Human knowledge of the atmospheres of the stars reverses the usual order of acquaintance with more familiar things.
We knew what they were made of long before we found out what they were like. Knowledge of their composition came suddenly about seventy years ago with the invention of the spectroscope. The work of Huggins and Secchi showed that the stars, like the sun, had continuous spectra crossed by dark lines--indicating that they wer e opaque, incandescent bodies surrounded by absorbing atmospheres. Some of the more prominent constituents of these atmospheres, such as hydrogen, magnesium, calcium, and iron, were immediately recognized and it became evident that the same kinds of matter were present, and doubtless the same natural laws operative to the utmost distance at which observations were possible. The star's spectrum, then, reflects its chemical composition, or, more precisely, the composition of the atmosphere. One might have expected to find very great variety among them; but from the earliest investigations it became evi.dent that such was not the case. Stellar spectra fall into a very few main types, the first characterized by a relatively small number of lines belonging with few exceptions to the permanent gases, hydrogen, oxygen, nitrogen, and the later * Presented at the Medal Meeting held Wednesday, May I6, I934.