EXPEKIMBNTS AVITII IONIZED AIK. 



25 



out in Cba[)ter I. (§4) should be I'ecalled, as shown in the foi'egoiug figure 5. The 

 pressure temperature locus, whicli determines the margin of the opaque zone, is here 

 fully given, and it is probal)le that the colors must lie along parallel loci. In my 

 earlier memoir I found the equation d = AW CPi(p-B)m^ whei'e A — 9, (7= .013, 

 y» = .35, ^=43, expressed the opaque margin [)retty well, and this has been 

 drawn in the diagram. B is the height of the asym[)tote and {'p--B) to be taken 

 numerically. 



In the following table 2, 1 have added some special results for the position of 

 the color curves in question, obtained from direct experiments made at diflferent 

 tempei'atures. These agree with the view stated pretty well, but are not as decisive 

 as one would wish, pt)ssibly because the actual tempeiviture at which the reaction 

 takes place is not given by the thermometer. The peculiar feature is the distribu- 

 tion of coloi's at 27°-28°, as seen in the chart, figure 6. All attempts to clear up 

 these discrepancies have thus far failed. 



TABLE 2.— VARIATION OF COLOR WITH STEAM PRESSURE IN THE ABSENCE OF 



ARTIFICIAL DUST. 



Color. 



Yellow 



Brown-Yellow , 



Opaque 



Violet 



Blue 



Blue-Green. . . 

 Faint Yellow. . 



Purple 



Violet 



White 



Pressure. 



70 cm. 



40-50 



40 

 3.3-35 



3° 

 28-25 

 15-20 



10 



Temp. 27°-28° 



Color. 



White- Yellow . . . 



Orange 



Brown 



Violet 



Blue 



3° 



Pressure. 



100-95 cm. 

 90-70 

 65-60 



55-50 

 50-40 



Color. 



Opaque 



Violet 



Blue 



Gray-Blue 



Faint Yellow, ... 

 Faint Purple. . . . 



V. 



21 



Pressure. 



35 cm. 

 33 

 30 

 27 



25 

 25-20 



7. Effect of Steam Pressure. — The following table 3 gives data bearing on the 

 practical question as to how far pressure is to be kept constant in work like the 

 present. The unlooked for independence of the color obtained, considered as sub- 

 ject to the pressure actuating the steam jet, is very striking. Persistence of color 

 thnnifdiout wide variations of pressure (much wider than are incident to the inves- 

 tigations below) is the rule. The observation is capable of two explanations: 

 either the influx of air keeps pace with the additional steam supplied by the jet 

 when the pressui'e rises, or the color criterion is not a sensitive one, showing indis- 

 tino-uishable gradations for large valuations of dust contents, etc. Probably both 

 reasons apply. In the table L/m denotes the number of liters per minute of air 

 passed over phosphorus to produce the full blue in cpiestiou. 



