A difference in kind was very soon perceived to separate these objects into two well-marked classes. Its natural and obvious character was shown by its having struck several observers independently. The distinction of "cloud-prominences" from "flame-prominences" was announced by Lockyer, April 27; by Zöllner, June 2; and by Respighi, December 4, 1870. The first description are tranquil and relatively permanent, sometimes enduring without striking change for many days. Certain of the included species mimic terrestrial cloud-scenery--now appearing like fleecy cirrus transpenetrated with the red glow of sunset--now like prodigious masses of cumulo-stratus hanging heavily above the horizon. The solar clouds, however, have the peculiarity of possessing _stems_. Slender columns can ordinarily be seen to connect the surface of the chromosphere with its outlying portions. Hence the fantastic likeness to forest scenery presented by the long ranges of fiery trunks and foliage occasionally seeming to fringe the sun's limb. But while this mode of structure suggests an actual outpouring of incandescent material, certain facts require a different interpretation. At a distance, and quite apart from the chromosphere, prominences have been perceived, both by Secchi and Young, to _form_, just as clouds form in a clear sky, condensation being replaced by ignition. Filaments were then thrown out downward towards the chromosphere, and finally the usual appearance of a "stemmed prominence" was assumed. Still more remarkable was an observation made by Trouvelot at Harvard College Observatory, June 26, 1874.[603] A gigantic comma-shaped prominence, 82,000 miles high, vanished from before his eyes by a withdrawal of light as sudden as the passage of a flash of lightning. The same observer has frequently witnessed a gradual illumination or gradual extinction of such objects, testifying to changes in the thermal or electrical condition of matter already _in situ_.