With reference to the nebular hypothesis of Laplace, Dr. A. R. Wallace argues that "if there exists a sun in a state of expansion in which our sun was when it extended to the orbit of Neptune, it would, even with a parallax of 1/60th of a second, show a disc of half a second, which could be seen with the Lick telescope." My reply to this objection is, that with such an expansion there would probably be very little "intrinsic brightness," and if luminous enough to be visible the spectrum would be that of a gaseous nebula, and no known _star_ gives such a spectrum. But some planetary nebulæ look like small stars, and with high powers on large telescopes would probably show a disc. On these considerations, Dr. Wallace's objection does not seem to be valid.
It is usually stated in popular works on astronomy that the spectra of gaseous nebulæ show only three or four bright lines on a faint continuous background. But this is quite incorrect. No less than forty bright lines have been seen and measured in the spectra of gaseous nebulæ.[373] This includes 2 lines of "nebulium," 11 of hydrogen, 5 of helium, 1 of oxygen (?), 3 of nitrogen (?), 1 of silicon (?), and 17 of an unknown substance. In the great nebulæ in Orion 30 bright lines have been photographed.[374]
D'Arrest found that "gaseous nebulæ are rarely met with outside the Milky Way, and never at a considerable distance from it."[375]
Mr. A. E. Fath thinks that "no spiral nebula investigated has a truly continuous spectrum." He finds that so feeble is the intensity of the light of the spiral nebulæ that, while a spectrogram of Arcturus can be secured with the Mills spectrograph "in less than two minutes," "an exposure of about 500 hours would be required for the great nebula in Andromeda, which is of the same spectral type."[376] Mr. Fath thinks that in the case of the Andromeda nebula, the "star cluster" theory "seems to be the only one that can at all adequately explain the spectrum obtained."[377]
Prof. Barnard finds that the great cluster in Hercules (Messier 13) is "composed of stars of different spectral types." This result was confirmed by Mr. Fath.[378]
From observations with the great 40-inch telescope of the Yerkes Observatory (U.S.A.), Prof. Barnard finds that the nucleus of the planetary nebula H. iv. 18 in Andromeda is variable to the extent of at least 3 magnitudes. At its brightest it is about the 12th magnitude; and the period seems to be about 28 days. Barnard says, "I think this is the first case in which the nucleus of a planetary or other nebula has been shown to be certainly variable." "The normal condition seems to be faint--the nucleus remaining bright for a few days only. In an ordinary telescope it looks like a small round disc of a bluish green colour." He estimated the brightness of the nebula as that of a star of 8·2 magnitude.[379] Even in a telescope of 4 inches aperture, this would be a fairly bright object. It lies about 3-1/2 degrees south-west of the star [Greek: i] Andromedæ.