( 5©92 ) 
1. Whether the unequal RefradlionSj made without refpeft 
to any ioequah'ty of iocidence, be caufed by the different Re- 
frangibility of feveralRaysj or by the fplitting, breaking or 
diflipating the fame Ray into diverging parts > 
2. Whether there be more than two forts of Colours ? 
3. Whether Whitenefs be a mixtur^i of all Colours f 
, , The F/W? of thefe Siaerei you 
That tht Kny li not fplit, cr any ^ j„ A • 
.thtr^trtciM «*ay already dctermiod by 
an Experiment in my former Let- 
ter ; the defign of which was to fliew. That the length of the 
coloured Image proceeded not from any unevennefs in the 
Glafs,orany other contingent Irregularity in the Refradions, 
Am^ngft other Irregularities I know not, what is more obvi^ 
om to fufpe<9:^ than a fortuitous dilating and fpreading of 
Eight after fome fuch manner, as Des-Cartes hath defcribed in 
his Ethereal Refrafliions for explicating the T^y^ of a Cornet^ 
or as the Animadverfor now fuppofes to be effected by the 
«Sphttingand Rarifyingofhis iEthcrcal|pul(cs» And toprc^ 
vent the fufpieion of any fuch Irregularities, I told you, that 
I refradted the Light contrary ways with two Prifmes fuccef- 
fively, to deflroy thereby the Regular tSt&% of the firjl Prifme 
•by the fecmd^ and r© difcover the hregtdar efiefts by augment 
ing them with iterated refraAions. Now, amongft other Ir* 
regularities, if the firff^ Prifme had fpread and diffipated every 
ray into an indefinit number of diverging parts^ the fecond 
ihould in like manner have fpread aad diffipated every one of 
thofe parts into a further indefinite number, whereby the L 
mage would have been ftill more dilated, contrary to the e- 
vent* And this ought to have hapned, becaufe thofe Linear 
diverging parts depend not on one another for the manner of 
their Refta^ion, but are every one of them as truly andcom- 
pleatly Rays as 1 ha whole was before its Incidences as may ap- 
pear by iatercepting them feverallyi 
i The reafonablenefs of this proceeding will perhaps better 
appear by acquainting you with this further circumfliance* I 
fometimes placed the fecond Prifme in a pofition Tranfvcrfe to 
the firfi^on defign to try, if it would make the long Image be- 
come four-fquare by refraftions croffing thofe that had drawn 
the round Image int6 a longone» For, ifamongflothcr Ir- 
regularities the Refradiien of the/'j? Prifme, did by Splitting 
dilate 
