538 On the Constitution of [September, 
unequal length by the culminating point of its heat-radiation 
(or the epoch of its greatest brightness) and the culminating 
point of its eradiation temperature, or that epoch in which 
the colour of its light approaches most nearly to the blue 
extremity of the spe< 5 trum. 
During the first period the eradiation of heat was copti- 
nually on the increase. At the beginning of this period, 
when the star is still in the condition of a nebulous spot, 
this change of state takes place very slowly, since at that 
time the emission of heat was very small. Subsequently it 
is much accelerated, when by increasing density the eradia- 
tion of heat reaches its culminating point. 
During the second period the eradiation of heat is already 
decreasing, whilst the temperature of radiation is still on the 
increase. The speed of the change of condition is at first 
great, and becomes subsequently gradually less when the 
temperature of eradiation reaches its maximum value. 
During the third period both the heat-radiation and the 
temperature of eradiation are constantly diminishing, and 
during this entire period the change is very slow. 
With reference to the second stage of development, the 
author has shown elsewhere that the relative speeds of the 
changes of condition of two fixed stars are as theii masses. 
The duration of the transition of a fixed star from the cul- 
minating point of the heat-radiation to the culminating 
point of the temperature of eradiation is approximately in- 
versely as its mass. 
The culminating point of the temperature of eradiation 
forms the beginning of the period of cooling, and the duia- 
tion of the latter will be in any case the longer, the greater 
the mass of the star. For the duration of cooling of the 
sun, or the assumption that its contraction will proceed at 
least until its density is everywhere as great as the present 
density at the centre of the earth, and supposing an equable 
decrease both of the surface-temperature and of the surface 
itself, we obtain as inferior limit a time of about 7b million 
years, and we may assume that during the greater part of 
this period of refrigeration— -therefore at least during the 
next 40 million years— the sun will continue to send out 
luminous heat-rays. 
The duration of the change of the sun from the culmi- 
nating point of its heat-radiation to the culminating point 
of the temperature of eradiation amounted, on the other 
hand, to about 4 million years only. For a fixed star whose 
mass is greater than the mass of the sun there will be a 
greater distance between these two epochs, and as, fuither, 
