490 



NA TURE 



[March 21, 1901 



the plotted observations from the graph of the law are to be 

 considered. For instance, when the equation is written 



y^ - axj -b — V, 



the divergence 5' is measured along the ordinate ; but when it is 

 written 



Ci - x^- d= 8", 

 the divergence S" is measured in the direction of the abscissa. 

 Now if the divergencies were measured at right angles to the 

 graph, and the sum of the squares were made a minimum, the 

 graph would be the principal axis of inertia of the system of 

 points. This line passes through the intersection of the other 

 pair of lines, and gives a smaller sum of squares than any other 

 line. When the number of points is very great, all three lines 

 become sensibly coincident. 



We may conclude, then, that when the observations are 

 numerous and fairly concordant, the method of least squares, 

 applied in the manner -commonly taught, will give a practically 

 unique result. But if in any particular case there be any doubt 

 on this point, by reason of the number of observations being 

 small, or the discrepancies between the observations being very 

 great, it would appear to be desirable to find both the lines 

 corresponding to the values of a, b and a' , b', given above, in order 

 to test the question. In the event of the difference between the 

 pairs of constants obtained not being negligeable, the proper 

 line to be made use of, in preference to either of the two others 

 found as above, would seem to be the principal axis of inertia. 



A. F. Ravenshear. 



I02 Croxted Road, West Dulwich, S.E. 



The Collection of Material for the Study of " Species." 



Strangely enough, while the publication of " The Origin of 

 Species" and the research which has been carried on since 

 Darwin wrote his epoch-making work have completely revolu- 

 tionised the morphologist's conception of what is a " species," 

 nearly all the " systematic " work which is published even at 

 the present day, especially in those branches of biological 

 .science where the amateur collector exerts most influence, is 

 based upon the principles founded by Linnaeus. These 

 principles, while they were perfectly logical in pre-Darwinian 

 days, are now, however, quite obsolete and out of harmony with 

 the current state of biological knowledge. With a view to 

 bringing scientific practice more into accord with scientific 

 theory, a paper to which I listened at a recent meeting of the 

 Linngean Society suggested a reform in the present system of 

 species nomenclature. Since, however, the great majority of 

 those who describe "species" are unfortunately not in a posi- 

 tion to realise how great indeed is the necessity of some such 

 reform, it will probably not obtain immediate support from the 

 systematists. The average " systematist " still holds the pre- 

 Darwinian view of "species"; and, as the great bulk of the 

 material at his disposal in public and private collections is of 

 little value for the proper study of taxonomy, he quite fails to 

 see how absolutely untenable his position really is. He does 

 not realise how utterly impossible U is in certain groups to assign 

 limits to the variability of "species," and it will never occur to 

 him that two specimens superficially alike in all respects may 

 quite possibly have been evolved along entirely diff'erent paths. 



It is not my intention here to enter into a discussion as to 

 whether a system of provisional names, such as is suggested by 

 Mr. Bernard, might not be of distinct advantage in at least 

 some branches of zoological inquiry ; what I do wish to call 

 attention to now is the necessity, if any decided advance is to be 

 made in the study of taxonomy, for a sweeping change in the 

 present system of collecting material, and in its representation 

 in collections. The following remarks bear particularly upon 

 the case of the mollusca, of which group I have personally most 

 knowledge, but they are, of course, more or less applicable to 

 other branches. 



The taxonomist requires as a basis for his investigations to 

 know as exactly as may be the range of variation which those 

 forms in which he is interested experience at the different stations 

 over which they occur. This knowledge is obviously most 

 satisfactorily acquired by personal observation on the spot ; 

 but, as this is seldom practicable, the student is in general 

 forced to rely upon material collected by others. Unless this 

 material has been properly collected, the conclusions based upon 

 it are likely to be erroneous ; and most of the material avail- 

 able to the student of such groups as the Mollusca is eminently 



NO. 1638, VOL. 63] 



unsatisfactory. Apparently the aim of many conchologists is 

 to represent (?) in their collections the greatest possible quantity 

 of "species," each by a certain definite number— usually three — 

 of "fine specimens." These maybe as unlike the ordinary 

 examples of the form as can be ; and, whether they are localised 

 or not is of little account, so long as they be fine. If a larger or 

 more brilliantly coloured specimen is obtainable it replaces one 

 of the mystic three. 



