204 EXPERIMENTS MADE IN VIRGINIA. [MEMOIR XIII. 



(<?.) Crystallization is not necessarily connected with 

 these results; the vapor of mercury in a torricellian 

 void is condensed towards the light; so also the dew 

 which settles on the inside of a jar containing water 

 is always on the side nearest the window. The rays of 

 the sun have also the power of decomposing chloride of 

 gold : the metalline spangles are deposited on the side 

 of the glass nearest to the light. 



Artificial light gives none of these results. 



(<?.) Having removed the piece of tinfoil used in ex- 

 periment (V), it was placed on a stand in front of the re- 

 ceiver; it hindered crystallization taking place on the 

 parts on which its shadow was cast, and also for a cer- 

 tain space in the vicinity. 



(jf.) A piece of tinfoil was placed before a jar that had 

 been already coated with crystals. It removed all those 

 crystals that were within its shadow. 



(</.) Instead of using a piece of tinfoil as in experiment 

 (<?), the receiver was made hot, and a piece of resin 

 rubbed upon it, so as to leave a transparent circle of 

 that substance on it. On exposure to the light it was 

 found that the resin could not protect the glass. 



(^.) Along the inside of a vessel about to be exposed 

 to the sun a glass rod was rubbed. Rows of crystals 

 were deposited on the lines described by the end of the 

 rod. But the vessel must be very dry for this experi- 

 ment to succeed. This curious fact was first 

 observed in the case of an exhausted vessel 

 having a small siphon gauge shut up in it, 

 the extremity of which rested against the 

 glass. By accident the gauge was moved 

 half round the glass, and shortly afterwards a 

 line of crystals was observed coinciding with 

 the line of motion. The appearance was such as is rep- 

 resented in Fig. 33. 



