gles yes Pee | Aitcibatelig sets eetbiet cigar ite bianca Salah cit sheli satiate dead este heros etalst fet tise ps ‘j : Pyaseey) Seta Cake ey tte: ; 4 herd} ree wes haere ata % Pept ide setts Adv rye Ra Maeeveuyee: A peaabae or Sint aint te orate paaaigtent saint isaebiogai tit nha mage sas feat nists Veni hd Fhe dte ahd atte reer bee fi Sas at ba Shai ataass as eae itn at ts bia 45 Metedd bi Pat ae tay ge {hy cae as Ne ir taien boieae os aoe pid sh +3 Pet ni ees ii le ase od these othe Heise hy bias *heeed Hepes Vehesss byeke- ae sfotatetett asa Nate tees eM, wane bd itt aii culate i a ati aerate et Posie 7 Bee cian deneat He Pee , feed mst rafts rie Moberiiaeee enbthwe resting. SP BAA bbe rte) +oeees. beats ted ; : ‘ Sh eaichytaset it eS Pires sete aticeneist é at ‘oe q Hit 4h 4) rea ait Ttave sensi rtp a “ 7 4 Tee phPeberve sree Keveeek is | ah Abee, evetet rae +3} tbee ech de ou ii eye poerbeies ie Tei Psteran f fete fon ina abaas tsar Me vires es Rs Srvcghses ted oper tthe ae tats ihctbh > efahineye 2h 24 eee tty ate veeh eo ss atti ge oe whe a ihineal ebeal blaine rs qhory ite! apie ited ssa 2 watelee} if rasta P y 1908 421 By tile of be r SheVer ese heeeral Beastie oe oH seid “ sha Bhtncateniy ts ae Repti vais sist the mater reset sine! Saieit clay etter: Aale ring na Hevcharetah hae padicidueie alee Beceesituie tyes. ae PPVOE) LER degre bint wes Paes aha laeaaahelaiediaaial Tapeh es ereye easy pepb vey, jobs a rharete hace) oe Ret feat Spear tte rb agen tas r iit Searsesgameacone Pah pe rdoseder. DREAD A ohne vp hat oe! Prhee “ 7 ri Me wed tei fF i alata reheat cod iv paariteatte Lents betei 8B MP NS be CPF Leb bee ny Liberate sae ati bbttalet aaenites nn va ee a Cad Keeyettt yalelavaseni Hat Tabi ceeleealedteets tial leiasbbrbs je5ry bstapapescoste Hee tri ies sia 4 fen thee sie ie (verre rsiitaeey: tiie vide sety, sete manthel eae Wert aah Maleate Bie saci? petit Hiei tteresan wrbed he] baptaet feat set eta se eheled eveeis Milasreae eet o cto Patelaletr betaine se rt $e rie init tert Shed reset LARA Mesa a some erie, as Piel b yo eow as op Mice sed; eatis ten treet be Are od! aatdh kaceiabsanmaenls WMitederetetttraatstatitises ess: iat nie pets we arity Tebtentematteiticmation ipteditate Mists Mysavese) top pine beeereted ihe tad goede. eee set +44 eee Raper i esl fuleeett rie ite me Sauateea siete ahah " Fett Beate ty atete th} Ber 1s a¢ete Demehalaleda keg beletaceront erat Shihrhth bbb pant bends + en ey be oid tote ehre er» ee 4 er Pee yhe ny ee) avs bat biahbebeahebeannanatarad teat cnetits weit Viidert pat lata ; tet it Boek best hestenbe ” eke os vuehas et hel cedsstassl its eee See ehh tped» ul sib \ ah reeebeebee tapes dere: ‘aha vi ‘¢94 iy ee ASH Sb ones Spe is heap ES i wae Raphi 4 ira primhe va ot y it aia i us iter! i Wtinie! hades 4 po bith sestrag rigtel ebay iphe jortdeohhgnenre tt Sha trettemite PRED S bse Slecdeaeerteeer es Liner try; wey Bias 4a) alee ain f ASO) mdr * Server hie reky sees hs thas ates Pea Hee tJ aheee bb bAstetets hel alton ein es a ae ehh tei naivaa SY Sh Sob Hhve hen h diy peep pond 7 AB ih 941 pent niet myst cS } emaeaetat ifs: fat ln fe ane UM Dy Sa ealbetye Peis aves Lababal eh Wied hays fy a} At -Vppek ¥4y any eohinby tee: _ ‘eaben peas bataesd techies froth Apseee sit peonons st feds . a <¢ OF meh Ne Ot at etre a bagea)« Ari erbiwe hips tae be b et ri be rt Vien et rhe the hone er eels os yeils ’ Theta) pstustie ts Aedes eee baies ages aren ron HanH bes goed AGE ela ¥ Vda baad bes y Ue eatt os bean aaa shana catered erbate iets aaite Vee gow oS eri igititas hateapesiss Spe saa Miata Alp tieda dictate epi reee shes pyieds ud mitt eh sabanape : TRING ese rae ft Yih as id lein'¢ oy +n Naser cee Sheet Fabs hesat oiteracanr yids Libra ious Ys ie naa vist Me Ag Pali Hehe ne sive capnnncts Papier * Soh tp Rete gaa p ary tore rebebnases as RSLERR TO ots byatierhsshpdede pepe tits bi ars she Mbp g Sta ber = aig ati He x ; - hae ayisdegits vhene Rute aeea : hye < HASORD SPE en esdied eat cece rsetey. Setnasie be ibuhabebseenel shatalisasbaiaaa heleer eee ers VEL ei brs ray dey Me | asta eke pies eave crite err stein a Satin agit antl ft iat at feet i oe a rAd Raiue VEL Gaeta frist y Feige pent paders dee 10} gard silith fader sh 5 a i shy So dit rere) i mye rid bevepien er i if Athans bas te tee + +4, a peyegeies thre hitele Sash aye ue peat eer oy asst Vien Ave etetire #4) Bab Tees boletacsa ree aT bP einiatelt dno bades reece tye byeteqsinas Midi te sgtielitdatarers Seid: The tes Fashet thé) abbtedalblaeen jetted Berar ttt rhe dhe, tite 4 ehdey sj 4s i8de le rg ep priblehsioaiioie tse Te Hit eat i tele a ie aa sleet ips ot ast) ited taetsistel 2 thee He Meee saaherthaarg de Vest aya OPE) © rhe le Fae % aaah Ver greta ie chai ii ieietaio Pian bere iege sti ley . + hae Me ) y apeethe shh Bomieie rpynbeleb vad ry Pty eye dias apes 91 ! cha sisletebe tie! rs 4) 619) + east then ibe tat i yeti: sae ae eine eae ah bers! is rae 4 ii fie tear bg bhte eter heat ora Bes . ht caren fie aia Rt asin ue i veh erat bi +) pig 19 996) pris Se dee M4 Rise ae ae Hee : {iSsateenbeletenesseasse pithy) 4 pene pepe eyed 7) i ete? 20 Pe x er jor Fosend bedasctgny cs pe phot y 4 * ‘hese as ietea.\ sitet ne en ees mia Hsnple yah } ee iathvere eee pir bene iettesd ote a Test Pwo e bbe by 7 da iter 1 ain jievee Moist beh Renee Girt as He ets Teles ishateansassioittetoaircatt tos gcogs oss es Brash ss gnepacgin ees | epee tse vere See way ed aierealtatae ties ree 77 Bistrones ss ur + yan) poe. eleee st Fy ; ra) stasetiaa tee PYIPEN bed hp hebonad reese Deno Sy athe ert Heit i = re as oe apt et seetatiehtt nsiflaiteeret Hambeloecibintens it Jura Joye Hate sels pitat: rey pipes treelunes tte aga euterear Hira hee Ayteab % eeris | is ee Airey 46 abetoratt (Pypthe Puseeit be tit dee Hisar ape ohh Slave © Ae briprias Wise + lead ois eres red be are seleka fp eiteat. + vy might rghars mitnl adhe ty AYeiele Rye yal | nob hs ie ciate ‘4 Sythe Hevea tien Teerhier beter ese Fok Bh ~ » + Penep eset 4h Ff wrest Westley abi * poe nebitad fh 922 75 beehse | b4 teeyig ba ble eed, Sh at ie bag Wel strat sGyetias «leas al t a cf ie felbrsbentatehetete Meese ies +tosperse i ite aaa t Hib ird\ atbdidiinyis He bet hes yoeeaaass +} 6 dotet obannes ty sabes i arepateley thus 4 heer tas Saiante sit paste amt beipaett sure partreadety beatetete loan an ia senea he i pore iets} aie seats aati oe sean lis Bie aa ii sees Si te eat ie Jive aaets ide YI at eet haem! i" teedersetl at +) es iets “ Bede sev ss bf Mig poe Jt'¢) (elebebe penser arte = penn tag setae Bete ; sys — ver . cai — - pireosten Pt yerchmbelebneed Hvhah a4 ra bear eel teit tedheh: pian tare eesgte Tyletel oF Seige on ‘ye pares ope vol gs 0a $044 stieeati praretys: pest Seersptit bas by beaten lehs beloncab eases togeedes =o ~ [aot s! Sesneprentonsoutersp ot arerbesiks if tate gears Stes (Guihant Erne ttis e ieneseneacoeesiee: Reel Free iels| Dhan 4 thetay Fone babe babe} iy ENIVERSITY: STUDIES. UNIVERSITY OF NEBRASKA Ra VOLUME III — ie. _ PUBLISHED BY THE UNIVERSITY (i ae . 