128 



E. Jorgensen. 



speet to tlie dependeuce of the radiolaria ou temperature and sali- 

 nity. Unfortunately, there are no samples from deeper water. 



It may, nevertheless, be concluded with certainty that all the 

 species here mentioned of the divisions SpumeMaria and NasseUaria 

 are oceanic species. There may, however, be diiferences of opinion 

 as to whether they are northern or southern forms. Cleve states 

 (L. 40) that some species which belong here (from the Northern 

 Atlantic) must be considered to be derived from the Northern Pacific 

 by way of the Northern Arctic basin. He comes to this conclusion 

 partly because several of these species have been found in the 

 northern part of the Pacific, and partly because they do not occur 

 in samples taken farther south than 40 " n. lat. 



Gran (L. 7(). p. 149 — 154) considers these species to be of 

 southern origin. 



As far as I can gather, Clbve's statements are generally based 

 upon examinations only of surface samples. My impression is that 

 the majority of these radiolcma come from the northern part of 

 the Atlantic, where they appear to occur in the mixed Atlantic 

 waters. In the deeper waters of the Norwegian Ocean, thei-e are, 

 perhaps large numbers of these northern species. In a deep 

 water sample from the sea between the Faeroe and vShetland Isles, 

 which was taken by Mr. Clark and kindly placed at my dispo- 

 sition for examination, I found a comparatively large number of 

 those species which occur on the coasts of Norway, and some of 

 these species were numerous. 



It appears to be certain that ocean water of between 34 and 

 35 "/oo salinity contains many species of these radiolwia. Perhaps 

 here too they oceui' for the most part in the deeper layers. 



On the other hand, there appears also to be species (of those 

 mentioned in this paper belonging to Spiimclhiria and XasseUaria) 

 which belong to the salter, warmer waters of the Atlantic, but 

 their number is comparatively small. 



Key to the genera of NasseUaria. i) 



Shells with plainly latticed cephalis and thorax, or 

 more joints. 



More than 3 joints; rather narrow sliell 



with no Ol' inconspicuous spines 



3 joints (cephalis, thorax and abdomen). 



I'horax with 3 long protruding 



spines, each with a bi'oad base; 



no broad tophorn 



Thoi-ax with 3 to 4 slender 

 and short spines which may easily 

 be overlooked. 



2 diverging, 3-e(lge(l 

 tophorns; broad thorax: 

 short, expanded, l)iini- 



shaped abdomen 



1 conspicuous, 3-edged 

 tophorn ; thorax narroAver. 



not brim-shaped 



2 joints (only cephalis and thoi'ax). 



Cephalis rudinuMitary; largo 

 thorax LUhurach 



Stichocoru^- 



Dictyocerm. 



Chdhroriicliis. 



A)>>lrori/rlfi 



(Inly to 



le (let(;rminati 



Thorax little developed with 

 narrow spines which form <i or 



more „basal feet" Aennfhocorys. 



Thorax with piercing, 3-edgcd 

 spines or with 3 regular liasal 



feet Didyojiliimus. 



Thorax with piercing, slender 



spines without conspicuous edges. 



Thorax broad and low, 



campanulate ; cephalis \\ith 



numerous spines Helotholus. 



Thorax narrow and high. 



Cephalis very 



broad, not plainly 



constricted from 



the thorax Amplihnellssa. 



Cephalis nar- 

 rower than the 

 thorax, easy to 



distinguish Lithomelissa. 



Only cephalis, plainly latticed. 



Cephalis with a sagittal I'ing (bilocular) Cerafospyris. 

 No sagittal ring. 



5 long, 3-edged spines, one of 

 them forming a nearly central 

 columella with an inner verticil of 

 3 branches and protruding above 



as a tophorn Cladoscenhmi. 



The same, but the columella 

 wanting the inner verticil of bran- 

 ches Euscoiium. 



No broad, 3-edged spines. 



No columella Fcridiiini. 



(Incompletely latticed 

 shell with indistinct, margi- 

 nal columella Phormncanflia). 



More or less incomplete skeletons, not plainly latticed. 

 A strong sagittal ring with spines and 



some few meshes Dictyoc'urus. 



No sagittal ring. 



The ventral sagittal s]iine is 

 present. 



Rather well doveloped 

 network (at last with 

 numerous slender ai-i'lios 

 and spines at the outside) Phormacantha. 



No network, only strong, 

 bent spines with nnmei-- 

 ous scattered branches . . Campyhicdiitlm. 

 (5 strong, sti'aight. 3- 



edged spines Ensceniuni). 



The ventral sagittal spine is 

 wanting. 



A very incomplete ce- 

 phalis with very I'ew. 

 large meshes and nrai-ly 



central columella Protuncenium. 



No columelhi. 



