« PreviousContinue »
investigations along exactly the lines printed in the Bulletin, with many others, but to have him do so in conjunction with simultaneous field studies * * * etc.
To one who is unacquainted with either party to the controversy, on the nature of which he is also very little posted, the following questions naturally arise:
1. By what uncharted route under the civil service did Dr. Buckingham reach his present position in the Department of Agriculture?
2. Does one who secures a position in the department by the means implied, consider himself owned by one outside of the department, so that the latter shall complain, ‘Is the creature greater than the creator?'
3. Is Dr. Buckingham owned in Madison, Wisconsin? Arthur John HOPKINS.
October 22, 1905.
THE METHOD OF ELIMINATION IN FIXING GENERIC
TYPES IN ZOOLOGICAL NOMENCLATURE. One of the most perplexing problems in zoological nomenclature is to decide on the proper application of a generic name used in a comprehensive sense by an early author, to one of the component parts of the original group. The genus of Linnæus and his followers of the eighteenth century corresponds fairly to the family of the twentieth century. It is agreed that a generic name should stand or fall by its typical species. But the writers of the eighteenth century had little conception of type-species in the modern sense. We must, therefore, find some method of fixing their types for them.
This may be done by choosing the best known European or officinal species,' to quote an expression attributed to Linnæus. When such a species is clearly indicated, this ought to settle the matter. But it does not do so in all cases, as some genera have no species either European or officinal. As many of the earlier writers took Linnæan specific names for their genera (tautonomy), it is safe to regard such a practise as fixing the type in question. Bodianus bodianus is an example of this sort. Virtual tautonomy (as Tetrao
tetrix, Scomber scombrus) amounts to the same thing.
The method of beginning with a leading species or chef de file, as typical representative of each genus to be described in full, while the others were disposed of in comparative sentences, was adopted by Lacépède, Cuvier, Valenciennes, Poey and other authors. In ichthyology this has given reason for the choice of the type of the genus by page precedence. This method was raised to the dignity of a universal rule by Dr. Bleeker and others. It is a pity that it was not systematically adopted earlier, for it would have given fixity, a matter which in nomenclature far outweighs all others. But Linnæus, among others, usually placed his type-species in the middle of the series, the less known or more aberrant forms at either end of it.
- The rule of the first reviser is generally recognized, and is given precedence over all other methods of fixing the type by many authors. The obiection to it is that no one has yet defined the first reviser, so as to separate his rights from the rights of different meddlers. If we admit none to be revisers, unless they definitely limit a genus and definitely associate its name with some one or more of its original constituent species, to the exclusion of others, this rule may be available, although its application involves a good deal of otherwise profitless labor in bibliography.
In recent years a rule of fixing types by elimination has come into vogue, the Amercan Ornithologists' Union having given it especial prominence. As a guide to the operations of a first reviser, who finds no type assigned by previous writers, the rule is not open to serious objection.
But it has been largely applied without regard to previous revisers, and the meanings of various generic names have been frequently shifted in accordance with its supposed demands. It is evident that it is in great need of definition.
For example, let A, B, C, D represent the species of a comprehensive genus called X. If each of these is successively made the type of a new genus U, V, Y, Z, then Z, the last of these, is invalid and its type, D, becomes the proper type of X. This is the simple condition of the problem. But let A and C be set off to form a new genus; C and D, another. Let a new genus be formed which would probably include B in it. Let still another be framed which might possibly include D. Let it be further uncertain whether A and B should be placed in different groups. Let still another writer definitely connect the old genus with A, while another uses it, not for any of its constituents, but for some new form probably congeneric with B, and you have a not unusual statement of the problem.
There is no way out of this by the rule of elimination. By accepting the first reviser rule, itself subject to the Linnæan rule and the rule of tautonomy, we may well fall back on the rule of page precedence, and let the rule of elimination be simply a recommendation to the first reviser, without direct validity of its own. This is the position of the rule of elimination in the new International Code.
I give two concrete illustrations of the difficulties of the rule of elimination among genera of fishes.
