31r. G E O I I ~E E. lhvrs. in rcply, mid. rcfcrring to tho rcmnrks of 1)r. Grossmun, hc did not conr;idcr thnt condensntiou in tho mniiia wns n ficrious mnttcr. IIe hnd bccn scrrcd with gna some niilcs mvny from the gnssorks, nnd it w i s prncticdly cIcIivcred to him :it the snnic cicgrcc of illuminnting powcr ns it left tlic works. thcrc not Iieingmorc thnn half ii cundlc cliffcrencc. I~csidcs, if condcusntion kept trrking place in the mains, this condcnsctl liquid slioiild nppcnr in tbo strcct syphons, wliere it is well kuown thnt nothing but foul ivntcr iisunlly nppenred. A deposit of Itydrociirhn from the fins could only tnkc plncc if such gns were siiturntcd with Iiydrocurboii vnpours. lic Iind iicvcr yet met with it finrnple of gns thoroitglily sntiinitccl, nntl lie cluitc iigrcecl with Ik. Gmssrnnn tlint ppq compiinics slioiild supply nll their consiimcrs with g : i ~ of the stnndnrd illiiniinnting powcr. I t was not suflicicnt to test nt tho works, which he thottglit Iind bccn siifficicnfly provctl. If P L ~ \rere tcsted outsitlc thc works in sonic public building froin tlrc orclinnry muins, lie felt quite siirc tlic coruplnints of bnd p i s would be rcdiiccd to n niininiriin. I n rcfcrcncc to I)r. 1)rcyfits' rcniiirks, lie wns of opinion tliiit the ~~l t g ~i c n i test of photometry wns fnr niorc rclialilc tlinri it iviis posuililc to rnnkc :my clicniicnl tcst. For iustiiiice, tlicrc wns it W n t diffcrcnec in the illuriiiuiitiiig power of tlie vnpoar of benrcric (CJ1,J and hcxyl ligdritlc (C,JIl,), wliilc thc clicmiciil dillcrencc, so fnr ns rcgiirdcd Iierccntngc of ltytlrogcti nntl enrbon, wns w r y sinnll. I h . Fninldnnd nnd otiicrs l i i i d nttcnipterl to viiluc . tlic illiiniinntiug power of con1-gnw liy II~C:IIIS of its cliciinc:il composition, lint it tvns his oliinion tlint nll site11 cni1c:ivonrs iiiiist cnd in fiiilnrc. I n rcply to Mr. Stcihoit?;c, the tests of tlic gns wcre iiinilc in tlic M:tldlestcr tcclirnciil lnlionitory at 10 o'clock in the ritoriiiiig nnd bctwccn 3 lint1 $1 in tliu aftcriioon, nnd IIC ilitl not attribute tfic diflcrciIcc 11ctn.ecn tlic tcsts ninrlc :it tho gnscvorks iind his own tests to itnytliiiig inorc tlinn tlic Iircsencc of nitrogcn nnd nir. I t iv:is inipossililc to ltnvc g:is of 11 high qiiiility tvlicn it couteiircd iis inucli ctirlioiiic wid, nitrogen, iind o q g c i i IIS tlic gns did in tlic iliiytiinc, iind this h r c upon thc question put by IJr. Ilnilcy. l*:vcry