159 
form of a hollow cone, which would account for the 
frequently observed apparent division of the tail into two 
lateral branches, for this hollow envelope being oblique to 
the line of sight at its borders a greater depth of illuminated 
matter would there be exposed to the eye. 
As the comet proceeds along its path it will project a 
newer shadow at an angle from that which it has already 
cast, the mist formed in which latter will be dispelled by 
the unimpeded action of the solar rays, whilst another 
portion of the comet’s atmosphere will suffer partial con- 
densation, thus causing the formation of a new tail and the 
dissipation of the old one to take place simultaneously, and 
accounting for the enormous sweep which the tail makes 
round the sun in perihelio in the manner of a rigid rod, and 
in seeming defiance of gravitation and all mechanical law. 
The extent to which condensation in the cometary atmos- 
phere will take place will obviously depend, amongst other 
things, on the difference of temperature within and without 
the shadow, and on the length of time during which that 
difference of temperature is allowed to operate. Now the 
further from the nucleus we go the fainter and the more 
diffuse the shadow will become; and apart from this, as 
well as in consequence thereof, the less the difference of 
temperature within and without that shade, and the longer 
the time required to effect a condensation. Accordingly 
the axis of the conoidal envelope will lag behind the axis 
of the shadow, the more so as we recede from the nucleus, 
thus producing the observed convexity on the tail’s orbital 
preceding side. 
The further we are from the nucleus, however, and for the 
same reason, the longer will be the time required to evapo- 
