r I^lS[C>WLKEf0E 



[July 17, 1885. 



I i 



m. i i 





,. If li 



tufaceous suifiLC (f ^[^ A\ ilh 

 Its light thiou-h in itiii sphd ill surli"i stUt f 1 

 £ am in no desfrec iiit ust. 1 in tint tluor^ ^\hlch 

 aeems to me not so mttch to he outside kno-n ledge, as to 

 be utterly inconsistent with known facts 



Turning to another point — the difference between the 

 behaviour of a pencil of rays in pisMnir to a focus and 

 •^he course of the axis of a ] i il is - hum hat more 

 important than Mr Wilhiin-, m- 1 i instance, 



failing to- 'recognise this dill i Ah "W lUnms has 



been kd hh imagine that the f cui of i ptULil of rays 

 proctelin,' f r mi i point on the suns suiface to the 

 earths uuiosphii thence to ])ass towards the prolonga- 

 tion of the line joining the centres of the sun and earth, 

 must he SMe'f^htn w, this hn I am quite surt thit 

 nothing m the tieatment of opticil jn .bh nis when Mr 

 Williams was a lad woald h u t enccuii ui 1 tin, mi;,tikt 

 As a mere matter of fut the focus (, r nther the 

 primary and secondii\ foci) of su. h x [kuci], l tufwt in 

 V ithm m my mdli -«s of mil, ^ of th it point on the line in 

 question to -nhich the axis of the ptncil is deflected by 

 the aetiou of tht tilth. ,tn,w,,h,H ,t ,„„f,unln 

 bet II 



^ 1 1 lilt \h A\illnm.-. 



^I'J'^ 1 . . I lint r 1 Imii tht 



con\tx kns, tht. imil n t i i i 

 oUk^v Cm he r<_c ill iii\ t th f imul i 1 it ii ' t. 

 the coustructi.m of cptitd iiisliui. mt, ului.iu it is~nn 

 an exires, cjiidition thit tlic illiquid t mtideme 

 should be so small th it it mn U «.uli i u .1 tt^d oi 

 itadilj coritatd'' 



I cm tike two simila p.ism- ml U m m. of them 

 make thi axts of two jitnuK i lu- fi m i lummoii.s 

 point mtcr,oet eith othtr J, tli.t j lut ot mteisection 

 the focus f .r thobeiajs - On thi loutrai^, e>ch pencil 

 :e has its own pan of joci and its c 



. ik lull 

 rth 

 1 U 1 tt 



.lib I'^l 

 .1 . t 1 

 1 Ins 1 t 



1 the rays forming 



f their axes) This 

 ion of two parts of 

 ,' t t d lunar eclipse, 

 I V 

 I hnd myself per- 

 ■^uiqiose, misunder- 

 1 1 1 I have read and 



mit of 



iL lb : 



f th 





xsii,tent 



of 



'rfictorj for ex.ict inqairiea; 

 lut unl s. this law, which is found to be true for 

 1) ironutiie \ iriations of considerable extent, fails entirely 

 it hi^'h le^els, it is certam that the air ibove 3^ piiles, 

 \>here the barometer stands at a height of only 15 lachea, 

 cim send ?io light vhateier to the moon at the time of 

 mid-tot ihty^ ' 



Secondly, it is absolutely certain tliat air not sa cldud- 

 laden.or so dustladen, or thick enoutrh from whatever 

 cause, to render the sun's light f unt t ^ us iv han te is 

 setting, tnust transmit a considerible amount of ligSt to 

 the moon evtn m the case of rays^azisg the earth, — 

 while riys passm;,' a mile or twaifom the earth's surface 

 would be absorbed in much les»^egree, and rsvys passing 

 lies fiom the' earth's surf ice' Vo'ild ^ 



Thii 

 hei^'ht rf M^ 



rb.d I 



^^hln tl 



cl uK 



tk 



those 



obstr 



t hi 



athwart their 1 u 

 clouds in the sk\ 

 clouds are so sp u 

 scarcely any h^'h 

 the hori/on 

 But Ihi sti 



dii 



1 





to 



y^reihly 

 blnuely 



im^oi that 

 by 1 60th 



after t 



-, 1 1 1 ' 111 Of course he quite 



11 '1 f In I f the suns decrease of 



lu.ti t \N iKih 1 111 -.in_r the effective extent 

 .fth uiulii But tki 1 hm h Milkn diminution of 

 Instil istli. _' 111 tu 1 slul Ns'i. ipirnaehed, and the 

 t,'eniril unii. milt} fth tint iftcru uds need no such 

 cxplmiti.m ashe hisckM^id Let the refr letive effect 

 of our eirth's itmosphere do \\hat it miy in diminishing 

 the extent of the retrion m ibsolutc* shido^N, it cm never 

 do more thin throw a "ierj feeble light -i^ithm the geo- 

 metrical shadow Moreover the light thus thrown by 



