AND BOILER SETTINGS' 343 s Khl mphtudo of the temperature surges at C is due to '» <««•< that this roupk. is located at a point where a pocket of hot gas occurs I lus would hold the temperature within rather r'T ""! * '''"n! r falrly Probllbl° that these SU1^S <** due, in par, to ,h«. mhirahon of cold air, as those couples were located fairly close lo the center line of the boiler. This location would briiiR them Mwmi 1 ho two drums at a point where cold air might drop em thorn from leakage. The air leaking in at this point would be heated by the* brickwork. Stratification would not be likely FKJ. 177. Dinfcrnm nh«iwhiK th<« flow of gases in two pass horizontally baffled boiler. Tempera- tiin* rnt'iutiirrtui'iitft miufo in thin wotting arc shown in curves, Figs. 178 and 179. This t*«»il«*r JK the itiiwi* niyn" und make an the single pass shown in Fig, 174. Presumably the ntcftrn was K<'"<'rnt«*r2 At about the nunio rato in both boilers. The location of cool gas jMnrkH.fi m indicated in tin* same mariner as in Figs. 170 and 179. at this locution tin the tubes have a tendency to mix the gases very completely UH they pans upward. The large* amplitude of the temperature surges at A is probably due* to the fact that this couple was at times immersed in the jet of aHrttruHng hot. gases and at other times in the eddy of com- parativ(»ly cool g'anes at this point. Another factor that is men- tionwi in the caption of the curve is the formation of balls of chilled guBCH among the tubes. These balls of cool gas will have a tendency to drop into the lower portion of the setting, but will be sustained by the up flowing current of hot gases. An analogy in found in the submerged bubbles of air caused by a stream of water falling into a body of water.