The genus Clupanodon Lacépède, 1803, was based on “toothless herrings,' the chef de file being Clupanodon thrissa. This species as described by Lacépède, is the Clupea thrissa of Broussonnet, the American species, later called oglinus by Le Sueur. This is, how ever, not the original Clupea thrissa of Linnæus, 1758, which was based on the Clupea thrissa of Osbeck, 1757, a Chinese species, later called Clupea nasus by Bloch, a species of Konosirus. The second species of Lacépède, nasicus, is the same as Clupea nasus of Bloch. The third, pilchardus, is the Clupea pilchardus of Linnæus, a species of Sardinia, which is probably the same as Sardinella. The fourth species of Lacépède, sinensis, is apparently the species called later Clupea ilisha, and is probably not the original sinensis of Linnæus. It is a species of Clupeonia or Harengula. The fifth, africanus, is a species of Ilisha, and the sixth, jussieui, is the original type of the genus Clupeonia.
Arranging these according to the modern genera:
1. thrissa. The type of Thrissa Rafinesque, 1815,
the name given as a substitute for the hybrid
name Clupanodon. Chatoëssus Cuvier, 1817, based on Lacépède's
thrissa, the generic name later transferred by Valenciennes to punctatus, the thrissa of
Linnæus. Opisthonema Gill, 1863, based on thrissa of
Lacépède = oglinus of Le Sueur. Konosirus Jordan & Snyder, 1900, based on
punctatus Schlegel, which is a congener of Clupea thrissa Linnæus (= Clupea nasus Bloch) and not of Clupanodon thrissa Lacé. pède, which is oglinus of Le Sueur. Most writers unite Konosirus with Dorosoma Rafinesque, 1829; but the two are probably
distinct. 2. nasicus. This is the original thrissa of Le Sueur
and is congeneric with Konosirus punctatus. 3. pilcharus. This has never been made type of a
genus. It is certainly congeneric with Sardinia Poey, 1870, with Amblygaster Bleeker, 1855, and I now think with Sardinella Valenciennes, 1845. Most writers (wrongly I think) unite all these with
Clupea Linnæus, 1758. 4. sinensis. This is referred by Valenciennes to
Clupeonia Valenciennes, 1845; which genus is probably identical with Harengula Valenciennes, 1845, earlier page. Most writers (I
think wrongly) place it in Clupea. 5. africanus. This is congeneric with the type of
Nisha Gray, 1836, and with that of Pellona
as type of a genus. 6. jussieui. Type of Clupeonia Valenciennes,
1845, apparently congeneric with types of Harengula and Kowala of the same author
on earlier pages. Usually referred to Clupea. By the first reviser' after Lacépède, Rafinesque, 1815, Thrissa is substituted for Clupanodon, and Lacépède's thrissa is doubtless to be taken as Rafinesque's type. By the next, Buchanan, 1822, ilisha (=sinensis Lac.) is described as a new species of Clupanodon. The genus Clupanodon then dropped out of notice until revived by Dr. Jordan in 1882, by a process of elimination for Clupeonia jussieui. Later the same writer, by another process of elimination, substituted Clupanodon for Sardinia. Still later, by the same process with further light, the newly defined genus Konosirus, being congeneric with Clupanodon nasicus, was suppressed in favor of Clupanodon. The change of result depends on the status assigned to K'onosirus, Sardinia and Clupeonia. By the process of elimination the name Clipanodon can be used for any one of several species, its use depending on the views one may hold of these closely related generic or subgeneric types. If restricted to the chef de file, the matter is at once settled. The species involved become:
Harengula jussicui. The rule of the first reviser, if the rule of the chef de file be disregarded, would cause Clupanodon to replace Harengula, Clupeonia and Kouala, jussieui being its type.
Another illustration is taken from the genus of flounders, Pleuronectes Linnæus, 1758.
In this genus, the European species mentioned by Linnæus and by Artedi, from whom the genus is derived, are: hippoglossus (type of Hippoglossus Cuvier,
1817). platessa (type of Platessa Cuvier, 1817). flesus (type of Flesus Moreau, 1873, a genus
very close to Platessa, perhaps, in fact,
identical). limanda (type of Limanda Gottsche, 1835). solea (type of Solea Quensel, 1803, of Solea
Rafinesque, 1810, and of Solea Cuvier, 1817). rhombus (type of Rhombus Cuvier, 1817, name
preoccupied: of Rhomboides Goldfusz, 1820, substitute name; also, as Bothus rumolo, the first species named under Bothus Ra
finesque, 1810). maximus (type of Psetta Swainson, 1839, not
Psettus Cuvier, 1817; first species named of
cludes also rhombus).
