176 



SCIENCE 



[N. S. Vol. XXXII. No. 814 



Thus the three investigators of the a group 

 in the Martian spectrum, Slipher, Campbell 

 and Very, are in complete agreement as to the 

 failure of the psychrometer readings to give 

 reliable information about the humidity of the 

 total air column, vyhich is the important 

 datum of this test. If any further demonstra- 

 tion of this point is needed, it may be found 

 in Campbell's spectrogram No. 3, September 

 2, 1909, where the mean light paths were 5.15 

 for Mars, 3.61 for the moon, and the moisture 

 by sling psychrometer was about 2.9 grams 

 per cubic meter at the time of the Mars 

 spectrogram, but only about 0.3 grams ±, 

 when the lunar images were being recorded. 

 Yet notwithstanding the presence of a quan- 

 tity of terrestrial aqueous vapor about four- 

 teen times as great for the Mars spectrogram 

 as at the time of the lunar impression, if we 

 are to trust tjie meteorological records as Dr. 

 Abbot wishes, Professor Campbell merely 

 notes that little a " seems to be a shade 

 stronger in Mars than in the moon." Evi- 

 dently, either the psychrometer readings are 

 not to be relied on, or the photographic process 

 must have been very insensitive. Perhaps a 

 doubt may be permitted on both of these 

 scores. 



Dr. Abbot holds that, while the weather 

 may have been bad during a large part of his 

 stay on Mt. "Whitney, the conditions as to 

 humidity were favorable on the nights when 

 Director Campbell made his observations, and 

 that the spectrograms are " entirely conclu- 

 sive," while there is " no evidence at all of 

 water-vapor on Mars." Per contra, the fact 

 is that in spite of the low relative humidity 

 on the summit of the mountain on the nights 

 of September 1 and 2, we have no knowledge 

 of the conditions in the air column through 

 which the rays passed, except as these may be 

 surmised from the general seasonal and 

 regional meteorological data. The top of a 

 high mountain is the seat, during the day 

 time, of an abnormal local ascensional move- 

 ment of air, heated by contact with the inso- 

 lated slopes of the mountain. At night the 

 convection is reversed. Air from an elevation 

 above the summit descends, and with relative 



humidity reduced by virtue of compression 

 in the dovmward movement, a nocturnal ab- 

 normal condition of local dryness is liable to 

 be produced. In the free air, far from the 

 mountain top, quite other conditions may pre- 

 vail. 



The upper air is afFected by the great gen- 

 eral and seasonal movements of the atmos- 

 phere. In summer, a mantle of aqueous vapor 

 distributed through a wide range of altitude 

 prevents excessive radiation to the celestial 

 spaces. The night temperature at the summit 

 of the mountain in September descended to a 

 little below the freezing point, but this does 

 not indicate the complete removal of summer 

 conditions in the upper air over the whole 

 surrounding country. The depression of hu- 

 midity may have been, and probably was, 

 largely local, and in any case, considering the 

 altitude (14,500 feet), the cold was not ex- 

 ceptionally severe, and does not point to any 

 extensive withdrawal of a protective envelope 

 of vapor from the surrounding region, such 

 as occurs in winter. On the contrary, since 

 the weather over the whole southwest had been 

 for some time excessively rainy, the entire 

 air column over the region, taken as a whole, 

 was probably unusually replete with moisture. 

 The great mass of air through which an in- 

 clined and long line of sight passed, as in 

 Campbell's spectrogram No. 3, where the alti- 

 tudes were 11°. 2 and 16°.l, was compara- 

 tively unaffected by the local air movements 

 of the mountain top. The spectrograms prove 

 nothing definitely. Interpreting them as Dr. 

 Abbot would have us do by the psychrometer 

 readings, they are barely able to detect a vari- 

 ation of moisture in the ratio of fourteen to 

 one in the case noted above. By Campbell's 

 own account they are poor specimens, being 

 weak in the neighborhood of a, and having 

 other photographic defects. It is in the 

 photographic process that the real crux of the 

 problem lies. I can best illustrate this by an 

 example. 



Being engaged in a revision of Eowland's 

 intensities of the solar Fraunhofer lines, I 

 have had occasion to note the exceptional un- 

 certainty of those estimates which lie 'on the 



