Metamorpliosis and Lite-history of Gnathia maxillaris. 470 



corresponcl to tlie mediocre size of least frequency in the adult males, 

 i. e. circa 4 mm. 



AVe bave so far ouly considered tlie variability of the males: 

 the fem ale s are very much rarer than the males, a fact which has 

 beeu noted by Sars in the Norweg-ian species. From a study of 

 2(50 adult ovig-erous females I bave determined that the Variation in 

 size extends from 1 to 7 mm : the curve of frequency distribution is 

 bimodal, the modes being at 2 and at 4 mm (thus in 260 females 

 13 measured 3 and 33 measured 4 mm). It is probable that these 

 two modes correspond in reality to the two modes in the males; 

 the bodies of male and female not being strictly comparable with 

 regard to length because the female when distended with eggs is 

 stretched transversely and loses in stature. The modal condition of 

 2 mm is of far commoner occurrence than that of 4 mm, just as the 

 small males preponderated in numbers over the large. It is therefore 

 of interest to find that the large females always produce a far greater 

 number of eggs '. Consequently to account for the persistently and greater 

 frequency of small adults in general we must either suppose that 

 there is a much greater selective death-rate among the offspring of 

 the large adults or else the numbers of the small adults are continually 

 being supplemented from the offspring of the large forms. A definite 

 answer to this question can only be obtained by rearing the larvae 

 from the youngest stage upwards, and besides that one would bave 

 to know at least both the parents of the brood; but at present my 

 efforts to get the adults to pair in captivity and the young- to infect 

 fish in acquarla bave been in vain. So we may proceed to inquire 

 from other evidences what it is that determines the size to which 

 auy Prauiza may attain. Is this a purel}' hereditary character or 

 does it depend upon the individuai experience of each larva? Certain 

 fticts point strongly to the conclusion that although iuheritance may 

 play a part in determining the size of the adult yet an equally 

 imi)ortant róle must be allowed to the direct action of nutritional 

 conditions upon the larva during its parasitic life. For firstly it 

 must be noted that although the larger females produce slightly 

 larger segmented larvae than the smaller females, yet this initial 

 difference in size is not sufficient to account for the immense varia- 

 bility of the adults, and so the diöerentiation in sizes must occur 



1 Females spanning 2 mm had a meau nuuiber of 50 eggs, those spanning 

 4 mm a mean number of 104 eggs. 