Scophthalmus and Bothus are based on three species each, the two categories being essentially the same, Scophthalmus being based on literature, Bothus on specimens. But the order is changed in the two cases, maximus occurring first under Scophthalmus,
rumolo (rhombus) under Bothus. Under Rhombus and Bothus and Scophthalmus, both maximus and rhombus are included, and Psetta, although based on maximus alone, by implication is a substitute for Rhombus.
The first reviser, Rafinesque, 1810, leaves no species in Pleuronectes, unless, as he refers all the other species to other genera, we might regard hippoglossus, which is not mentioned by him as the type of his Pleuronectes. The next reviser, Cuvier, 1817, recognizes the genus, Pleuronectes as used by Linnæus, but at once separates it into four genera or subgenera dropping the original name. These are Platessa (platessa, flesus, limanda), Hippoglossus (hippoglossus), Rhombus (maximus, rhombus) and Solea (solea). Meanwhile Solea had been set off previously by Quensel (1803) and by Rafinesque (1810), the latter author very erroneously referring to it, platessa, flesus and limanda also. Swainson, 1839, the next reviser, recognizes Pleuronectes (platessa), Hippoglossus (hippoglossus), Psetta (maximus) and Solea (solea). This is the first restricted use of Pleuronectes since the time of Linnæus and his followers. Later Pleuronectes was restricted by me to maximus by the rule of elimination, flesus being then regarded, as it is still regarded by most authors, as congeneric with platessa. Limanda is also near platessa. But neither limanda nor flesus is the best known European species' of the Linnæan genus Pleuronectes. The rule of the first reviser would fix Pleuronectes with , platessa, the rule of the best known species with platessa or maximus, the rule of elimination would place flesus as type of Pleuronectes, if defined as dealing with a species at a time. But Rafinesque took out solea, platessa and flesus together, to form his genus Solea, leaving only hippoglossus not provided for. This fact. some would hold. restricts Pleuronectes to P. hippoglossus. Cuvier next took out all the species, leaving no genus Pleuronectes, and placing Rhombus last, next to Solea. On the other hand, platessa was placed first by Cuvier, its subgenus Platessa being apparently the chef de file subgenus in Cuvier's genus Pleuronectes.
With this group nothing in particular can
be settled by the process of elimination unless to the ovicapsules of Harriotta and Rhinowe agree beforehand as to whether Flesus is chimæra lies in the absence of differentiation a valid genus, or as to what were the unex between the anterior and posterior portions pressed purposes of Rafinesque.
of the lateral alæ of the capsule and the But common usage and common sense uniformity of the transverse costal ridges all agree in placing platessa, the common Plaice, through. It differs from the ovicapsules of as the type of Pleuronectes.
both Harriotta and Rhinochimæra by the DAVID STARR JORDAN. greater width of the alæ and especially the
greater width and extension forward along the AN INTERESTING CRETACEOUS CHIMEROID sides of the archidome. The resemblance is EGG-CASE.
greatest to Rhinochimara. ALMOST nothing is known of the structural The genus Harriotta was set apart as the characteristics of the holocephalous fishes of type of a subfamily (IIarriottina) by Gill, in the Mesozoic period except dental plates or 1896, and it was associated with Rhinochimæra teeth. The remains of such, however, are in a family (Rhinochimærida) by Garman, numerous and about a score of generic names
in 1904. It is to this group (if a family, have been proposed for them, although A. properly nameable Harriottide) that the Smith Woodward only fully recognizes five, Wyoming fossil belongs. It can not be correGanodus, Ischyodus, Edaphodon, Callorhyn- lated with any one of the many generic names chus and Elasmodectes. I was, therefore, (Eumylodus, Mylognathus, Dipristis, Sphagemuch interested in a fossil which Drs. Frank pæa, Diphrissa, Bryactinus, Isotænia and H. Knowlton and T. W. Stanton referred to Leptomylus) that have been especially coined me for identification, if possible, and which for American Cretaceous fossils, but the I at once recognized as a chimæroid ovi- naming of it, if such must be done, I leave capsule apparently most nearly resembling that to Dr. Dean who is now publishing (through of modern deep-sea forms.
