90 
THE VOYAGE OF H.M.S. CHALLENGER. 
On the other hand, we have seen that each group has different affinities — those of the 
Theocosomata being with the Bulloidea, those of the Gymnosomata with the Aplysioidea. 
We must conclude therefore that the Pteropoda are polyphyletic in origin. 
We shall now endeavour to show in the case of each group what has been the line of 
descent. 
A. Origin of the Thecosomata. 
Hitherto those authors who have believed that the affinities of the Pteropoda are 
with the Tectibranchia (de Blainville and Boas) have contented themselves with indi- 
cating the proximity of the Thecosomata to the Bulloidea, but without going further and 
trying to ascertain whether the Thecosomata are phylogenetically derived from these 
latter, and in what way this descent may have taken place. 
It may be most confidently affirmed that the Thecosomata are descended from ancestors 
resembling the Bulloidea, and that the cause of the modifications which they have under- 
gone is to be found in the increase of natatory habits and the adaptation to pelagic life. 
If now we try to ascertain by what process the passage from one group to another 
has taken place, and by what successive modifications a Bulloid has become a Thecoso- 
matous Pteropod, we are met at first by an apparent difficulty, in the fact that the most 
primitive Thecosomata, the Limacinidse, are sinistrorsal, whilst all the existing Bulloidea 
are dextrorsal. But is this a real difficulty ? Is there in fact a great morphological 
difference between a dextrorsal and a sinistrorsal animal ? 
Of what importance is the direction of the spiral ? It is of scarcely any value, for 
we see among the species of a single genus ( Neptunea , Pyrula, Vertigo, &c.), or among 
the genera of a single family ( Lanistes and Ampullaria), forms coiled in opposite 
directions . 1 If this be the case with forms so nearly related, there is a fortiori no reason 
for astonishment that the same thing should happen in the case of the Bulloidea and 
Limacinidse. 
The examples just quoted show that it is very natural and simple that among the 
Bulloidea there should have arisen in course of time sinistrorsal forms, which, however, have 
preserved the dextrorsal asymmetry of their internal organisation ; that is to say, that in 
these animals the “ sinistrorsity ” has only affected the coiling of the visceral sac and the 
shell, and these sinistral forms would bear to some of the Bulloidea the same relation that 
Lanistes bears to Ampullaria. (Bouvier 2 has shown that Lanistes is not sinistral as 
regards its organisation, and that it differs from Ampullaria only by the contrary 
twisting of its visceral sac.) 
These forms, which are still unknown to us, are the extinct ancestry of the Lima- 
1 In the Pyramidellidae we have a case in which in the same specimen the first coils are sinistral and the subsequent 
ones dextral. 
2 Sur le syst&rne nerveux typique des Prosobranches dextres ou senestres, Comptes rendus, t. ciii. p. 1276. 