Large Series necessary. — UiliXt can be known about a species 

 until large series have been examined ; yet a collector or 

 museum curator almost invariably prefers a " new species " to a 

 specimen which might lead to a clearer understanding of others 

 already in the collection. Again, in most museums two or 

 three shells, for instance, are considered to amply represent a 

 species. When one has only a few examples under examination 

 it is a fairly simple matter to assign these to so-called species ; 

 but the task becomes very different when one comes to deal 

 with a large series, particularly if the specimens are from 

 different localities. In all groups the range of specific vari- 

 ability is very much greater than will be admitted by those who 

 confine their attention solely to museum specimens : in the 

 case of the marine Mollusca, it is often quite easy to select from 

 a large gathering of a single species two or more series which 

 will readily pass as distinct species if the intermediate forms are 

 suppressed. In the past, many "species" have been thus 

 formed ; and, if rumour speaks truly, this has sometimes been 

 done quite knowingly, the connecting forms having been care- 

 fully destroyed ; though more generally it has arisen in- 

 advertently through the study of insufficient material. Again, 

 a museum series, in addition to demonstrating the range of 

 variation of a form, should also illustrate its life-history ; but 

 only too frequently an immature individual is regarded by the 

 collector as a " bad specimen " and thrown away as valueless. 



Exact Localities. — The most important consideration in a 

 collection is that every specimen shall be accurately localised, 

 and the more minutely the exact locality has been recorded the 

 more valuable will the specimen be for study. At the present 

 day, perhaps, few collectors are satisfied with such records as 

 " Australia" or "America," but such scarcely less vague ones 

 as " S. Africa," " W. Indies," &c., are to be commonly met 

 with ; and specimens with inexact localities, or without any 

 record at all, abound in our museums. One unsatisfactory 

 •feature about specimens purchased from dealers is that there is 

 a temptation for the dealer to suppress the true origin of his 

 specimens. 



Those whose knowledge of species has been derived mainly 

 from museum specimens seldom realise how greatly these spe- 

 cies often vary in relation to their environment. Thus, in 

 the case of the marine Mollusca, specimens obtained from the 

 sandy portions of a shore will frequently differ perceptibly from 

 others of the same species collected on the neighbouring rocks 

 or mud. By keeping the series from different stations distinct, 

 the collector will often be surprised at the considerable local 

 variation which his specimens will manifest. 



Fossil ajtd Recent Forms. — The treatment of palaeontology as 

 a distinct science is one of the greatest obstacles in the way of a 

 proper appreciation of the problems of taxonomy. In most 

 museums, as in our own National collection, the fossil forms of 

 a group are widely removed from their recent allies ; and the 

 not unnatural result is that writers on existing species of, say, 

 the Mollusca, seldom make even the slightest reference to 

 fossil members of the group, A true knowledge of the relation- 

 ships of the living members of any group can only be attainable 

 by the study of those forms which have preceded them in the 

 process of evolution ; and this research will be greatly simplified 

 when recent and fossil forms can be examined side by side. 

 The comparison of recent specimens with the closely related 

 subfossil ones from the same locality and elsewhere is of most 

 particular importance, but is as yet seldom possible. 



The Condition of Specimens. — Among conchologists, and in 

 this respect they are almost the only sinners, insufficient atten- 

 tion is generally paid to the condition of the specimens. The 

 collector of shells too often prefers to gather up the mis- 

 cellaneous debris of a "shell beach" rather than search for 

 living examples, and unfortunately he is in the habit of founding 

 "species" upon material so obtained. A very large percentage 

 of Molluscan species has been based upon single, dead and 

 unlocalised specimens : what wonder that so many of their 

 names are absolutely worthless ? 



^^ Faking'^ Specimens. — Any interference with the natural 



