46 



B. Doubtful species. Misaki Harbour. Siboga area. Maldive Archipelago. 



hispida -I- . -- 

 Jiaccida 



Gardineri . . -f- 



septata . . -\- 



lyra + 



foliacea + 



C. ON THE GEOGRAPHICAL DISTRIBUTION OF CHAETOGNATHA. 



The study of the geographical distribution of terrestrial organisms and of their habitats, 

 has been pursued for many years with considerable success. We know the mammals of (say) 

 India, Peru, and Austraha; we know that this is nocturnal, that arboreal, a third of burrowing 

 habits, and so forth. But how far would such knowledge have been attainable, had the sole 

 record of poorly-described species consisted of such vague statements as 'N.E. Asia", 'Africa', 

 with no indication of the exact localit)-, of the nature of the country, nor of the conditions 

 under which the capture was made? Yet the data provided for the study of the distribution of 

 oceanic organisms have been for the most part of this vague kind lill within the last few years, 

 and even some of the more recent expeditions have left a good deal to be desired. 



Before the distribution of an oceanic species can be satisfactorily grasped, two things at 

 least are necessary: a fair probability (certainty is rarely obtainable) that the captures recorded 

 of a species really dealt with that species and with no other; and an exact record of the 

 geographical positions. The difficulty of the first of these is proverbial, and is not less in 

 Chaetognatha than in other groups; the second is often vague or difficult to tind. 



In the case of planktonic oceanic species, these two are not enough for a proper study 

 of distribution, for the depth of capture is also urgently needed. Recent studies tend to suggest 

 that a great factor in oceanic distribution (possibly the greatest) is temperature, that an isotherm 

 may be as great a barrier at sea as a mountain range or wide river is on land, and that the 

 'distributional area' of a planktonic species is often not an 'area' at all, but a solid tigiire in 

 three-dimensional space bounded by isotherms and isothermobaths. At present we are unable 

 to draw such a figure for a single oceanic species, not bccause such species are necessarily of 

 world-wide distribution (for they are obviously not all so), but because our records are so 

 scanty, notably as regards vertical distribution '). 



Those species, which remain apparently valid after revision, have been clearly marked in 

 the Synonymie List above, which constitutes an attempt at the first of the |)reliminary steps 

 to a study of distribution. The .second step, a compilation of the recorded occurrences of these 

 .species, exacted so much time and work that the resulting lists have been printed in full below, 

 in the hope that they may ease the labours of future students of Chaetognatha. They attempt 



l) .\nothcr puint of great impuitance in the matter is that counsel has been daikencd by marine zoologists having allowcd 

 themselves in many cases to accept the 'regions' of terrestrial geography, and to apply thcm to the occan; but this point is too large 

 for discussion in this place. 



