Jax. 1, 188G. 



♦ KNOWLEDGE ♦ 



8c 



hardly yet be tnken as effectively defining tlie limit of 

 demarcation between 6 mag. and"? mao-. 



CoUuiouoo^J, July 28, 1870. 



I thank you foi- calling my attention to that section 

 m my " Outlines." * Undoubtedly there is a discordance 

 of statements which requires correction. But the appeal 

 there is rather to the statistical result of actual enumera- 

 tion ; and assuredly on a cursory view of the heavens on 

 a clear night, stars down to the 7th and 8th mag- 

 nitude do atfeet the eye, though we cannot n.v them ly 

 reason of that strange law which curtails a.star directly 

 looked at of a very large aliquot part of its photometric 

 effectiveness. 



I am very glad you are taking an independent line, and 

 utiiismg the immen.so additions which have been made 

 to rranogrjphy in the way of numerical accumulation 

 and improved knowledge of proper motions, ko. We 

 may — indeed must— form theories as we go along fwhat 

 else do we observe for.' R. P.] and 1 hey serre as 

 guides for inquire-, or suggestions of things to inquire- 

 but as yet we must hold them r.ither looselv, and for 

 many years to come keep looking out for side-l'ights. 



[I had submitted to Sir John Herschel nfj-ldea of a 

 method of presenting his gauges of the southern heavens 

 (and also, if possible, his fjther'.s) on an equal surface 

 projection. I may remark that the idea was never 

 carried out by myself, becuise Mr. Sidney Waters kindly 

 undertook to carry it out : and placed his results, pnb- 

 h-shed in ihe Monthly Xotices of the Astronomical 

 Society, at my dispo.sal. His valuable chart will be 

 found in my " Universe of Stars." What prevented me ! 

 from attending to that work was that I had entered on 

 another work of a far more laborious nature, and cal- 

 culated to throw much clearer light on the subject of 

 stellar distribution in space. I refer to my equal-surface 

 chart of the stars in Argelander's .series of forty large 

 charts. This chart, which I was not able even to begin 

 (so few were my leisure hours) during Sir John Her- 

 schel's life, occupied me about 400 hours, which I could 

 only with great difficulty make free out of my working 

 time. I give a small copy of this chart as Fig. 6, an 

 electro of which was kindly given me bv if. Flammarion, 

 in company with the two equal - surface charts of 

 lucid stars, referred to by Sir John Herschel later on, as 

 Figs. 7 and 8. Electros of these also were presented to 

 me by M. Flammarion,— in consideration for the use of 

 these illustrations— perfectly free to him however in any 

 case— in his Ihvue MensudU d,^ V Asfronomie Foi.uUire.-- 



CuUingicood, July 28, 1870. 



I have been thinking over pl-.ns for laying down the 

 star-gauges, and I can light on none that seems to promise 

 better than that which you suggest. The chief obstacle 

 to carrying it out will, I suppose, be found in the fact 

 that often all the gauges do not cover the whole surface, 

 and that there are necessarily considerable areas 

 ungauged. 



July 31. I have kejit this unfinished in hopes of being 

 able to lay my hands on the ma.ss of graphical prqjec- I 

 tions. &c., which formed the groundwork of what is said j 

 about the distribution of stars in galactic parallels in my 

 Cape observations, thinking they might save you .some 

 trouble. I know they exist, but after a great search and 

 turning over the cojitents of many portfolios and boxes, 



' I had noticed that Sir John Herschels recognition of a band of 

 bright stars running nearly on the line of the Mi'lkv Way was not in 

 accordance with his remark that stars of the higher orders are not 

 more richly =trewn in the Milkv War than elsewhere -R r 



I have been unable to lay hands on them — so I will not 

 longer delay replying to your letter— and doubt not that 

 you will quite as well accomplish your object from the 

 registered statement of gauges — both those of my father 

 f.nd my own. 



CulUngwood, Feb. 7, 1871. 

 The two star-charts you have been so good as to send 

 me (Figs. 7 and 8), are very interesting. The contrast 

 between the star density under the arch of the Via Lactea 

 to the right of the Nubecula in the southern chart as com- 

 pared with that in the diametrically opposite place in the 

 northern is striking. By the bye" I miss the Xttbectila 

 Minor in the former. [See map "on p. 89."' 



EVOLUTION OF LANGUAGE. 

 Br Ar.A S. B.allis. 

 ir.— SOUND CHANGES. 

 REFERRED (New Series, Xo. T.) to the 

 natural selection of .sounds according to- indi- 

 vidual and racial peculiarities, whi^h is one 

 rf the chief causes of change in language. 

 The extent of this selection may be gauged bv . 

 the fact thr.t, although, according to Prince 

 Lucirn Bonaparte, no less than 310 conso- 

 aiiids can b? articulrttd, even the most cultivated 

 languages make use of but r. small fraction of this number, 

 while the less cultivated .'how a quite remarkablj 

 poverty. Thus, Hindustani has 48 consonants, Sr.nscrit 

 37, or, if we count the Vedic I and Ih, 39 ; Turkish, 32 : 

 Ptrsi-n, 31 : Arabic, 28 ; Kafir (Zulu), 20 and severr.l 

 clicks; Hebrew, 23; English, 20; Greek, 17, of which 

 three are compound; Latin, 17, of which one is corn- 

 round; Mongoli:in, 17 or 18; Finnish, 11; Polynesian, 

 10 native consonants, of which many dialects have les.-. 

 Some Australian langunges have 8, with three variation.*. 

 The poorest of the Melanesi^.n group has 12, and other.s 

 13, 14, or more consonants. Looking at the chief 

 European languages in respect to vowels rs well as con- 

 sonants, we find that the simple sounds in English are 40, 

 French 3-5, German 28, Italian 28, Spanish 26, and 

 Portuguese 29. The number cf sounds is greatest in 

 languages whose speakers mix largely with other nation-, 

 and almost unconsciously ler.rn from them. The Hottentot 

 language, however, is said to have no less than twentj- 

 simple vowels and about " twelve diphthongs, but its 

 consonants are deficient, and largely consist of gutturals. 

 They are supplemented by four clicks— dental, palatal, 

 cerebral, and lateral — which are apparently relics cf 

 ordinal sounds differing little from those of the lower 

 animals. The poverty of such languages as Latin and 

 Greek may be attributed to the dominant character of 

 the minds of their speakers, which forbade their 

 adopting sotmds from the races they conquered ; this 

 exclusiveness hardly exists at the present day. But, 

 new sounds are rarely adopted into a language unless the 

 mental and physical constitution of the race is hardy ; 

 and even when they are adopted, the sounds take on a 

 different character in the mouths of the new speaker.-. 

 The h in French is a German peculiarity, and the gni r. 

 Romi.n attempt to render the German w. Thus, hameau 

 is our holne, gidchet our wicket. On the other hand, th^ 

 sound of «, as in j)ure, arose from an attempt to pronounce 

 the French u in pure. Ch and j, in English, are cf Romaii 

 or Xorman origin, but have b.^en adipted by analogy to 

 many words of Saxon pedigree, as child, Anglo-Saxon cild. 

 The Otyi-herero has neither ?,/. nor ii, r. nor c. Its 

 pronunciation is lisping, owing to the fact thr.t the 



