556 



combined hands, and tripled, 



NATURE 



\_Oct 13, 1887 



WW 



, to imitate the sign for 



five times double five — the hand added to the outspread feet. 

 The instinct in the primitive mind to represent concretely in the 

 symbol all that was contained in the idea would sooner or later 

 have to give way to the desire after facility and clearness. All 

 that was found unnecessary to the distinctness of the figure 

 would be done away with, strokes would be dropped or 

 shortened, and complex forms would be made simple. In pro- 



first 



mg m 



cess of time, therefore, ^A^' became \/; ^|^\J^ L 



\ X / and then jQ ; and ▼ ▼ , after coalesc 



VC/, was simplified into \ I/. Before this last change took 

 place, however, it would appear that a double-fifty form was 



employed, v yy , unifying in </C> . and rounding into (/^ 



the ^Etruscan hundred. It was probably because this double- 

 fifty form was used that no need was felt for a name or sign to 

 represent five times fifty, and that a new form only arose when 

 five times double fifty was reached. By this time the simplified 



fifty, \|/> had become contracted into I , and it was plain 



that a simpler figure could be got by representing double-five 



times fifty 1^ than by any way of symbolizing five times 



double fifty. This figure, 1^, would naturally change into 



li^, then be altered into I ^, and become joined into one in 



1^. By the simple process of doubling this symbol, the next 



higher form, -^|\. or v'lv . was attained. This figure does 

 not stand for a thousand, but for five hundred and five hundred. 

 If ^1^ had been, as is generally supposed, a unity mean- 

 ing 1000, the doubled length of the stroke would have been 

 unintelligible, and the multiple forms would have been 



ClOO' C 1300 ' ^^^ so on, and not qqIqq , 



With the double five hundred, ^|\ , we have probably 



reached the limit of what may be called the primitive numerical 

 notation — the notation of a people who spoke of, counted by, 

 and had symbols for ones, fives, double fives, fifties, double 

 fifties, double-five times fifty, and twice double-five times fifty. 

 And it was in the main with this cumbrous system of enumera- 

 tion that the Romans were content during their whole exist- 

 ence as a people, hardly making any advance beyond substituting 



a hundred form, C, for the double fifty, and a thousand form, M, 

 for the double five hundred. 



The "Sky-Coloured Clouds." 



The last distinct display of these that I have seen this season 

 was during the night of July 29-30, and there was a very slight 

 display, if really one, on August 9. They seem only to be 

 visible for a short period before and after the summer solstice. 

 I have looked up all the recorded dates to be found in Nature 

 and elsewhere when these clouds have been seen, either by 

 myself or others, and the following are the first and last noted 

 dates each year : 1885, June 8, July 7 ; 1886, May 28, August 12 ; 

 1887, June 18, July 30 — but a suspected display of the pheno- 

 menon was observed four days earlier (June 14), and an equally 

 doubtful one (mentioned before) as late as August 9. 



It does not appear to me possible to attribute their luminosity 

 to anything but direct solar illumination. Mr. Rowan in his letter 

 (Nature, vol. xxxvi. p. 245) thinks otherwise, though in an 

 earlier letter (vol. xxxiv. p. 192) he seems to express a similar 

 opinion to mine. With regard to Prof. C. Piazzi Smyth's letter 

 on the subject (vol. xxxiv. p. 311) I have had a conversation 

 with him since that letter was written, and he said he did not 

 intend to convey the idea (which it had done to Mr. Rowan as 

 well as to myself) that the auroral line appertained to the spec- 

 trum of the clouds ; and my observations with the spectroscope 

 quite agree with his belief that they are not self-luminous, for I 

 have failed to see any bright line. Prof. Smyth says the 

 spectrum of these clouds is purely that of twilight. 



The earliest observations on these clouds would appear to be 

 Mr. Rowan's. Some authoritative assertion of when they were 

 first seen would be very interesting in considering their cause. 

 1 do not know that any one else perceived them before 1885, 

 but Mr. Rowan, writing in 1886, says they had occupied his 

 attention during the previous ^wo or three years- — if two years, the 

 date, 1884, would be after the Krakatab eruption, and would 

 add to the plausibility of their suggested connexion therewith ; 

 but if three years, it would disprove such connexion. 



Sunderland, September 30. T. W. Backhouse. 



A Light Fog. 



At Blowing Rock, Watauga Co., N.C., a part of the main 

 chain of the Blue Ridge Mountains, at this point possessing an 

 elevation of about 4000 feet above the sea, on the night of the 

 6th inst., while the writer was crossing a causeway through a 

 mill-pond a light .fog, obscuring objects at a hundred yard: ' 

 covered the water. The moon, a little past the full, produoei 

 upon the bank of fog a very distinct bow. The bow was luminoui 

 white, without any trace of colour, about 2° in breadth. Th 

 ends apparently rested on the water, the entire arc being re 

 fleeted in the water. The segment of fog within the bow was 

 faintly lit up, the lighting up being distinctly seen by contrast 

 with the fog outside the bow. At the same time on turning and 

 looking at the moon it was seen surrounded by a corona about 

 2° in diameter (four times the moon's diameter), the colours— 

 fairly bright — being in order, going from the moon's limb, 



a { 



I 



