cycle. In 

 compensation for this slowness is the vastly greater number of 

 eggs produced, probably double that of the myrmarids, and, so 

 far as is known, the purely parthenogenetic reproduction. 



" If we judge the effectiveness of the two groups of parasites 

 merely on rate of increase (reckoning the life-cycles as twenty 

 and forty days respectively), and suppose that the myrmarid 

 produces twenty and the tetrastichid forty female young, at the 

 end of six months the latter will have produced 4,096 million 

 descendants, but the myrmarid in the same time will have 

 produced one million more." 



Other factors, however, come into play, so that the problem 

 is really more complex. The tetrastichid {Ootetrastichus) has, 

 for example, the advantage that each individual is bred at the 

 expense of the whole contents of an egg-chamber of the leaf- 



15 



