52± 



SCIENCE. 



[Vol. IY., No. 97. 



sion through the earth's atmosphere, he com- 

 puted that twent}'-seven thousand million million 

 candles must be spread over the moon's earth- 

 ward hemisphere, painted black, to send us as 

 much light as we receive from her. Probably 

 forty thousand million million candles would 

 be required to allow for absorption. Sir Wil- 

 liam carried his computations a little farther, 

 and figured, that, if the face of the moon which 

 we see were painted black, and covered with 

 candles standing packed in square order, touch- 

 ing one another, all burning norm all}', the light 

 received at the earth would be about the same 

 in quantity (as estimated by our eyes) as it 

 really is. 



How does moonlight compare with sunlight? 

 On the 8th of December, 1882, Sir William 

 Thomson in Glasgow measured the brilliancy 

 of the sunlight at one p.m., and computed that 

 it was about fifty-three thousand times greater 

 than that of a candle-flame. This, he says, is 

 more than three times the value found by 

 Arago for the intensity of the sun's light. 

 ' So much for a Glasgow December sun ! ' 

 Hence he derived the conclusions that the Glas- 

 gow sunlight was seventy-one thousand times 

 the York moonlight, and that " we cannot be 

 very far wrong in estimating the light of full 

 moon as about a seventy-thousandth of the 

 sunlight airywhere on the earth." Those who 

 are curious to know more of this inquir}' will 

 find the note to which we call attention in the 

 proceedings of the Glasgow philosophical so- 

 ciety for 1882-83. 



LETTERS TO THE EDITOR. 



*V Correspondents are requested to be as brief as possible. The 

 writer's name is in all cases required as proof of good faith. 



The oldest living type of vertebrates. 



I was gratified to have my own conclusions as to 

 the systematic relations of the galeoid Selachians 

 verified by so competent an original investigator as 

 Mr. Garman. The differences between us now are 

 fictitious rather than real; or better, perhaps, they 

 are chiefly differen<jes of expression. 



As to the characters of the Opistharthri, it must be 

 remembered that I assigned them long before Chlamy- 

 doselachus was known; and then the statement that 

 among living sharks they ' alone exhibit ' the ' peculi- 

 arities ' specified, was literally true. 



' The palato-quadrate, not articulated with the 

 skull,' is a true character of the typical sharks and 

 Khinae. Of course the apparatus, being the suspen- 

 sorium of the lower jaw, must have some connection 

 with the cranium; but it is indirect, and not direct. 

 The name ' Anarthri ' is therefore quite appropriate, 

 contrasting well with ' Opistharthri ' and ' Proarthri.' 

 The newly proposed term, ' Mesarthri,' is, however, 

 unobjectionable, although I should still, independ- 

 ent of priority, prefer Anarthri. No one who took 

 an intelligent interest in the subjects in question 

 would be misled by the name ' Anarthri,' or the diag- 

 noses of the Anarthri and Khinae. 



I must dissent from the opinion that the Clado- 

 dontidae are related to the Chlamydoselachidae rather 

 than to the Hybodontidae. To traverse the question 

 would, however, infringe too much on your space. 



Mr. Garman, in his substitute for my provisional 

 diagnosis of the Selachophichthyoidei, ' vertebral 

 condition unknown,' has added to .our knowledge of 

 the group by verifying my suggestion (Science, April 

 11, 1884) that the " anatomy will probably reveal a 

 structure most like that of the Opistharthri." 



I am pleased to find that the views of Mr. Garman 

 as to the remoteness of the Xenacanthini or Ich- 

 thyolomi from the true selachians agree with those 

 expressed by myself. The Xenacanthini, in fact, 

 appear to me to be true fishes rather than selachians, 

 although not teleosts, as has lately been urged. 



Theo. Gill. 



Hornblende andesite from the new Bogosloff 

 volcano. 



A short time since, there were received at the Na- 

 tional museum, from Lieut. George M. Stoney of the 

 Ounalaska, several fragments of rock from the new 

 volcano on Bogosloff Island in Bering Sea. On ac- 

 count of the interest just now attached to this locality, 

 it is thought a brief notice of these may not be out 

 of place here. 



The rocks are hornblende andesites. Two varieties 

 were received, — one very light gray and slightly pur- 

 plish in color, fine-grained, friable, and somewhat 

 porous; the other dark gray in color, and much more 

 firm and compact in texture; both varieties contain- 

 ing macroscopic hornblende and plagioclase, and, un- 

 der the microscope, seen to be nearly identical, each 

 consisting of a gray groundmass in which are em- 

 bedded deep reddish-brown, strongly dichroic horn- 

 blendes, light green augites, and numerous crystals 

 of a plagioclase felspar. Sanidin is also present, a 

 very little apatite, and the usual sprinkling of iron 

 oxides, which seem to be largely magnetite. The 

 groundmass consists of a microfelsitic base, carrying 

 colorless microlites, grains of opacite, and minute 

 yellowish and greenish particles which are probably 

 hornblende and augite. The light-colored variety 

 contains small patches of a nearly colorless glass, 

 while the dark variety seems felsitic throughout. A 

 more detailed description of these rocks will be given 

 later. Geo. P. Merrill. 



National museum, Washington, 

 Dec. 1. 



Edison's three-wire system of distribution. 



Referring to the article with the above heading 

 in ISTo. 94 of Science (Nov. 21), it is not difficult to 

 show that the conclusions reached are not in har- 

 mony with the fundamental proposition governing 

 the size of electric conductors. This proposition is, 

 that ''the additional running-expense due to the re- 

 sistance of the conductor shall equal the interest on 



