March 30, 1883.] 



SCIENCE. 



221 



used for steadying the body in walking. In man the 

 arms are used, because most movable; but in lower 

 animals the head Is most often used. The domestic 

 fowl moves the head back and forth alternately with 

 the movement of the legs ; the horse moves the head 

 up and down ; the cow moves the nose back and forth. 

 Are these movements ghosts of a former real walk- 

 ing with the head ? Joseph LeConte. 



Berkeley, Cal., March 7, 1883. 



Suggested improveinent in lighthouses. 



The articles in Science during March, on the use of 

 the electric light in lighthouses, recalls what I think 

 is a most useful improvement suggested, or at least 

 advocated, by an English yachtsman, — Mr. E. F. 

 McMullen, — in a littte book called the 'Voyage of 

 the Orion.' In the vicinity of a powerful lighthouse, 

 whether lighted by electricity or othersvise, the great 

 glare of tlie light completely blinds the eye of the 

 navigator. To remedy this, Mr. McMullen proposes 

 that a colored shade should be fixed so as to change 

 the color of, and diminish, the light within a fixed 

 radius of from one to three miles, according to cir- 

 cumstances. Tims, besides the protection given the 

 eye from too much glare withiu the radius of the 

 shade, the navigator would also be warned that he 

 was within a known distance of the light, — a con- 

 sideration which would often be of mucli value. 



I sincerely hope that our Lighthouse board may 

 make some experimental trials of this plan, as well 

 as turn its attention to the adojjtion of flashing 

 lights, instead of revolving lights with long periods 

 of darkness. Indeed, in our whole system of lights, 

 and also of buoys and other ' day-marks,' we are 

 behind the times. Edward Buegess. 



Fluidal cavities in quartz-grains of sandstones. 



It is interesting to note that the minute cavities 

 containing a liquid and moving bubble, so common 

 in the quartz of granite rocks, are also to be found in 

 sandstones. This is especially the case with a hard, 

 compact Potsdam sandstone quarried at Fort Ann, 

 Wasliington county, N.Y. The cavities, though very 

 minute, are abundant, and the included bubble very 

 sensitive, being in a constant state of rapid move- 

 ment. G. P. Merrill. 



U. S. national i 



The copper-bearing rocks of Lake Superior. 



In Science, No. 5, Professor Irving takes issue 

 with my statement that there is no evidence what- 

 ever of the Lake Superior copper rocks holding any 

 other place in the geological series than that which 

 includes Potsdam and primordial Silurian or lower 

 Cambrian. 



In making the statemeiit, I referred only to those 

 parts of the north shore, extending from Sault St. 

 Mary to Thunder Bay, which I have myself exam- 

 ined. 



I could not presume to discuss, much less to dis- 

 pute, the evidence which Professor Irving adduces, 

 in disproof of my statement, from the St. Croi.K re- 

 gion and the south shore, neither of which I have 

 ever seen; but I maybe permitted to s.ay, that the 

 unconformities mentioned by Professor Irving, and 

 which I have no doubt are real, do not, in my opinion, 

 in the least invqjidate my statement. Unconformi- 

 ties, even if locally very great, are not necessarily 

 any indication of a great time-gap. And it seems 

 to me that too much importance has been attached to 

 these by Professor Irving, and far too little to the 



immense difference in the physical condition of the 

 groups he now correlates; viz., the original Huronian 

 of the north shore of Lake Huron, and Hunt's Anim- 

 ikie group, lower Cambrian of Thunder Bay, or, to 

 come closer, the cleaved roofing-slates of Thompson 

 in Minnesota, and the liorizontal micaceous argillites, 

 black dolomites, and cherty rocks, of Pie Island, 

 McKay's Mountain, Thunder Cape, etc. 



In Canada, at least, these two formations are abso- 

 lutely and undoubtedly distinct, physically, minera- 

 logically, and geologically; while the latter, as seen 

 around Thunder Bay, is followed in almost conform- 

 able sequence by the red and white quartzose sand- 

 stones, conglomerates, amygdaloids, etc., of the so- 

 called upper copper-bearing or Keweenian series of 

 Hunt. These 1 have examined from Thunder Bay, 

 around the north shore to Gros Cap, where they rest 

 directly on the Laurentian gneiss, the Animikie 

 group and the underlying Huronian being wanting. 

 A short distance to the east, however, the latter ap- 

 pears in full force, hut overlaid jieither by Animikie . 

 nor by Keweenian (i.e., lower Cambrian), but by the 

 Sault St. Mary sandstones, which, in view of their 

 relation to the Black River limestone above them, and 

 to the Keweenian in Gonlais and Bachewarmg Bays, 

 are much more probably representative of the hori- 

 zon of the St. Peters sandstone, or Chazy and cal- 

 ciferous, than of the St. Croix Potsdam. The re- 

 spective limits of the two sandstones on the south 

 shore seem uncertain. 



The arrangement above indicated brings the whole 

 succession of the Lake Superior, Cambrian, and 

 Cambro-Silurian formations into perfect accord with 

 that of the same formations in the Appalachian re- 

 gion, where, as I have elsewhere stated, indications 

 of local contemporaneous volcanic action are not 

 wanting at about the same horizon — lower Cambrian 

 and upper Huronian — as that at which they occur 

 in the Lake Superior region; the chief difference 

 being, that the formations in the former region are 

 folded and metamorphosed almost past recognition, 

 and in the latter not more so than are many simi- 

 lar rocks of cretaceous and tertiary age. 



I think, if Professor Irving could visit Michipi- 

 coton Island, he would be able to recognize plenty 

 of volcanic detrital matter or tuffs among the copper- 

 bearing rocks. The vast areas over which I have 

 examined the ejectanienta of the extinct tertiary 

 volcanoes of Australia enables me very readily to 

 recognize such rocks when seen ; but their occurrence 

 at Michipicoton, and elsewhere on the north shore, is 

 no proof that they also occur to the south, and there- 

 fore I fail to see why Professor Irving should dissent 

 from my statement on this point. 



Alfred R. C. Selwyn, 

 Director Geol. and nat. hist. sure, of Canada. 

 Ottawa, March 14, 1883. 



Sncw-drifts. 



Having often noticed the drifting of snow in paral- 

 lel lines over the ice on our lakes, this explanation 

 has suggested itself. Very often, when the wind 

 drives the snow against any object, as a tree or fence- 

 post, tlie snow will be hollowed out on the side to- 

 ward the wind, and heaped up on the other side. 

 This is explained by the fact that the tree acts as a 

 reflecting surface, creating a counter-current of air, 

 and preventing the accumulation of snow on the side 

 toward the wind. 



Might not the parallel ridges of snow on ice be ex- 

 plained in the same way ? The first deposit of snow 

 is caused by the flakes catching on some inequality 

 or damp spot on tlie ice. This deposit acts as a re- 