1903 = _ é ‘ s PORES id ge ‘ / CONTENTS CLEMENTS—Greek and Latin in Biological Nomenclature. ..... EA EGE 1 ENGBERG—The Degree of Accuracy of Statistical Data............... 87 ~CARTMEL—The Anomalous Dispersion and Selective Absorption of MSM a Se Mik, wd tae Taki a eae ey vi era Sm Selene fs. ake 101 CARRIKER—Mallophaga from Birds of Costa Rica, Central America... 123 FryYE—George Sand and her French Style.......................00.. 199 PounD—Notes on Certain Negative Verb Contractions in the Present. 223 Moritz—On the Variation and Functional Relation of Certain Sen- tence-Constants in Standard Literature..................... 229 BATES—On the Errors in the Methods of Measuring the Rotary Polar- ation of Absorbing, Substances, ... 20.04 en eee ae 255 BaTEs—The Magnetic Rotary Dispersion of Solutions of Anomalous Bisporbing Substances... £05 Fc). veo 4. aN asda hdd ein: 265 Harcitt—Regeneration in Hydromedusae..............0.. 0200 eee. 275 WHITE—Some Peculiar Double Salts of Lead........................ 307 Bring homWemolres Ge pally atta meek ss... hed eile Pow catbvog howe 331 Moritz—On the Representation of Numkers as Quotients of Sums and Differences of Perfect Squarés.......5...5........4.... 355 »e-4¢ VS Vor. ITI. DECEMBER, 1902 UNIVERSITY STUDIES Published by the University of Nebrasse——— STUN Gin Uligg oo L. A. SHERMAN H. B. WARD W. G. L. TAYLOR F. M. FLING, EpItTor CONTENTS GREEK AND LATIN IN BIOLOGICAL NOMENCLATURE ; Frederic E. Clements LINCOLN, NEBRASKA Entered at the post-office in Lincoln, Nebraska, as second-class matter, as University Bulletin, Series 7, No. 13 8S] CONTENTS OF THE UNIVERSITY STUDIES, Vou. I No. l On the Transparency of the Ether By DEWITT B. BRACE On the Propriety of Retaining the Eighth Verb-Class in Sanscrit By A. H, EDGREN On the Auxiliary Verbs in the Romance Languages 7 By JOSEPH A. FONTAINE 7 No. 2 On the Conversion of Some of the Homologues of Benzol Phenol into Primary and Secondary Amines By RACHEL LLOYD Some Observations on the Sentence-Length in English Prose By L. A. SHERMAN On the Sounds and Inflections of the Cyprian Dialect By C. E. BENNETT No. 3 On the Determination of Specific Heat and of Latent Heat of Vapori- zation with the Vapor Calorimeter By HAROLD N. ALLEN On the Color Vocabulary of Children By HARRY K. WOLFE On the Development of the King’s Peace and the English Local Peace Magistracy By GEORGE E. HOWARD No. 4 On a New Order of Gigantic Fossils By EK. H. BARBOUR On Certain Facts and Principles in the Development of Form in Literature By L. A. SHERMAN On the DIKANIKOS LoGos in Euripides By JAMES T. LEES S1 TWNIVERSITY STUDIES Vor. Il DECEMBER, 1902 No. 1 Greek and Latin in Biological Nomenclature FREDERIC E. CLEMENTS “Nomina Veterum Graecorum et Romanorum plantis iniposita laudo, ad conspectum vero Recentiorum plurium horreo. Nec mirum factum! quis enim Tyro de nominibus fuit unquam in- ‘structus? quis unquam dedit circa denominationem plantarum praecepta, demonstrationes, exempla?” Linnaeus Critica Botan- fr ie ie Bae ty The following treatise is intended to serve as a compendium of the principles of word-formation in Greek and Latin of suf- ficient thoroughness to enable the biologist to construct in proper manner any derivative desired. Further than this, various un- fortunate usages which have obtained in nomenclature and the many types of malformations will be considered in detail, and suggestions will be made for their correction or elimination. The treatment throughout is based upon the conviction that no biologist should be content with a nomenclature that is doubtful or crude in its philology. On the other hand, ultra-purism, to- gether with the mooted questions pertaining solely to the classi- cal philologist, will be avoided, since nomenclature for the sake of uniformity and stability must rest upon the assured. For these reasons, also, it is felt that, while he must conform to the best usage of the language, the nomenclator must go a step fur- ther, and, in the case of uncertain or various usage, establish a definiteness which the language itself did not know. Further warrant is found for this in the fact that the careless hand of UNIVERSITY STUDIES, Vol. III, No. 1, December, 1902. 321 2 Frederic E. Clements analogy is always busy throughout the life of a language, and, also, in the fact that the lexicon must take account of all usage, with the result that the cruder derivatives of formative and de- cadent periods of the language are found alongside of the purer, or at least more refined forms of the classical period. While Kuntze’s important contributions and the Rochester Code have been notable achievements on the way toward nomen- clatural reform, it has been evident from the first that botanists had merely reached a temporary resting place, from which they must sooner or later go forward to the ultimate goal—a uniform and stable nomenclature and terminology of international recog- — nition. The failure to deal with the matter of generic types and word-formation, both only less important than the cardinal prin- ciple of priority, made a reopening of the question inevitable, an event which is rapidly being brought about by the increasing frequency of papers upon nomenclature. The zoologists, while they have not gone so far in certain lines as the botanists, have greatly anticipated them by their action at the Zoological Con- gress of 1901, when they agreed to place zoological nomen- | clature upon a classical basis. Sooner or later, botanists must take the same action. When this time comes, biological nomen- clature will be in a fair way to become a symmetrical, stable structure, based upon the two cardinal principles, priority and classicity. There can be little difference of opinion in regard to the repeated statement that