October 14, 1892.] 



SCIENCE. 



219 



from a least-square adjustment of all the observations between 

 1580 and 1890 : — 



D. 



(1). 



+ 6.24° + 17.75° sin [0.7° (i - 1850) + 112.7°] 

 ± .10 ± .31 ± .2 



"Where D stands for the declination at any time t, positive when 



•west, and 0.7° = = anarular motion in one year. The dif- 



514 

 ferences between the observed and computed values would seem 

 to indicate a fluctuating smaller period of about 80 years, having 

 a variable parameter of about J°. Somewhat similar results vpere 

 found at Paris by Mr. Schott, an'd at Christiania by Professor 

 Weyer. No attempt was made at present to establish this second 

 term, it lying within the probable error, which for a computed 

 result is ± 20'. 



The inclinations datevs^ith Norman's in 1576. The observations 

 between this dfite and 1S91 can be represented by the following 

 formula : — 



J= 70.40° — 3.98° sin [0.7° {t 

 ± .065 ± .09 



1850) + 33.0°] 

 ±.1 



(2). 



Where /= inclination at any time t and the period involved, 514 

 years — apparently the same as for (1). 



The probable error of a computed value is ± 10', which, con- 

 sidering the material, is satisfactory. Both (1) and (3) apply to 

 latitude 51° 30' and longitude 0° 07' west of Greenwich. The 

 mean of the Greenwich and Kew observations was taken to ap- 

 ply to this station. 'Comparing (1) and (2), a remarkable result 

 peculiar to this station will be noticed — that the epochs are prac- 

 tically complementary, hence the following approximate relation 

 betvyeen the declination and dip can be found : — 



— ^— -h '' =1 (3). 



Il7.75j= (3.98)" 



Where (! = Z)- 6.34 and i = /- 70 40. 



From (1) and (2) the following results are obtained : — 



Zero 



0.0° 



1660 I 

 1976' 



From which it appears that for London the mean declination 

 takes place about the epoch of maximum and minimum inclina- 

 tion, and vice versa. 



With the aid of formula (1) and (3) the curve described in space 

 by the north end of a free magnetic needle was now accurately 

 constructed and graphically exhibited. It was shown that the 

 first approximation of the curve could be taken as a spherical 

 ellipse, the period being about 500 years. For Paris, a similar 

 result was obtained and exhibited, using provisional formate. 

 For both stations the curve lay to the greater part west of the 

 true meridian, and the direction of the motion (standing at the 

 centre of the needle and looking towards the north end) was that 

 of the hands of a watch, or opposite to that of the earth. A 

 rough survey was then made of the globe in an easterly direction 

 approximately in the latitude of London, and elucidated by a 

 diagram. It was found that the needle was farther along in its 

 secular orbit at every station. The curves for some stations in 

 the southern hemisphere were also exhibited. 



The following conclusions were reached: — 



1. The direction of the secular motion of the north end of a 

 free magnetic needle in both hemispheres is that of the bands of 

 a watch. 



2. That if the secular orbit is a single closed curve, then are the 

 periods different for different stations. 



3. That if the period is a common one, then must the orbit be a 

 closed curve of two or more loops lying within each other. We 

 are getting then, at present, a small loop in America and a larger 

 one in Europe. 



4. That our present feeliug is rather that, strictly speaking, we 

 have no such thing as a i^riod of secular variation, but that the 

 needle partakes of a sort of spiral motion, returning approximately 

 to a former position, but never exactly so. 



Future study may possibly modify some of these conclusions. 

 The possibilities opened up by such a study as outlined were next 

 briefly alluded to, and reference made to a possibI>^ extension of 

 the well-known Gaussian analysis by the introduction of the 

 variable t — time. 



In conclusion, can we not say with Sabine : "Viewed in itself 

 and its various relations the magnetism of the earth cannot be 

 counted less than one of the most important branches of the physi- 

 cal history of the planet we inhabit." 



LETTERS TO THE EDITOR. 



»*♦ Correspondents are requested to be as brief as poasible. The writer's name 

 is in all cases required as proof of good faith. 



On request in advance, one hundred copies of the number containing his 

 communication will ^e furnished free to any correspondent. 



The editor roillbe glad to pnblish any queries consonant with the character 

 of the journal. 



On Biological Nomenclature. 



I HAVE read with interested attention the discussions by bota- 

 nists in Science on this subject. It would appear that they are 

 fully alive to the need of some canons of nomenclature in their 

 branch of biology — a need which has been felt, and supplied of 

 late years, by the zoologists. 



The nomenclature of botany has always seemed to me to be 

 more stable and uniform than that of zoology, for the reason, as 

 I supposed, that the naming of new genera and species has, for 

 the most part, been reserved to a comparatively few leaders in the 

 /Science; and the same cause has contributed to the fixity of bo- 

 tanical classification, in comparison with the incessant taxonomic 

 fluctuations which zoology has suffered. 



With the late great increase in the number of working botanists, 

 the distinction of a small select "caste" of authoritative namers 

 and describers in botany seems to be breaking down, with the 

 various good and evil results attending this transfer of power 

 from a privileged oligarchy to more democratic rule. 



I think that not improbably the botanists who are now exer- 

 cised over names may examine with much confidence the canons 

 of nomenclature lately formulated and rigidly enforced by the 

 American Ornithologists' Union. These rules have been found to 

 work admirably in practice. They may not be the best possible, but 

 on the whole they are the best extant. A number of leaders in other 

 departments of zoology, besides ornithology, as in mammalogy, 

 herpetolcgy, ichthyology, malacology, entomology, etc., have 

 found them entirely available. If some two or three of the rules are 

 not so acceptable as the rest, yet it seems to be generally conceded 

 that it is better to abide by them all, than to dissent from the 

 code as a whole on account of a few comparatively unimportant 

 points that may not be liked so well as the rest are. 



Referring to the excellent article of C. H. Tyler Townsend in 

 Saience of Sept. 16, it seems to me that the moot points he raises 

 have all been considered carefully by the ornithologists, who 

 have settled each of these questions to very general satisfaction ; 

 and that the considerations upon which their conclusions have 

 been reached are entirely applicable to the botanical questions in- 

 volved. 



I wish to say a word respecting the somewhat epigrammatic 

 rule, "once a synonym, always a synonym," for the form of 

 which I am measurably responsible, if I remember rightly. Like 

 any other curt sectentious saying, the rule is, as I perceive by Mr. 

 Townsend's remarks, liable to be misunderstood. There is no 

 question that, as he correctly says, "If a form which had been 

 described and then thought to be the same as some other species, 

 is later proven to be a valid species, the name originally proposed 

 should stand." Certainly it should. That is not the application 

 at all of the phrase " once a synonym, always a synonym ;" and I 

 never heard before of its application to the case Mr. Townsend 

 adduces. What the aphorism really means is best illustrated by 

 a concrete example: 



Let there be a genus Smithia in botany. Let a genus Jonesia 



