146 Notices respecting New Books, 



4th. Broad and pale lines and shadings in their proper breadths, 

 but produced by close thin lines drawn in any direction except the 

 vertical. 



These are the rigid principles on which the spectra are con- 

 structed in the author's plates. An example per contra, by way of 

 warning, is adduced from Brewster and Gladstone's map (ex- 

 ample I.), where every line in the Great- A band has been evidently 

 doubled, not by the observers but by the engraver. 



On the whole therefore the author decided to magnify all the 

 old drawings to his own scale; and to symbolize their shadings, 

 giving dates and the number of times linear of enlargement. 

 Hence, and from the small original scale of some of the maps, come 

 the " cyclopean " columns in some of the examples. 



Great A and its preliminary band take up two plates and 14 

 examples (subjects 1 and 2). The beautiful splitting up of the 

 preliminary band into a rhythmical series of double lines is well 

 shown in the Madeiran examples. The Royal Society's spectrum 

 is represented by a blur of shade for the band and an immensely 

 broad line for Great A, showing no splitting-up whatever, and in 

 this respect is in no way equal to Brewster and Gladstone's much 

 earlier drawing. 



As regards Great A itself, Prof. Langley's spectrum alone seems 

 to compare in number of lines shown with the two Madeiran 

 examples. 



While discussing Great A , the author points out the need of a 

 fine slit and accurate focus to bring out the doubling of the lines of 

 the preliminary band. He considers that the origin of these lines, 

 if telluric, is not watery vapour, but some dry gas in the upper 

 regions of our atmosphere, or in the millions of miles of space 

 beyond, and suggests a mode of research "by looking through 

 500-feet long metal tubes with glazed ends filled in succession with 

 every known gas, at a light of constant quality." 



" Little a " and its band of lines form subject 3. This line and 

 group were ascertained by Angstrom to be caused by invisible 

 watery vapour in the lower part of the earth's atmosphere. 



The earlier spectra fairly represent the object, except that we are 

 told that, in the Eoyal Society's plate, vertical "engraver's" shading- 

 lines mislead. The two Madeiran spectra show a great number of 

 lines ; and we find (example 5) a very interesting Lisbon spectrum 

 of the author's, with high sun and in a dry air, in which the lines 

 become very fine. On page 10 we have a suggestion that on the 

 Peak of TenerifEe they would possibly cease to have an apparent 

 existence in the solar spectrum. 



Great B with its preliminary and attached bands, as seen in a 

 high and low sun respectively (subjects 4, 5, 6, and 7), occupies 

 3 plates and 28 examples. Here (preliminary band) the earlier 

 map of Angstrom's, as respects double lines in rhythmical series, 

 contrasts favourably with the later maps until we come to the 

 Madeiran views. The author discusses whether one of Prof. Lang- 

 ley's lines is, or is not, outside a couplet in the series ; and records 



