1872.] The Construction of the Heavens. 309 
all been intended to subserve terrestrial purposes rather 
than to extend our knowledge of celestial relations. 
And it is particularly remarkable how few of the astro- 
nomers whose names stand highest in the roll of scientific 
fame, have taken any considerable degree of interest in the 
problems presented by the stellar heavens. Hipparchus 
and Ptolemy studied the stars in order to obtain information 
respecting the earth. Copernicus neither formed nor 
attempted to form any theory of the sidereal system. He 
expressed, indeed, the opinion that the universe of stars is 
spherical, but he based this opinion on considerations far 
removed from the actual study of the stars. Kepler, in like 
manner, did not direct his attention to the distribution of the 
stars, and the evidence which that distribution may afford 
respecting the constitution of the heavens; he, like Coper- 
nicus, expressed an opinion respecting the sphere of stars, 
but the opinion was based on fanciful analogies, not on 
observed facts.* Galileo limited himself to the observation 
of the stars with his telescope, enunciating only the theory 
that the Milky Way consists of a multitude of stars, as 
demonstrated by his telescopic researches. Hevelius did 
* His reasoning is so remarkable that I venture to quote it in this place, for 
it affords a very suggestive insight into the nature of Kepler’s mental organi- 
sation, especially when we consider that the work in which these ideas are put 
forward—the ‘“ Epitome,” was published partly in 1618, and partly in 1620, 
while the three laws of Kepler were published in 1609 and 1619; in other © 
words, that he was bringing out a farrago of fanciful and baseless speculations 
during the very period which saw him force from nature the key to one of the 
noblest of her secrets. After explaining that the stars are necessarily 
enclosed within the substance of a spherical shell having the sun at its centre, 
he proceeds to reason thus :—The radius of the concavity enclosing the sun is 
determined by a simple proportion. Saturn, the outermost planet, travels at a 
distance from the sun equal to 2000 times the sun’s radius; therefore by 
the harmony of relations the distance of the sphere of the fixed stars is equal 
to 2000 times the distance of Saturn from the sun. But the thickness of the 
crystalline (the spherical shell enclosing the stars) can also be determined. 
All the matter of which the universe is formed is divided into three equal 
parts. One third is included in the body of the sun; another third forms the 
substance of the planets and of the celestial ether which fills up the space 
within the sphere of the fixed stars; the remaining third forms the crystalline. 
Now since the ether fills a space exceeding the sun’s volume 64 trillions 
of times [(4,000,000)*], its density must be proportionately small compared 
with his. And the density of the crystalline must be a mean between the sun’s 
density and that of the ether. Thus the crystalline has a density 8000 million 
times [(4,000,000)?] less than the sun’s; and as its mass is equal to the 
sun’s, its volume is 8000 million times greater. Hence, since its inner 
diameter is known, it is easy to calculate its thickness, which is found to 
be equal to the 6oo00th part of the sun’s radius—or, according to nineteenth 
century measurements, to about seventy miles. 
It would appear, as Struve points out, that Kepler in thus reasoning was 
chiefly striving to accommodate the Copernican theory with the ideas enter- 
tained in his day respecting the waters above the firmament spoken of in the 
Pentateuch. 
