,35,. BIOLOGICAL THEORIES. 505 



without wings, and for the same reason. Not a single ancestor 

 of either of them was, so far as we know, devoid of either wings 

 or eyes. Winglessness and bHndness are constant characters among 

 soldier termites, not only in one species, but in several ; not 

 only on one continent, but on several ; and there is reason, therefore, 

 to believe that winglessness and blindness were already constant 

 characters, even before the existing species were differentiated 

 from one another — were even constant characters before any existing 

 species was evolved from the common ancestral type from which 

 they have descended. 



In the lungs of a common frog are frequently found several 

 parasitic worms belonging to the species Rhabdonema nigro-venosa. 

 Each individual Rhabdonema is male when young, and later in life 

 becomes female— in other words, it is a protandrous hermaphrodite. 

 The old ones (that is the females) produce ova, which, after being 

 fertilised by the spermatozoa of the younger ones (that is of the 

 males), are carried from the lung to the mouth of the frog, and are 

 then swallowed and pass with the food through its intestine, develop- 

 ing, as they go, into Rhabdites. These are ejected with the excre- 

 ment of the frog, and develop, in moist earth, into sexually-mature 

 worms [Rhabdites), of which some are males and others females. They 

 are never hermaphrodite, and they are, in other respects, unlike their 

 parents. The few ova produced by a female are fertilised by the 

 spermatozoa of a male, and, after feeding for a time upon the tissues 

 of the mother's body, escape, and find their way into the lungs of 

 frogs, there developing into Rhabdonemata, like their grandparents and 

 unlike their parents, and passing through a condition in which all are 

 male to a later condition in which they are female. 



I fail utterly to see how this resemblance of individuals to their 

 grandparents; and unlikeness to their parents, this regular alternation 

 of generations of hermaphrodite Rhabdonemata with generations of 

 male and female Rhabdites, is to be " explained " by the continuity of 

 germ-plasm, and I see no necessity for assuming the transmission 

 of two sets of pangenetic gemmules which take it in turns to deter- 

 mine the structure of the successive generations. It is very easy to 

 believe that the influences at work producing ova and spermatozoa 

 in one pair of Rhabdites are very like those doing similar work in 

 another pair, and seeing that this likeness in itself will explain the 

 likeness of the Rhabdonemata which develop from these ova and 

 spermatozoa in one case, to the Rhabdonemata produced in the other 

 case, there is no need for any assumption beyond that. So long as 

 causes are alike the effects are alike, and the assumption of the 

 existence of pangenetic gemmules, or of an unproven continuity of a 

 special germ-plasm appears to be uncalled for. 



I am almost afraid to refer to the fact that external influences 

 have some share in determining the form, not only of the developing 

 individual, but also of its offspring, because, whenever I have done 



