134 METAMORPHOSES OE MAX 



for a certain period, but that it could only be attained 

 in proportion to the number of generations which had 

 preceded the birth of the insect. This explanation 

 did not hold its ground any more than those which 

 had preceded it, and took no position in science, for 

 it deserved none, seeing that it never touched on the 

 real difficulty — the fecundity of the virgin aphis. 



According to another author, the aphides invariably 

 produce ova in the same manner as other insects, but 

 the fecundation instead of affecting merely one gene- 

 ration, extends its influence to several successive ones ; 

 hence, a second impregnation is unnecessary till the 

 entire of this transmitted influence has been exhausted. 

 According to this hypothesis, the ova are developed 

 in a fecundated condition, and are hatched in the 

 mother' s uterus, just as they are among the ovo- 

 vivipara.* The phenomena observed in the aphides, 

 when seen from this point of view, resembled those 

 already known, and although somewhat exceptional, 

 still they came almost within the limits of the general 

 law. The very vagueness of this explanation was a 

 recommendation to some zoologists, and caused it to 

 be accepted generally ; and until of late years it re- 

 mained almost unquestioned. 



Shortly after Bonnet made his curious discoveries, 

 naturalists observed another series of phenomena far. 



* Animals are termed ovoviviparous, which produce ova like 

 true ovipara but which retain them in the uterus until the period 

 of hatching ; the young being then expelled without however 

 possessing any of those connections with the mother's organism 

 which are seen among mammalia. There are instances of ovo-vivi- 

 parity among reptiles (the viper), fishes (the blenny), mollusca, and 

 annulosa. 



