February 14, 1902.] 



SCIENCE. 



259 



thick, if the sun and moon were relatively 

 at rest during the period of the ' flash ' ; 

 but considering the gradual advance of the 

 moon in covering successive layers of the 

 sun's atmosphere, Ave see that in the emis- 

 sion spectrum the flash line of the fainter 

 gas would be manj^ times more intense than 

 that of the brighter. The absorption lines 

 of the two gases would be very nearly the 

 same. The extent of the metallic vapors 

 of the sun's surface probably varies in- 

 versely proportional to their atomic 

 weights. 



In consideration of these facts, it seems 

 altogether likely that the gases of the 

 metals of Group II. extend very high, and 

 are noAvhere very much condensed. The 

 flash lines are to be regarded as true re- 

 versals of the corresponding solar lines. 

 The metals of Groups I. and III. are some- 

 Avhat denser near the sun's surface and do 

 not extend so high as those of Group II., 

 but as it is the upper portions that con- 

 tribute most to the formation of the emis- 

 sion lines, the flash lines are to be regarded 

 as only partial reversals of the Fraunhofer 

 lines, the solar intensities being greater 

 than the flash intensities. Most of the 

 strongest lines in the solar spectrum have 

 been found in the flash ; and this, taken in 

 connection Avith the meaning of the differ- 

 ences of intensities, leads us to further re- 

 new our faith in the existence of the 're- 

 versing layer.' 



The Total Solar Eclipse of May 18, 1901 : 



C. D. Perrine. 



The expenses of an expedition to Padang, 

 Sumatra, from the Lick Observatory, to 

 secure observations of this eclipse, were de- 

 frayed by Mr. William H. Crocker, of San 

 Francisco. Eclipse day dawned Avith light 

 clouds covering the sky. But little change 

 occurred during the morning. At the time 

 of first contact, the sun shone through a 

 rift in the clouds. At the beginning of 



totality all parts of the sky near the sun 

 Avere covered Avith light cirrus clouds and 

 haze. The inner corona only and Mercury 

 and Venus could be seen during the early 

 part of totality. The clouds became very 

 much heavier towards the end of totality. 

 The time of beginning and ending of 

 totality Avas 3 or -4 seconds later than the 

 time of these phases computed from data 

 given in the American Epliemeris, but the 

 uncertainty of longitudes in Sumatra may 

 account for nearly if not all of this. 

 Twelve photographs of varying exposures 

 Avere secured with a camera of 40 feet focal 

 length. These shoAV the inner corona and 

 prominences as well, probably, as if the sky 

 had been free from clouds. The longest 

 exposure, one of 150 seconds, shows the 

 streamers to a distance of one and one- 

 third diameters from the limb — more than 

 could be seen Avith the unaided eye. A 

 number of small prominences are visible on 

 the east limb of the sun. One of these 

 at position-angle 115° is covered AAdth a 

 series of coronal hoods or envelopes. At- 

 tention is called to a remarkable disturb- 

 ance in the corona in the northeast quad- 

 rant. At a position-angle of about 65° 

 there is a small compact prominence, over 

 which there is a disturbed area resembling 

 roughly an inverted cone. From the ap- 

 parent apex of this area a number of 

 irregular streamers and masses of matter 

 radiate as if thrown out by an explosion. 

 I am not aware that a disturbance of this 

 kind has been observed before in the corona 

 proper. Eight photographs were secured 

 AAdth the Floyd telescope of 70 inches focal 

 length. These negatives shoAV the same ex- 

 tensions of corona as those taken Avith the 

 40-foot camera. Twelve negatives were se- 

 cured of six regions on either side of the 

 sun in the direction of his equator for the 

 purpose of detecting any planets existing 

 there. These negatives were obtained Avith 

 lenses of 3 inches aperture and 11 feet 4 



