592 



SCIENCE 



[N. S. VOL. K No. 13M 



dongknut and the writer's impression is that 

 they were from two to four inches in diam- 

 eter and about a half inch in thickness. Each 

 had left behind it a track in the snow which 

 led from the foot of the overhanging portion 

 of the drift wall down its side into or nearly 

 to the road. A few curred, columnar pieces 

 of snow were also found which had fallen 

 from the top of the drift and had rolled down 

 the side without forming rings. 



It was evident that the rings and columnar 

 pieces had been formed from small tongues 

 of snow which had been built out over the 

 steep side of the drift at its top by the wind. 

 These tongues had separated from the snow 

 wall first at the top and had bowed them- 

 selves over until their free ends nearly or 

 quite touched the snow at their bases with 

 the result that they broke away and rolled 

 down the bank. 



As in the ease of the attainment of large 

 size by the "rolls" described by Karl M. 

 Dallenbach in your issue of October lY, so 

 in this instance the completion of the ring 

 form was a matter of balance during the 

 process of bending forward and rolling down 

 since a few fragments had broken away and 

 rolled on their sides without having attained 

 the ring form. 



While the wind seems in this case to have 

 operated in building out the tongues of snow 

 until they became too heavy to maintain their 

 equilibrium it was probably not involved in 

 the rolling process which seems to have been 

 due altogether to gravitational attraction. 

 W. Armstrong Price 



West Virginia Geological Survey, 

 morgantown, west virginia 



variation of fishes according to 

 latitude 



To THE Editor of Science : In the number 

 of Science for April 4, Professor Starr 

 Jordan discusses the generalization that in 

 certain families of fishes species living in cold 

 waters have a large number of vertebrse, while 

 the related ones of warm waters have a small 

 nimibgr. He interprets this as being the 

 result of a general phylogenetic process which 

 is favored either in warm or in cold water, 



depending upon whether the large or the 

 small number of vertebrse is considered as 

 more primitive. He has attempted to deter- 

 mine which is the more primitive by investi- 

 gating the ontogeny of the metameres in 

 Sebastodes, but has failed in this, because, as 

 seems to be generally the case, the number 

 of metameres characteristic of the adult is 

 attained at a very early stage. 



"We would refer to the fact that such varia- 

 tions in number of vertebrae with temperature 

 occur within the limits of a single species, 

 as Heincke'^ has shown for Clupea harengus. 

 Both sea-herring and coast-herring show a 

 decrease in (1) number of vertebrae, (2) 

 breadth of skull, (3) number of keeled scales, 

 and (4) length of body, as one goes from the 

 open ocean into the Baltic. We would suggest 

 that this shows the adaptation of the large 

 type with many vertebrse to water of great 

 density (very saline and cold) during the 

 critical and sensitive early stages of develop- 

 ment, and of the small type with few vertebrae 

 to water of low density (brackish and warm) ; 

 that is, that certain characteristics connected 

 with a large number of vertebrae make the 

 young of the large type develop successfully 

 in water of high density and that other char- 

 acteristics connected with a small number of 

 vertebrae make the young of the small type 

 develop successfully in water of low density. 

 The adults are comparatively hardy and able 

 to seek water of suitable density. In the 

 species Hippoglossoides platessoides, Collett^ 

 has shown that northern specimens have more 

 rays in the dorsal and anal fins than have 

 southern specimens. We have not been able 

 to find that the individuals of this species 

 on this side of the Atlantic have the numbers 

 of fin rays varying according to latitude. 

 Nor does the variation of fin rays correspond 

 with the temperature or density of the bottom 

 water in which the adults live. There are 

 however indications that it corresponds with 

 the density of the surface water in which the 



1 " The Natural History of the Herring, ' ' 17th 

 Ann. Rep. Fishery Board for Scotland, 1899, p. 282. 



2 ' ' Pishes, ' ' Norweg. North-Atlant. Expedn., 

 Zoology, Vol. in., 1880, p. 148. 