nomenclature is merely an in- strument in the hands of the biologist, and there should be just as little question that the instrument should be a worthy and ready one. I. Classical Greek and Latin are the basis of scientific nomenclature. “Tdiotae imposuere nomina absurda.”” Linnaeus Philosophia Botanica 158 1751. There has never been any serious question concerning the necessity of a universal language for the natural sciences. The ancient and medieval development of biology, carried on first a22 Greek and Latin in Biological Nomenclature 3 by Greek and Roman philosophers, and then perforce by men who had at least some knowledge of Greek and Latin, deter- mined irrevocably that this scientific language should be Latin, immeasurably enriched by Greek derivatives. So natural and complete, indeed, was this linguistic heritage from the ancients and the herbalists that Linnaeus merely simplified the syntax, definitised the vocabulary, and modified the use of Latin, with its incorporated Greek, to obtain a great binomial system, with- out which taxonomy as it is to-day would have been impossible. Since Linnaeus, no botanist has questioned the right of Greek and Latin to constitute the language of science. DeCandolle did indeed point out the many advantages English would possess as an international means of communication between scientists, but it was hardly his thought that English would supplant Latin as the language of taxonomy. The realization of the sugges- tion, in view of the fact that biological publication is made in sixteen languages, among them Russian, Magyar, and Japanese, is anything but imminent. Yet, while biologists are agreed that Greek and Latin shall furnish the materials for nomenclature and terminology, their practice, unfortunately, is still very far from uniform. Personal and vernacular terms from all possible sources have increased to such an extent that nearly a sixth of our present generic and specific names are derived from ver- nacular tongues. The economy of time and intellectual effort obtained by the use of such names is so considerable that they will always appeal to the poorly prepared or indifferent descrip- tive biologist. But they offend all the. canons of uniformity and taste, and the real taxonomist, whose work is thorough and pains- taking from the first glimpse of a new organism to the final pub- lication of its name and diagnosis, will avoid them. The best Greek and the best Latin available are alone good enough for biological nomenclature. The Greek and Latin of Linnaeus were the work of no very certain hand, and should not constitute the standard, when a better standard is obtain- able. Linné’s knowledge of word-formation in Greek was often elusive, though his names are far superior as a rule to those of more recent coinage. Similarly, the formations of Byzantine 323 4 | Frederic E. Clements Greek and Late Latin, as well as those of many preclassic au- thors in both languages, have little value for the nomenclator. Classic Greek and Latin only can be fully satisfactory, since they are not merely the best Greek and Latin obtainable, but, also, because they present the best conditions for securing es- sential uniformity. Again, it should be clearly understood that classic Greek and Latin are not necessarily the Greek and Latin of the extreme purist. ne A name orterm is invalid unless constructed according to the principles of word formation in classic Greek or Latin; alternatives are to be reduced toa uniform basis. Retroactively, all terms improperly constructed shall be cor- rected, exceptin the case of words of uncertain or unknown etymology, when no correction shall be made if any proper Greek or Latin construction will give such a word, with a possible meaning. “Nomina generica ab uno vocabulo . . ._ fracto altera in- tegra composita Botanicis indigna sunt.’ Critica Botanica 29 1737. “Nomina generica ex duobus latinis vocabulis integris et con- junctis vix toleranda sunt.” J[bid., 26. This rule finds its warrant in the fact that uniformity is a first requisite of nomenclature as purity is of linguistics. A malformation is not only unpleasant as well as incorrect philo- logically, but it is also extremely unfortunate by reason of the complications which it introduces into nomenclature. The philol- ogist is satishied only with most skilful handling of derivatives that is possible. He will no more be guilty of a malformation or a hybrid than the true scientist will be capable of a bit of superficial or bungling work. The latter must then learn to look upon linguistic matters with the same conscientiousness that he uses in scientific investigation. Ultimately, however, he must be prepared to go farther than the philologist even, for the sake of uniformity. The latter is chiefly concerned with the devel- opment of a language, or group of languages, and with him slightly different or alternative forms are of advantage rather than a source of difficulty. In science, where the form and ap- 324 Le ae Sa et Ns Greek and Latin in Biological Nomenclature 5 plication of each name or term should be absolutely fixed, alter- native forms of words and alternative methods of composition lead mevitably to grave confusion. The nomenclator must in consequence outdo the philologist in his own field. When it is possible to obtain essentially the same derivative in several slightly different forms by varying the stem of the first term, the connecting vowel, or the form of the last term, or by pro- ceeding from alternative forms of the same word or stem, then the nomenclator must make the most intelligent choice possible in the selection of the best form to use, or the best principle to govern. In so doing, he will often strengthen the hands of the philologist, since it is a well-known fact that many alternative forms are merely the bungling creations of the decadent period of a language. In choosing a principle for guidance in dealing with alterna- tive forms and methods of derivation, several courses have been considered. The first plan was to follow the usage in the case of each particular word, but it soon became evident that no one but a specialist in philology would be able to make derivatives at all, since the usage varied repeatedly in words of the same group. A similar attempt was made with regard to the best usage, but, while this led to somewhat greater uniformity, the results were not much