pot and
surrounding faculae, representing the various parts in their true mutual
relations. "I have ascertained in this way," he wrote,[449] "that the
faculae occupy the highest portions of the sun's photosphere, the spots
appearing like holes in the penumbrae, which appeared lower than the
regions surrounding them; in one case, parts of the faculae were
discovered to be sailing over a spot apparently at some considerable
height above it." Thus Wilson's inference as to the depressed nature of
spots received, after the lapse of not far from a century, proof of the
most simple, direct, and convincing kind. A careful application of
Wilson's own geometrical test gave results only a trifle less decisive.
Of 694 spots observed, 78 per cent. showed, as they traversed the disc,
the expected effects of perspective;[450] and their absence in the
remaining 22 per cent. might be explained by internal commotions
producing irregularities of structure. The absolute depth of
spot-cavities--at least of their sloping sides--was determined by Father
Secchi through measurement of the "parallax of profundity"[451]--that
is, of apparent displacements attendant on the sun's rotation, due to
depression below the sun's surface. He found that in every case it fell
short of 4,000 miles, and averaged not more than 1,321, corresponding,
on the terrestrial scale, to an excavation in the earth's crust of 1-1/5
miles. Of late, however, the reality of even this moderate amount of
depression has been denied. Mr. Howlett's persevering observations,
extending over a third of a century, the results of which were presented
to the Royal Astronomical Society in December, 1894,[452] availed to
shatter the consensus of opinion which had so long been maintained on
the subject of spot-structure.[453] It has become impossible any longer
to hold that it is uniformly cavernous; and what seem like actually
protruding umbrae are occasionally vouched for on unimpeachable
authority.[454] We can only infer that the forms of sun-spots are really
more various than had been supposed; that they are peculiarly subject to
disturbance; and that the level of the nuclei may rise and fall during
the phases of commotion, like lavas within volcanic craters.

The opinion of the Kew observers as to the nature of such disturbances
was strongly swayed by another curious result of the "statistical
method" of inquiry. They found that of 1,137 instances of spots
accompanied by faculae, 584