the Carnegie Institution) an elaborate work The interest arises from the assumption on the chimaroids. I have had the privilege that where likeness prevails between such of looking over the proof-sheets of that work products, not only the parts which frame them and my knowledge of the ovicapsules of the but other structures must correspond. The Harriottidæ is chiefly derived from it, though inference is not irrefragable, but in the ab- I had long ago seen those of Harriotta. sence of contradictory data, perfectly legiti- If these determinations prove correct and mate as a provisional hypothesis at least the groups named families by Garman are
The fossilized egg-cases previously known accepted as such the curious deduction folare few and the indications as to affinities lows that no fossil ovicapsule of a typical interesting as well as important. Three chimærid has been found as yet. figures have been published of Jurassic egg- Although the living harriottids are deepcases, two by Emil Bessels and one by Otto sea forms, it does not follow that a deep sea Jaekel. All are of the Callorhynchus type is indicated for the habitat of the extinct and it is significant that a right palatine harriottid. The character of the sandstone tooth,' obtained from the ‘Lower Greensand' as well as of the basin in which the ovicapsule of New Zealand, has been attributed by E. T. was found is opposed to the hypothesis of a Newton and Woodward to that genus and deep sea. It must be remembered, too, that named Callorhynchus hectori.
the same genus may have species ranging The newly found fossil was obtained by Mr. from shallow water to abyssal depths; ChimN. H. Darton, of the U. S. Geological Survey, æra, for example, has a species (C. colliei) from ‘massive sandstone' a few miles west of which may be caught from a city wharf and Laramie, Wyoming.
In the interest of conciseness of description The contour and general form are well pre. I would use archidome for the chamber for the
head and trunk of the chimæroid and urodome served but not the details. The resemblance
for that receiving the caudal portion.
another (C. affinis) which may descend to a depth of at least 1,300 fathoms.
ELECTROMETER FOR THE STAGE OF THE
MICROSCOPE. The capillary electrometer consists of a vertical tube drawn out at the lower end into a fine capillary and filled with mercury (Figs. 1 and 2). The upper end of the tube is joined to a cylinder in which a piston is
mercury in the capillary is kept from falling by the surface tension, it is obvious that whatever increases or diminishes the surface tension, for example an electric current, will raise or lower in corresponding measure the mercury in the capillary. The alteration in surface tension is accompanied by the movement of ions between the meniscus and the remaining electrode of the electrometer (the mercury in the acid reservoir). In practise it is found that this movement can be neither very rapid nor long continued, without injuring the sensitiveness of the instrument. The potential difference from even a single element (Daniell or dry cell) is far too large to be used safely. It is advisable to employ a potential divider, or rheochord, which shall permit only a fraction of the original potential (not more than 0.1 volt) to reach the electrometer.
The electrometer should be kept short-circuited, except during an observation, so that the capillary and the mercury in the reservoir may always be connected through a conductor. The short-circuit key is shown in Fig. 1. A strip of spring brass connected with one of the binding posts of the electrometer rests against a second piece of brass connected with the other binding post, except when depressed by the finger. The point of higher potential, when known, should always be connected with the capillary.
When the capillary electrometer is connected with two points of unlike potential the meniscus is displaced. The pressure necessary to bring it back to its original position is proportional to the electromotive force that displaced the meniscus. Thus by connecting the electrometer with known differences of potential it may be experimentally graduated. In practise, the relation between the pressure and the potential must frequently be redetermined. It is usually easier to measure differences of potential, such as the demarcation current of nerve or muscle, by compensation. In this method the electromotive forco of the demarcation current is measured in fractions of a Daniell cell, or any other constant element, by bringing into the same cir
moved by a screw, thus making pressure on the mercury column. The end of the capillary dips in a reservoir containing 20 per cent. sulphuric acid. A little mercury is placed in the reservoir. Platinum wires lead from this and the mercury in the capillary to convenient binding posts. When mercury is placed in the vertical tube it enters the capillary until the weight of the column of mercury is balanced by the surface tension. If the capillary be now dipped in the reservoir containing the sulphuric acid and the piston driven upward by its screw, mercury will be forced out of the capillary into the acid, and on lowering the pressure the mercury will retreat within the capillary, drawing the acid after it. As the