more satisfactory, and the labor involved was enormous. From the first it was seen that, while an occa- sional word would deviate more or less regularly from the for- mation typical for its group, as in the case of the imparisyllabic neuter, ordpa, oréuaros (mouth), which regularly enters into compo- sition in its shortened stem form, the philologically correct stem, or the correct connective, was overwhelmingly predominant. Furthermore, since such usage includes the best usage in all cases, it was concluded that uniformity and purity could best be obtained by making this the invariable usage for all the stems of any group, as well as for all combinations of each stem. The justification of such a rule may be readily found in a con- sideration of imparisyllabic stems, which have constituted the most fertile source of alternatives. The Greek neuters in -ya, gen. -paros, furnish a large number of examples in which. the 325 6 Frederic E. Clements shortened form of the nominative and the stem proper’ of the oblique cases alternate in word-formation. The Greek lexicon exhibits 1,782 neuters of this class, of which 231 appear in 969 derivatives as the first term. In the latter the proper stem ap- pears in 781 words, while the shortened form appears in 188 words. The alternation of these stems in Greek has of neces- sity given rise to corresponding alternatives in nomenclature. Thus, there are found Grammonema Ag. 1832 and Grammato- nema Kuetz. 1845, Lomaspora DC. 1821 and Lomatospora Reichenb. 1828, Spermodermia Tode 1790 and Spermatodermia Wallr. 1833, Stomotechium Lehm. 1818 and Stomatotechium Spach 1843. Unfortunate and confusing as the variants of the same generic name are, the case is very much worse when the variations of one stem furnish two otherwise valid generic names, as in the case of Dermatocarpus Eschw. 1823 and Dermocarpa Crouan 1858, Grammocarpus Seringe 1825 and Grammatocarpus Pres] 1831, Haemospermum Reinw. 1825 and Haematospermum (Wallich) Lindl. 1836. In the former, we are concerned merely with uniformity, desirable as that may be, while in the latter the validity of a generic name is destroyed because of its essen- tial identity with an earlier name, an identity of which the later author was probably unaware. Such fatal duplication of generic names can only be avoided by stringent rules for securing uni- formity in methods of derivation. Myosurus L. 1737 (Myo- suros Dill. 1719) and Myurus Endl. 1837 are again alternative forms of the same compound word, which have been applied to different genera. The former illustrates the rare and archaic type of syntactic composition, the latter follows the usual method of composition by stems. In the case of Coleosanthus Cassini April 1817 and Coleanthus Seidl July 1817, the latter, though correctly formed, falls by the working of priority before the former, which is a blunder, equally indefensible from the stand- point of syntactic or non-syntactic composition. Callitriche L. 1751 and Calothrix Ag. 1824 illustrate the confusion that arises from using alternative Greek words (xaAXt-, xaAds, beautiful) and from the variation of the termination of the last member of the compound. Either first term is correct, but their compounds are 326 Poe. Greek and Latin in Biological Nomenclature 7 identical in meaning and essentially so in derivation. They are to be regarded merely as different forms of the same compound, and Calothrix becomes a homonym. The confusion wrought by alternative forms and blundering construction is nowhere better shown than in the following series of names, belonging to five different genera: Asterothrix Cassini 1827 (Asterotrix Brogn. 1843, Asterothria Gren. 1850), Asterotrichion Link 1840 (As- terostrichion “Klotsch” 1840, Asterotrichium Witts.), Astero- trichia Zanard. 1843 Astrotricha DC. 1829 (Astrotrichia Rchb. 1837), and Asterotrichum Bonord. 1851. The retroactive application of this rule is imperative for the sake of uniformity and purity. By far the greater number of plant genera have already been recognized and named. The new names to be proposed for years to come will be relatively few, and a reform which affected even all of these would be barely worth while. Further than this, most new names are made after the pattern of names already in use, whether correctly or incor- rectly formed, a practice certain to perpetuate the blunders of the past. Arguments from the standpoint of purity are equally cogent, but, as they would perhaps appeal to the philologist alone, they will not be insisted upon here. A rule of this sort to be at all worth while must be retroactive, for by retroaction alone can confusion be avoided and uniformity secured. The retroactive operation of the rule must be so safeguarded, however, that changes for reasons of uniformity or purity will be made upon real and not upon supposititious grounds. Framers of generic names have been extremely careless in the matter of indicating etymologies, but this is not sufficient warrant for reconstruct- ing names upon the basis of supposed meanings. Many a genus has received a name of known or evident etymology, but of meaningless or mistaken application, a fact which should re- ‘strain us from correcting words of unknown derivation on the basis of an assumed etymology. In making changes to secure ‘a more uniform and stable nomenclature, the greatest care must be taken to minimize the error arising from personal judgment. In many words of uncertain etymology, several derivations are equally plausible, or at least possible, and the exercise of per- 327 8 Frederic E. Clements sonal choice would simply lead to greater confusion. Tor these reasons, changes in words where the etymology is not expressly indicated or clearly evident should not be made, unless the proper formation of such a word in Greek or Latin fails to give a name of any possible meaning. The correction of such words as fall under this rule can only be made upon the basis of greatest probability, which, unsatisfactory as it may be, will conduce to the ends sought. ; WORD FORMATION IN GREEK Greek words arise by derivation or by composition. In deri- vation, roots or stems acquire a new meaning through the addi- tion of a suffix, a termination having no separate existence in the language, except in the rare case of certain words which have lost their real significance and are now found only as suf- fixes. In composition, two, rarely