52 ON THE ECHINODERMATA OF THE 



dicular to the plane of the plate, midway between the extremities. When the young 

 Starfish is examined under the microscope from above, it will be seen that in the lateral 

 sutural furrows, which open on the dorsal surface, there are, in addition to the marginal 

 fringe of compressed spinelets, an inner series of fine, pointed, cilia-like spinelets, at 

 least at the upper portion. 



In the early stages of this species, the lower or ventral series of lateral plates lies 

 much more upon the actinal or ventral surface of the Starfish than it generally does 

 in the fully grown form. In the examples above mentioned the row of single or 

 true lateral spinelets, borne by the ventro-lateral plates, stands almost at the angle 

 formed by the vertical side of the test and the actinal surface, the plates themselves 

 arching sharply under, onto the actinal surface. The spinelets also seem to occupy a 

 position relatively nearer to the centre of the plate than at a later stage. In large 

 specimens no portion of the lateral plates curve onto the ventral area, and the spinelets 

 are situated much nearer to the upper extremity of their respective plates. 



Variations. Dr. Liitken records examples from a station off Norway, which pro- 

 bably lies near the southern limit of the area of distribution, that differ strikingly from 

 the normal short-armed form by their longer and more pointed rays, whereby the 

 contour approaches that of Archaster Parelii; but in other respects no differences were 

 noticed. We have observed a similar variation in the radial proportions amongst a 

 series of specimens from Barents Sea, but not developed to such a marked degree as in 

 those examined by our learned contemporary. The relative proportions of the greater 

 to the lesser radius in two specimens are respectively 17 millims. to 9'75 millims. 

 in the one, and 17 millims. to 7 millims. in the other; two smaller examples 

 measured similarly 10 millims. to 6 millims., and 10 millims. to 5 millims. the 

 difference in character presented by the wide and gentle curve of the arm-angle 

 of the one in comparison to the more acute and angular outline of the other being 

 much more striking to the eye than the figures which indicate the actual proportions 

 would seem to imply. From the occurrence of both these forms together, as well as 

 the identity of their general structure, it is, perhaps, not improbable that we have here 

 nothing more than a sexual character. 



Upon the whole, this species would seem to be remarkably constant. After a 

 careful study, however, of specimens from Greenland, Novaya Zemlya, and North 

 America, we are inclined to believe that a certain amount of variation does occur 

 (probably of locational permanence) in the features of the ventro-lateral plates and of the 

 ambulacral spines, after the manner indicated whilst treating of the phases of growth ; 

 and although this would seem rather like a confusion of the stages characteristic of 

 growth with the features presented by circumstantial variation, the evidence has been 

 such as to lead to the inference that certain characters of early growth-phases are, 

 in some localities, retained until a much later period of growth, perhaps even 

 becoming a permanency through life a state of things which is perfectly explicable on 

 the not improbable assumption that the exigencies of arctic existence have acted in 



