OF ARTS AND SCIENCES. 163 



the circumstances under which the observations are necessarily made. 

 An insjiection of the charts shows us that the maxima occur at those 

 times of year when the Milky Way cannot interfere with the observa- 

 tions. This is true of the eastern, as well as of the western light. 

 We cannot very confidently assume that the interference of the Milky 

 Way is the cause of the observed lessening in the elongations attained 

 at other times. If it is, we must infer that, when the Milky Way in- 

 tersects the zodiacal light, its brightness prevents the observer from 

 tracing the light so far beyond it as he might in the absence of its 

 disturbing influence. It is probable that the question must be left to 

 the decision of photometric observations. 



We may next consider the fact, hitherto resting solely on the evi- 

 dence of Jones, that the ai)parent elongation of the zodiacal light 

 extends as the evening advances. Most observers have contented 

 themselves with a sinijle drawing of the li^ht for each date of obser- 



o o o 



vation, so that they give us no aid in the decision of the question. 

 But it is highly probable that the common statements with regard to 

 the rising and setting of the zodiacal light cannot be accepted without 

 modification. The testimony of one industrious observer to a multi- 

 tude of particular facts must ordinarily jDrevail over mere general 

 assertions. Jones paid special attention at times to the mode in 

 which the light disappeared, with varying results (Jj. 186, 190, 192, 

 196, 248). Tlie darkness of the sky whicli he observed near the 

 horizon, in the same stellar region where the light had been seen 

 earlier in the evening, is certainly not conclusive, for ordinary atmos- 

 pheric absorption may have produced it. On the whole, I should 

 infer from these passages that the final disappearance of the light 

 occurs by its setting, rather than by its fading. But this does not 

 change the customary result, apparent everywhere in the work of Jones, 

 that long after ordinary twilight has ended the zodiacal light continues 

 to extend itself towards the east, contrary to the diurnal motion of the 

 stars. I have myself noticed this phenomenon at different times, but 

 only casually, and on occasions when no leisure for a precise record 

 of it could he obtained. The vertex of the light seems to remain at 

 the same altitude for a long' time. Serpieri thinks that this and other 

 results of the observations of Jones require us to abandon all i^urely 

 cosmical theories of the zodiacal light, and oblige us to consider 

 it as a terrestrial phenomenon (Sp. 73, 152). But it seems to me 

 that before coming to this conclusion we must know more of the 

 nature of twili<rht than we do now. We need to know whether the 

 sky is really as dark just after the arch of distinct twilight has dis- 