more, words are united ac- cording to certain rules to form a new term, or compound, in which the meaning of each may be traced. Formation by pre- fixes is really a sort of composition, except in the case of a few inseparable particles, which properly belong under the head of derivation. For the sake of convenience, however, all formation by prefixes will be considered under composition. Greek has obtained its stems by derivation, i. e., by adding suffixes to roots, a process to which the origin of all simple words may be traced. Derivation belongs chiefly to the earlier devel- opment of the language, and, indeed, is very largely prehistoric, especially in the case of primary derivation. Composition, on the other hand, is a much later development, and must. have at- tained its maximum in the classical period of Greek literature. Both derivation and composition afford the biologist the means of coining new words. For various reasons, among them con- venience ‘and usage, scientific terms have been taken directly from, the Greek lexicon (sometimes, of course, they have been found already borrowed in Latin), or new words have been formed by composition. Formation by derivation is equally valid, and the fact that it almost invariably gives shorter words. leads one to wonder that it should not have come into general use. The reason may be found in the fact that word-formation 328 s Greek and Latin in Biological Nomenclature 9 in biological nomenclature has been far from scholarly, and that derivation requires much greater care and knowledge than com- position does. It is also true that the possibilities of derivation in Greek, though large, are necessarily limited by the relatively small number of suffixes; while the sources of composition are practically inexhaustible. DERIVATION Derivation consists in the addition of one or- more suffixes to the primitive, irreducible portion of a word, which is termed a root. It may be distinguished as primary when one suffix is added to the root, making a stem, secondary when a second suffix is added to the stem, tertiary when a third suffix is at- tached, and so forth. For convenience, however, we may follow Henry,* and term those derivatives primary in which the root carries a single suffix, and secondary, all those in which the stem thus formed has been modified by one or more accretions. Fur- thermore, derivatives are classed as verbal when the suffix added permits of conjugation, and nominal when it permits of inflec- tion. It is important that this be kept distinct from the fact that certain suffixes can be added only to verbal stems, while others can be attached only to nominal (denominative) ones. Nomen- clature is not concerned with the construction of verbal stems, and the suffixes which follow are those which form nominal stems, i. e., nouns and adjectives. Primary derivatives are formed by attaching the suffix imme- diately to the root, though rarely an adventicious -o- intervenes. Secondary derivatives are made in similar manner by adding the suffix directly to the stem. In both cases, the groups of let- ters thus brought into contact conform to certain general pho- netic principles of the language. For convenience in making the changes, which arise in this way in derivation and composition, a short summary of the phonetic mutations in Greek is given. Mutations peculiar to verbal stems are omitted. A more‘ com- plete account of these phonetic laws may be found in any of the more comprehensive grammars. | 1Henry, Victor. A Short Comparative Grammar of Greek and Latin, 102. 379 10 Frederic E. Clements GENERAL PHONETIC PRINCIPLES Aspirates In composition, aspirates (x (kh), -¢ (ph), @ (th) ) arise when a surd (x, 7,7), usually by elision of the final vowel of the stem of the first term, comes in contact with an initial as- pirated vowel of the second term. dék (a) -7pepa — dexypepos, ten-day é(é) -edpa — Epedpos, seated upon avt(¢)-dpos — avOopos, an opposite limit Very rarely, this influence is exerted through an interposing consonant. | Tétp(a)-imros = Tépirros, with four horses abreast Accumulation of Consonants As a rule, groups of consonants are modified to prevent harsh- ness. Generally, three successive consonants, or a con- sonant and a double consonant, are avoided, or one letter is dropped, unless the first or last is a liquid (A, #, v, p), or y before a palatal (x, y, x, &). méprros, fifth; oxAnpds, hard; oadmyé, trumpet In composition, final « or o of the first term may stand before two other consonants. exotpogy, dislocation; ékpOeipw, to destroy utterly The concurrence of two consonants, when it produces harsh- ness, is avoided in several ways. (1) When, by the transposition or loss of a letter, mw or y stands immediately before ’ or p, the corresponding son- ant (8, 8) is inserted. Hes (0s )-9epa = peonu(€)pa = peonuBpia, midday avnp, genitive, *av(e) pos = dvpds, = dvdp0s, man (2) A consonant is sometimes transposed to a more con- venient position. : mukvos, genitive, rvv€, nominative, meeting place Assimilation. Two explosives can occur together only when the latter is a dental (7, 6, 8). In such a group a palatal or labial must be of the same order, and another dental is changed to o; 330 AA ay o Greek and Latin in Biological Nomenclature II x and zw can alone stand before 7, y and f before 6, and x and ¢ before 6, while o may occur before all three. gAextixds, burning, from ¢A€yw, to burn; tpertyp; rubber, from tpiBw, to rub mA€yonv, entwined, from zA€xw, to twist; ypaBdnv, grazing, from ypadw, to grave oxiotds, cloven, from oxiLw, to cleave; reoréov, persuaded, from ze/@w, to persuade éx, from, always retains its final palatal in composition. exdnpos, foreign; kv, pustule Before o, B and ¢ become z, y and x become x, while 7, 6, and 6 are dropped; xe is then written € and zo is written y. xaArAvBos gen., xaAvB-s nom. = xdAvp, a Chalybean ypadw, ypap-cw = ypayw paotvyos gen., paotry-s = paorie, whip TpLxos gen., Opry-s = Opi, hair xXapiros YeN., XapiT-s = xapis, grace Aaprados gen., Aaprad-s = Aapras, torch KopvOos gen., Kopv6-s = kdpus, helmet This rule applies to such groups as -«r-, in which the r is first dropped and the « then passes into €. vukTos YeNn., vuKT-s—vvé, night Before p, labials (7, 8, ¢) become p, palatals (x, x) become y, and dentals (7, 6, 6, £) become o. Brer-pa (BArAérw) — BrEupa, glance; rTpiB-pa (Tp/Bw) = tpippa, anything