June 16, 1922] 



SCIENCE 



651 



emphasizing some other words, with words denot- 

 ing family relationships, with suffixes, prefixes and 

 ease-endings, and sometimes, apparently, to in- 

 dicate punctuation. Although Quintilian says that 

 their proper use was to distinguish words and 

 case-endings which are alike in spelling but dif- 

 ferent in quantitj', that rule is comparatively 

 seldom followed in the extant inscriptions. 



A sketch of the modern faeroe dialect: J. 

 Dynelet Prince. 



The novcB or "new stars": E. E. Baenakd. 

 This paper dealt with the peculiarities of the 

 novEe or ' ' new stars. ' ' These are not new stars 

 in the ordinary sense of the word. They are stars 

 whose original condition was very faint or even 

 feeyond the reach of any telescope, that suddenly, 

 from some unknown cause, become very bright 

 even to the naked eye — in some cases increasing 

 their light as much as a hundred thousand fold. 

 They then fade away, perhaps never to become 

 bright agaiu. All of this great increase of bright- 

 ness occurs within a few hours ' time, or a few 

 days at most. The outburst of light is very sud- 

 den; the decline is at first rather rapid, then slow- 

 er, and with many halts and minor outbursts they 

 finally in a few years' time, say from eight to 

 ten or fifteen years, return to their original 

 brightness. This interval seems to vary with dif- 

 ferent stars. Some of these wonderful objects, 

 such as the great nova of 1572, have become 

 visible to the naked eye at midday. Two of them 

 in recent years became brighter than the first 

 magnitude. One of these, Nova Aquilae of 1918, 

 for one day outranked every star in the entire 

 heavens except Sirius and possibly Canopus. 

 At first we did not know anything of tlie previ- 

 ous history of these strange stars. They suddenly 

 appeared as if a new star had just been created. 

 But in recent years photographj' has added much 

 to our knowledge of them. Now when a nova 

 appears we search for it on our photographic 

 plates made before the star 's outburst. Some- 

 times we find that previous to this outburst the 

 star was beyond the reach of even the photo- 

 graphic plate, while in other eases they are shown 

 to have formerly existed as very small stars with 

 nothing to distinguish them from the millions of 

 other small stars that dot the sky. In several 

 instances we have found that previous to its out- 

 burst the nova had existed as a faint variable 

 star, rhythmically changing in brightness by a 

 small amount. The great star of 1901, Nova 

 Persei, is one of these that had probably existed 

 for ages as a small variable before it became 



a nova. Watched carefully now, this star fit- 

 fully varies through a couple of magnitudes, as 

 it probably did in its original condition. What 

 causes the tremendous outburst of light in these 

 wonderful stars is not known. 



The message of a meteorite: Monroe B. 

 Snydeb. 



The effect of diurnal variation of cloch rates 

 vpon longitude work: R. H. Tucker. From ob- 

 servations with the meridian circle, pendulum 

 clocks appear to run faster at night than the 

 average rate during a period of one day. The 

 excess each hour is small, but the daily rate at 

 midnight appears to be from two to three tenths 

 of a second larger than the daily rate at noon. 

 The largest error that would occur in predicting 

 the correction to an astronomical clock would be 

 between tv/o and three one-hundredths of a sec- 

 ond. The observed variation may be due to a 

 diurnal variation in the meridian plane. Such a 

 variation, with a period of fourteen months, does 

 occur, owing to the deviation of the axis of rota- 

 tion of the earth from the axis of figure of the 

 earth. There is a small diurnal term in the ob- 

 served latitude at the Lick Observatory, the full 

 amplitude of which is about three tenths of a 

 second of arc. Small corrections to the adopted 

 astronomical constants of aberration, or nutation, 

 may be indicated by these anomalies of observa- 

 tion. An exchange of longitude sigaals between 

 two stations, ninety degrees apart, might give a 

 resulting difference of longitude from two to three 

 one-hundredths of a second in error. Between two 

 stations on opposite sides of the earth the error 

 might be double that amount. Exchange of wire- 

 less signals, sent automatically by clocks across 

 the Atlantic, may give us a test of a variation 

 in clock rates. 



Discussion of a kinetic theory of gravitation, 

 II; and some new experiments in gravitation: 

 Charles F. Brush. 



Arc spectra and ionization potentials in dis- 

 sociated gases: K. T. Compton, with O. S. Dup- 

 FENDACK and P. S. Olmstead. The great com- 

 plexity of spectra of gases is due, in part, to the 

 fact that the molecules of the gas may exist in 

 various states of dissociation, association and 

 ionization, each type of molecule or atom giving 

 rise to its own characteristic spectrum. A dis- 

 covery of the exact state of the atoms or molecules 

 giving rise to each part of the spectrum of a sub- 

 stance is of great importance as regards both the 

 theory of spectral emission and the theory of 

 atomic and molecular structure. At the Palmer 



