June ii, 1891] 



NATURE 



m 



directions and to almost all degrees, some variations 

 becoming fixed while others remain indefinite. I am 

 at present examining (with the help of my former 

 student, Miss A. E. Warham, B.Sc.) the anatomical 

 characters of a number of colonies of various Botrylli 

 with the view of finding which characters, if any, can be 

 relied on in distinguishing species or "forms," and I 

 have just seen a series of ascidiozooids of Botryllus 

 sinaragdiis in which the branchial tentacles, usually re- 

 garded as important features in the diagnosis of species, 

 present all variations between eight and sixteen. Every 

 one of the numbers 8, 9, 10, 11, 12, 13, 14, 15, and 16, is 

 represented by one or more ascidiozooids, although 8 and 

 16 are those most commonly found. Also several definite 

 arrangements, such as 2 large pigmented tentacles and 6 

 small, 3 large pigmented and 13 small, are present, and 

 are connected by all possible gradations. Then, again, 

 we find that the smaller set of these tentacles may be all 

 alike, or may be of two sizes placed longer and shorter 

 alternately, or they may be 2 shorter and 4 longer, or 2 

 shorter and 5 longer, or 3 shorter and 5 longer, or 4 

 shorter and 5 longer, or 6 shorter and 5 longer, and so on 

 through the variations. Two or three of the extreme 

 forms, if examined by themselves, might easily be regarded 

 as distinct species. 



I have heard it said, and I fancy it may be often 

 thought, that since evolution has changed our conception 

 of a species, the modern biologist need not concern 

 himself with the description and nomenclature and 

 delimitation of those assemblages of variable forms which 

 are known as varieties and species. But to take such a 

 course would be a great mistake. The theory of evolution 

 has given taxonomy and speciography an additional and 

 a very real interest. Now that we know just how much 

 and how little the term species indicates, it has become 

 of great importance that species and varieties should be 

 re-studied from the evolutionary standpoint, that the re- 

 lations of allied forms should be carefully investigated, 

 the limits of their variation determined, and the effect of 

 their environment ascertained. The Botryllid^E form a 

 specially interesting group for such an investigation. 



Many of these more general remarks will no doubt 

 apply to other groups of organisms with as much force 

 as to the Tunicata, but some of the instances discussed 

 above may seem points of mere detail of no great general 

 interest. I believe, however, that they are typical cases 

 illustrating difficulties which may confront any specialist 

 in the course of his endeavour to attain to that important 

 object of biological investigation — a natural or genetic 

 classification of animals and plants. 



February. W. A. Herdman. 



PHOTO-STELLAR SPECTRAL 



pROF. PICKERING, while retaining the four types of 

 ■*- stellar spectra, finds that so many stars show an 

 intermediate stage of development, that, in the Draper 

 Catalogue, letters are substituted for the types. Thus, 

 let ters A to D denote stars of the first type ; E to L, 

 stars of the second type ; M, stars of Type III. ; while 

 N is reserved for fourth type stars. It seemed of some 

 interest to compare the photographic results with those 

 obtained directly with the spectroscope. For the first 

 and second types, the observations of Vogel (" Spect. 

 Beob.," — 1° to -f- 20°) were used. The stars in the first 

 four hours of RA. which occur in both works were ex- 

 amined and tabulated, those being rejected where there 

 was any uncertainty as to type in Vogel's observations. 

 The following table shows the results thus obtained : — 



' " Note on the Classification of Star Spectra in vol. .\xvii. Harvard 

 Annals, and on some Stars with Bright Lines." 



NO. 



I I 28, VOL. 44] 



Vogel. 



Eye observation. 



Class. 



I. ! ... 



II. ... 

 II.! ... 

 II. !!... 



Pickering. 

 Photographic observati( 



Letter. 



BE F H I K 



I 25 18 15 I I 



I 4 _ _ _ _ 



— 5 — 28 — 1 



— — — 2 — 2 



To show the differences in type, the following table has 

 been drawn up : — 



Pickering. 



Vogel. 



Stars. 



Number and Type. 



169 of I. 



42 of II, 



TypeL 

 105 

 4 



Type IL 

 64 

 38 



These tables show that, in the case of Type I., nearly 

 half the stars observed with the eye are really Type II. 

 according to the photographs ; in the case of Type I. !, 

 four out of the forty, although having a clearly pronounced 

 first type spectrum to the eye, are really second type stars 

 according to the photographs. In the case of the second 

 type, four stars out of forty-two are really first type. 



For the third type stars, Dundr (" Sur les Etoiles." &c.) 

 was consulted, and the following results were obtained: — 



DlTNER. 



Eye observation. 

 Type. 



III. 



III. ! 



III. !! 



III. !!! 



PiCKERI.VG. 



Photographic observation. 

 Letter. 



H I K M 



19 2 — 8 



24 2 I 22 



16 I I 24 



5 - - 12 



This table may be condensed as follows :- 



Dunes. Pickering. 



Type. Type. 



III. to III. ! 

 III. !!toIII. 



Total 



116 



The photographs therefore show that only 36 per cent, 

 are third type at all. In order to account for this very 

 remarkable result, the words of Prof. Pickering may be 

 quoted: — "The difference between this (the third) type 

 and the second is much less marked in the photographic 

 than in the visible portion of the spectrum. The most 

 noticeable difference is that, in spectra of the third type, 

 the intensity suddenly changes at the wave-length 4762. 

 Rays of greater wave-length than this are fainter than 

 those that are shorter." 



It will be seen that three stars of the third type appear 

 as first type stars on the photograph. These are : — 



(i) LL. 3717, ih. SSm. - 9"" o'-4, Dundr III.!!: "Les 

 bandes 2-9 sont fortement ddveloppdes, trfes larges et 

 sombres." 



(2) D.M. + i7'-i479,6h. 56m. + 17' 53'-8, Dunt^r III. !! : 

 " Les bandes 2-8, et peut-etre 9, sont visibles ; elles sont 

 tr6s larges et fort obscures autant dans le vert- bleu que 

 dans le rouge." 



(3) T^ Serpentis, I5h. 31m. ■\- \y 25'-9, Dun^r III. !! : 

 " Les bandes sont larges et fortes, surtout dans le vert 

 et dans le bleu." 



Prof. Pickering states, in the preface, that when the 

 brightness exceeds 6'5 it is difficult to classify the spec- 

 trum with certainty. The photographic magnitudes of 

 these stars are 665, 6"45, 6'44 respectively. 



As regards the fourth type, it is stated (p. 3) that " the 

 letter N is reserved for spectra of the fourth type, although 

 no star of this type is bright enough to appear in the 

 Draper Catalogue, owing to the red colour of all sucb 



