406 Mr. J. S. Grardiner on the Distribution 



do, use the argument that identical species in the above and 

 many other distant localities are cases of parallel or even 

 convergent evolution if they cannot raise ridges to span the 

 intervening gaps sufficient to permit of the migration over 

 of the species, so as to allow intercrossing to take place. We 

 ourselves — I have done so — use that argument ; but we ought 

 first to exhaust every possibility, in the absence of direct 

 observations of such evolution, to avoid getting into such a 

 strait. At most, by our faunas we get the possible range o£ 

 distribution of the larvae, but what we want in the first place 

 to know is their regular and average range. If we find 

 different species of the same genera or different varieties of 

 the same species of animals on the Ceylon and Maldive 

 slopes, we may infer that the larvae of these genera and 

 species seldom interchange, and we may deduce from the 

 known currents of the region the length of time for which 

 the larv£e can — or rather do — continue to live as such. 

 Although I consider that for the due progress of any science 

 we must have a certain amount of speculation, yet it seems to 

 me that this method is rather fatuous in the absence of any 

 direct evidence of the length of time to which any of the 

 larvse can live in the open ocean. 



Our second method is the direct study of the length of life 

 of larvEe as such. In considering it I divide the larvae for 

 convenience under three heads — the crustacean, the trocho- 

 sphere and its derivatives and allies, and the planula and its 

 allies, I deliberately omit certain larvae, partly because 

 they are unimportant, or I have not found any observations 

 to aid me, or they have not come under my notice in the 

 plankton with which I have had to deal. The most important 

 of those thus omitted is the ascidian tadpole, in respect of 

 which I have no direct observations, never having found it 

 more than twenty-four hours (measuring the current in time) 

 from land, though one is compelled to suppose from its 

 structure, so efficient for pelagic life, that it can be carried 

 for many days and considerable distances. In any case the 

 distribution of the Tunicata is so little known that it is of no 

 aid, but I have not heard of any occurrence that would give 

 it a life of more than five or six days. I mention it, howevei', 

 because it has saved me from one error, which I nearly 

 committed rather extensively, but which suggests a method 

 that may yet yield valuable results. Certain acinetid and 

 vorticellid Protozoa commonly settle on tunicate and other 

 larvae. They branch dichotomously, and, according to some 

 observations of my own at Naples in 1895, they divide and 

 hence branch dichotomously each night, the observations 