rubbed; ored-pa(oredw) = oréupa, garland; wAex- pa (rAexw) = rA€ypa, anything plaited; revy-po ( revyw) — Tedypa, a work; ad-pa (adv) —dopa, song; oxLds-pya(oxilw ) —oxiopa, cleft; reO-ya (7etOw) — reiopa, cable. éx remains unchanged; ékpayya, a wax impression The dentals (7, 6, 6, €) are retained only before A, v, p. Before p, they become o (see above), as also before each other; before o they are dropped. meb-rixos (ew) = rewetixos, persuasive; 75-Gnua (7dopar) = noOnpa, delight; ozeppar-o1, dat. (oréppara) — orréppacr; rxbo-o1s (oxilw ) = exiors, cleaving. Before another liquid (A, p, p), v is assimilated to the liquid; be- 331 12 Frederic E. Clements fore a labial (a, B, ¢, w), it becomes p; before a palatal (x, y, x, €), it changes to y; before o, it is elided and the preceding vowel usually lengthens; before another y, it is usually retained. ToAdipnkyns (radu), very long; madtdAvTos, teaed again; taXippoa, back water; wadirAavys, wandering to and fro; op Breua (ovv), seam; ovtppvors, a growing together; ovpyadpa, harmony; zadcykvptos, fishing net; ovyyovos, congenital; ovyxpoos, of like color; ovygéw, to smooth by scraping; péAavos, peAav-s—peAas; daipovos, Sarpov-o1—daipoor; ylyavros, yryavt-s = ylyas. wv drops its final before o and a consonant, or before Z, but the v is simply assimilated before o and a vowel. ovoTna, system; ovlwpa, girdle ovocaopos, earthquake; cvoocwpos, united in one body IlaAuv assimilates its v before o and a vowel, and usually retains it before o and a consonant; before another y, it is either dropped or retained. madtoouros, rushing back; wadivoxios or saAdtoxios, deeply shaded madiv€wos, living again; wadivvostos Or raXivostos, returning Ayav always drops v, except where doubling or assimilation-takes place. ayavvipos, ayappoos Ev does not change its final before p, o, or €. évpiCos, rooted; évoracis, plan; évévvypu, to boil in Doubling of Consonants (1) In word-formation, initial p is generally doubled when it follows a vowel, but remains single after a diphthong. Suappwyy, gap; yAvkvppila, sweet root; etpupewv, broad-flowing; evpios, well-rooted : (2) An aspirate is never doubled, but the corresponding surd takes the place of the first. argo for Sada, ete. (3) Doubling is a frequent phenomenon (mostly in verbs and comparatives) when the suffix ya (c) follows the final con- sonant of root or stem; final «, y, x, and, rarely, other explosives, absorbing «, becoming oo, 8 becomes £, and » becomes AA. oe Be eo J OIF a Nt ei ae oe Greek and Latin in Biological Nomenclature 13 drdrdcow (dvdak-t-w), peiLwv (pey-t-wv), GAdros (dA-t-os). Metathesis, or transposition of this « takes place when it follows final v or p. Oepdrrawva — * Pepar-av-ya — beparravia TWTELPA — TwTNpP-L-a. Syncope or elision of a vowel often occurs in the middle of a word. matpos for atépos Contraction of vowels should be ignored, when an occasion for it might arise in forming scientific terms. NOUN SUFFIXES General -o- (-wos, m.) primary or secondary verbal oxytone: 6v-p0s, heart; é€p-tc-pos, strife » -pa- (-pm, f.), primary (or secondary?) verbal paroxytone: 6ép-n, heat -o- (-os, -ov, m. or n.) chiefly primary: vop-ds, pasture; Av«-os, wolf -a- (-y, f.) chiefly primary: ¢vy-7, flight; po-7, stream; Acv«-n, white poplar -i- (-ts, m. or f.) chiefly primary, paroxytone: zo0A-ts, city -ev-, -ov- (-yv, -wv, m. or f.) primary or secondary, mostly verbal: dpo-nv, male; €ix-dv, image, ai-dv, age -pev- (-pynv, m.) primary oxytone: A-pyv, harbor -pov- (-pov, m.) primary paroxytone: rép-yov, boundary -pyo- (-pvov, n., -pvn, f.) primary, usually oxytone: otpw-pvy, bed ~po- (-pos, m., -pa, f., -pov, n.) primary, mostly oxytone: ¢-pa, seat, d0-pov, gilt -ho- (-Aos, m., -Ay, f., -Aov, n.) primary, mostly oxytone: qv-dy, tribe; $v-Aov, class. -tAos, -7A7, -wAov, -wAn, are widely ex- tended false suffixes used after a consonant: they show the accretion of certain stem vowels. -vo- (-vos, m., -m, f., -vov, n.) primary, often oxytone: d7-vos, s'eep; zo.-vj, penalty; réx-vov, child. An adventicious a has given the suffix -avo- (-avos, -avy, -avov, seen in: oréd- avos, crown; pyx-av7, device; dpér-avov, scythe. 333 14 Frederic E, Clements -vi- (-vs, f.,) primary: wj-vs, wrath -To- (-rTos, m., -Ty, f.) primary, usually paroxytone: yép-ros, yard; Kol-Tn, bed -atT- (-ap, -wp, n.) primary: 77-ap, liver; vd-wp, water -ax- (-a€, m.) primary paroxytone: pv-ag, torrent; dpz-ag, robber -a0- (-as, f..) primary or secondary, verbal or denominative, oxy- tone: Aap7-ds, torch; €86-ou4-ds, week -10-, -.0- (-us, f., rarely m.) primary or secondary, mostly oxytone when feminine, and paroxytone when masculine: dépv-ts, bird; xde-is, key; wats, child; jpep-is, oak with edible acorns; BacwW-is, queen ~wt- (-ts, f., -, n.) primary paroxytone: xdp-is, grace; peA-c, honey goes -wT- (-ws, mM.) primary paroxytone; yéA-ws, laughter -w- (-@, -ws, m. or f.) primary, feminine oxytone, masculine par- oxytone: 7x-#, sound; 7p-ws, warrior -ep- (-yp, m.) primary oxytone: é-#p, atmosphere; aié-yp, ether -op- (-op, -wp, n.) primary paroxytone: d-op, sword; zéA-wp, prodigy Agent -qu- (-evs, m.) primary, verbal or denominative, oxytone: ypa¢- evs, writer -ev- (-evs, m.) secondary denominative oxytone: ypappar-evs, scribe ' -Tep- (-Typ, m., -Tepa, f..) primary or secondary verbal oxytone: Av-rHp, deliverer; vK-y-tHp, Conqueror -Top- (-Twp, M.,) primary or secondary verbal paroxytone: f7-Twp, orator; vx-d-Twp, Conqueror -Ta- (-Tys, mM.) primary oxytone or paroxytone: xpi-rys, judge Secondary (1) verbal, usually oxytone, with short primary vowel or sigma, vai-é-rys, inhabitant, épa-o-rys, lover, or with long primary vowel, w«-7-77s, conqueror, or with long primary vowel and sigma, épx-n-o-rys, dancer; (2) denom- inative, generally paroxytone, oix-é-rys, servant, deo-po-rns, prisoner. From these words, the stem vowel has come to remain attached to the suffix, giving the agent suffixes, “ITS, -€(TNS, -OTYS, LOTS. 334 Lo at Greek and Latin in Biological Nomenclature I Means or Instrument Tpo- (-Tpos, m., -tpa, f., -rpov, n.) primary or secondary verbal, feminine and neuter usually paroxytone: da-rpds, knife; py-Tpa, agreement; Bdx-rpov, staff; dp-o-rpov, plough -tho- (-TAos, m., -TAov, n.) primary, usually paroxytone: av-rAos, bucket; xv-rAov, liquid (-rAn, f.) secondary verbal, usually paroxytone: éyérAn, handle -Opo- (-Opov, n.) primary, usually paroxytone: dp-Opov, joint " (-Opa, f.) secondary verbal, meually paroxytone: xkot-j47-Opa, chamber -Odo- (-AAn, f., -OAov, n.) primary (or secondary?) paroxytone: @vs-OAov, sacred implement; yeve-6Ay, race -xo- (-kn, f.) primary paroxytone: 67-«n, box -ev- (evs, m.) secondary verbal or denominative Series auedy- evs, milkpail Result -yat- (-pa, n.) primary or secondary, verbal (secondary rarely de- nominative) accent recessive: pay-pya, palisade; od-pa, body; 8-Ayn-pa, bane -es- (-os, n.) primary, mostly paroxytone: PeA-os, dart; €A-os, marsh Place -Typio- (-THpiov, n.) primary or secondary verbal proparoxytone: diuxao-rypiov, court house -eto- (-efov, Nn.) primary or secondary denominative paroxytone: WA-€lov, prairie -wv- (-ov, m.) primary or secondary denominative oxytone: dpred-wv, vineyard -tpa- (-tpa, f.) primary or secondary verbal paroxytone: zaAai-o- tpa, place for wrestling Action -ri- (-tus, f.) primary verbal, usually paroxytone: ¢a-ris, speech -ot- (-ows, f.) primary or secondary verbal, usually paroxytone: v-o.s, nature; adpav-t-o1s, disappearance -owa- (-ova, f.) usually secondary verbal paroxytone: doxima-oia, testing 335 16 Frederic E. Clements Quality -a- (-ta, f..) primary or secondary denominative, mostly paroxy- tone: appov-ia, harmony -os-, -es- (-ws, m. or f.) primary oxytone: #-s, dawn (-os, n.) primary recessive: Bdp-os, weight, épev6-os, redness -rnt- (-rys, f.) secondary denominative paroxytone: Aert-d-rys, thinness; whence -oryr- (-érys, f.) mavr-drns, universality -ovva- (-cvvy, f..) secondary denominative paroxytone: cwdpo-cvvy, prudence State or Object f -dov- (-dwr, f..) secondary verbal oxytone: dAy-7-der, suffering -povo- (-povn, f.) primary verbal oxytone: yap-povy, joy -Tu- (-Tvs, f., -Tv, n.) primary: Bpw-rvs, meat; do-rv, town -vd, -v6- (-vs, f.) primary, often oxytone: yAap-vs, cloak Dinunutives -lo- (-uov, nN.) primary or secondary denominative paroxytone: arop-tov, little spore. Various suffixes of stems have be- come attached to this diminutive, giving the common di- minutive suffixes, -aptov, -wdvov, -vdptov, -vAALov, -vduov, all forming neuter proparoxytones. -toKo- (-toKos, m., -uoKy, f.) primary or secondary denominative paroxytone: veav-ioxos, youth; zud-icxy, little girl. This suffix sometimes combines with -vov to form a suffix -wrxvov, neuter proparoxytone; dom6d-‘cxov, small shield. Patronymics -6a- (-dys, m.) secondary denominative paroxytone -6- (-s (ds) f.) secondary denominative oxytone Stems of the first declension add the suffix directly: Boped-dys, son of Boreas; Boped-s, daughter of Boreas. Stems of the second declension replace o of the stem with t: IIpvap-idys, son of Priam; Ipiap-is, daughter of Priam. Those in -vo, however, change o to a, giving the suffixes -uddys and -tds. Stems of the third declension insert-s before the suffix, ev drop- ping the v before 1: Kexpom-idys, a son of Cecrops; Kexpoz- is, daughter of Cecrops. 336 yr Greck and Latin in Biological Nomenclature 17 . ADJECTIVE SUFFIXES General -es- (-3, m., f., -es, n.) primary, rarely secondary denominative oxytones: wevd-ys, false; evyev-ys, well-born; Arz-ap-ys, persistent -0-, -a- (-os, m., -y, -a, -os, f., -ov, n.) primary or secondary (de- nominative when secondary) always oxytone, except in compounds: WA-ds, 7, ov, bare; Enp-ds, 4, dv, dry; Bov-vopos, ov, grazed by cattle -ad- (-as, m., f.) primary oxytone: omop-ds, scattered; Aoy-as, selected -t0- (-ts, f..) secondary denominative oxytone, feminines of nouns or adjectives, most having become substanives: Ae\d-is, Delphian Ownership or Relation -to- (-tos, -ta, -cov) primary denominative proparoxytone: orvy-vos, hateful: secondary denominative; (1) the stem vowel may be elided before 1, as 6aXaoo-tos, marine, from OdAacoa, or (2) it may be retained, as 8/ka-vos, just, 8mo-v0s, similar, whence arise new forms of the same suffix, i.e., -atos, -ovos, -€l0S, wos, etc. -o.0- (-ovos) arose from adding -to- to stems in -r, but is now regularly used as a suffix; @avpa-ovos, wonderful. -L60- (-vdt0s ) arose from attaching -wo- to stems in -1d-, but has become a regular suffix (especially frequent in the neuter to form diminutives): 6aAaco-cdvos, marine. -Ko- (-kos, 9, ov) secondary denominative oxytone: ¢voi-Kds, natu- ral: whence has probably come -txo- (-«Kds), zoAep-uKds, warlike, dep-par-txds, cutaneous; whence -rxo- (-rTiKds ), espe- cially applied to nouns of agent in -rys. The addition of -xds to stems in -ua has given the suffixes -waxds, and -akos; to stems in -v, -vKés. Material -wwo- (-vo-) (-wvos, 9, ov) primary or secondary denominative pro- paroxytone: dpv-wos, oaken; €¥-A-wos, wooden The modification of the initial vowel of the suffix has pro- 337 18 Frederic E. Clements duced the suffix -nvo- (-nvos), which is a secondary verbal oxytone, zer-«-yves, winged. -to- (-€0s, -ea, -eov) secondary denominative proparoxytone: the nominative form arose from primary stems in -e, and the intervocalic « was then elided, dpyvpe-to-s — dpyvpeos, silver; por¥B8-eos, leaden. -.veo- (-wveos) secondary denominative paroxytone: formed by add- ing -to to -wwo; dyy-weos, oaken. Quality -.0- (-mos) primary oxytone: Oep-yos, hot. The addition of this suffix to stems in -7, dpa-ct-pds, active, has produced a secondary denominative suffix -ijmos, é6-4d-yos, eatable, and this, by further combination, has given -aAuos, «id dA1pos, beautiful. iving -po- (-pos) primary, nearly always oxytone: Aauz-pds, bright; épv6- pos, red: secondary, mostly denominative, usually oxytone: dav-e-pos, plain, whence the suffix -ypos; xuuat-npds, billowy. -Ao- (-Aos) primary, nearly always oxytone: de-Ads, timid. . Secondary denominative oxytone: ovy-y-Aés, silent, whence the suffixes -nAds, -wAds, etc.; dmam-ndos, deceitful; duapr- wXds, used to sin. Fulness -evT- (-es, -eooa, -ev) secondary denominative, usually paroxytone, the feminine proparoxytone: yapi-eas, graceful; zrepd-as, winged, whence the suffixes, -0es, -yes; oxi-das, shady, devdp-nes, woody. -vo- (-vos) primary oxytone: cepu-vos, holy -.vo- (-wos) primary and secondary oxytone: zéd-wés, quite level; 6p-e-.vos, mountainous, whence -ewvos; evdv-evos, quite cheerful. -pt- (-pus) primary paroxytone, (6-pis, skilful -adeo- (-adeos) secondary paroxytone: popy-adéos, strong; Wwp-adéos. itchy -pov- (-pwv, -wov) primary, usually paroxytone: i8-por, skilful Ability or Fitness -uko- (-tKos) secondary verbal oxytone: ypad-txos, able to write; apx-tkos, fit to rule 338 Greek and Latin in Biological Nomenclature 19 -TiKo- (-rTuKos) secondary verbal oxytone: zpax-rixds, practical -yto- (-yos) primary or secondary, mostly verbal, proparoxytone: wot-ysos, drinkable; @avd-o-yuos, deadly, hence -oupos? -ysato- (-twaios), troBoA-atos, spurious Time -wo- (-wos) usually secondary, denominative oxytone: jpep-wos, of day; érwp -wos of late summer; x6eo-wos, of yesterday Likeness -wde- (-ddys, m., f., -8es, n.) secondary denominative paroxytone, arising from etdos, ré, form, in composition as the last term, whence the form -oedys, and, by contraction, -odys; Av -0dys, like a marsh, marshy. This suffix is often used to indicate fulness, also. Verbal: Capability or Obligation -To- (-Tos) primary or secondary verbal oxytone: oyw-ros, split; kXv-Tos, renowned; @¢uA-7-Tds, loved -Teo- (-Teos) seconca-y verbal paroxytone: ¢iA-n-ré0s, lovable COMPOSITION Greek exhibits two types of composition, syntactic and non- syntactic. Syntactic composition is the union under a single ac- cent of two words, one being merely a modifier of the other and in the case demanded by this relation. Such forms arise often from juxtaposition, for reasons of convenience, and are not, properly speaking, compounds, e. g., kvvds-Baros, dog thorn (kiwr, kuvos, dog), pvos-wris, mouse ear (pido, pwvds, Mouse). The subor- dinate word is usually in the genitive, though, rarely, it may occur in practically any case. In non-syntactic composition, the two terms of the compound are morphologically coordinate, though the one is usually subordinated to the other in meaning. The second word is attached to the stem of the first in the same way that secondary suffixes are added to stems, with the very important exception that the final vowel of the first member has become a universal thematic vowel, or connective, e.g., paxpo-c7opa, Kopuvn-popa. Non-syntactic composition is the only real compo- sition. It is so overwhelmingly predominant in Greek that it 339 20 Frederic E. Clements alone needs to be taken into account. Indeed, syntactic compo- sition must be sedulously avoided by biologists, if confusion is to be prevented, and the few syntactic compounds already in existence in nomenclature should be made to conform to the rules for non-syntactic composition. Compound words consist of three elements, the first term, the connecting vowel, and the last term. For reasons of conven- ience, the last term will be considered first, then the connective, and, finally, under the first term, will be given a detailed exposi- tion of composition in the different classes of words. THE LAST TERM The last term is always a nominal stem, i. e., noun, adjective, or verbal adjective. The form of the last term is necessarily determined by its character, as follows: I. lf the last term is a noun, it may (1) stand without change, and the resulting compound is properly a substantive, though Greek often employs such words as adjectives, or (2) it may take adjectival endings, according to its declension, and the resulting compound is an adjective. Again, in Greek, prac- tically all compound adjectives may be used as substantives. II. If the last term is an adjective or verbal adjective, it may stand without change in the resulting compound, but usu- ally it becomes an adjective of two terminations (-o0s, m.,f., -ov, n., rarely, -ys, m., f., -es,n.). The adjective may take substantive suffixes, in which case the compound will, of course, be a noun. The following examples will illustrate the form of the last term and the character of the resulting compound. I.1. The last term is a noun, undergoing no change. 10d0-c7ropa, 7 (mods, 7od0s, 6, foot, oopa, 7, seed) foot-spore aiparo-KoKkos, 6 (atpa, aiparos, To, blood, koxkos, 6, berry) blood- berry ddio-crapvAy, 7 (Odis, Opios, Ohews, 6, snake, oradvAy, 7, bunch of grapes) briony 1 For accent of compounds, see Buttmann, 292. 340 Greek and Latin in Biological Nomenclature 21 E.ho-Onxy, 4 (Eidos, Eipeos, 7d, sword, OyKy, 9, box) scabbard mept-BAnua, 7d (epi, around, PdAjpa, BAnparos, ro, throw) covering F puxpo-KadvBy, 4 (pexpos, small, xadAvBy, 7, hut) small hut. I.2. The last term is a noun, changed to an adjective, usually by a suffix. The various changes of the noun depend upon its declension to a large extent.* a. If the final term is a noun of the first or second declension (stem in -a or -o, nominative, -ys, -as, -os, masculine, -y, -a, -os, fem., -ov, neut. ) the compound adjective will termi- nate in -os, masc. and fem., -ov, neut. év-rofor-os, -ov (ev, good, rogdrys, 6, archer) with good archers KaAXt-veavi-os (KaAXe-, beautiful, veavias, 6, youth) beautifully youthful modv-Aoy-os (zoAvs, much, Aoyos, 6, word) talkative NevKo-Kou-os (Aevkds, white, Kopy, 9, hair) white-haired evpv-xwp-os (cdpvs, broad, x#pa, 7, space) roomy Tpaxv-0d-0s (tpaxvs, rough, édos, 7, road) with rough roads Babv-dvdd-os (Babds, thick, PvAAov, 70, leaf) thick-leaved, leafy é. If the final term is a noun of the third declension with the stem in any consonant except y, p, 8, or -es, the compound adjective ends in -os, -ov. peAavo-prcB-os (peas, péAavos, black, PAcp, PAkBos, 7, vein) black-veined puxpo-pacrty-os (puxpds, short, paorié, paortiyos, 7, whip) short- ciliate rodv-opvib-os (zodvs, Many, dpvs, dprios, 6, 7, bird) abounding in birds mukvo-gapk-os (wu«vos, thick, odp£, capxos, 9, flesh) with firm flesh d-cwpart-os (4-, without, copa, cdpatos, ro, body) incorporeal xpuco-crop-os (xpvoeos, golden, orépa, ordpatos, to, mouth) golden-mouthed 1 This account has been largely based upon Miller, Scientific Names of Latin and Greek Derivation, 134. 341 22 & Frederic E. Clements If the final term is a noun of the third declension with the stem in v, p, or 8 (nom. o), the compound retains this form, i. e., it is properly a noun used adjectively. Some- times the noun is inflected in two genders, e. g., -wv, -ov, Or -wp, -op, or, more rarely, it takes the adjective termina- tion, -os, -ov. paxpo-xeip (paxpos, long, xeip, xerpos, 7, hand) long-armed aito-xdwv, ov (adres, self, xOuv, xPoves, 7, ground) native aito-xOov-os, ov, country and all okAnpo-ous (oxAypos, hard, ods, zodes, m., foot) hard-footed kako-7rous, -7ovv (Kakos, bad, wots, zodes, m., foot) with bad feet If the final term is a noun of the third declension with the stem in -es (gen.-eos, nom. -ys, m. f. -os, n.), the com- pound adjective will terminate in -ys, masc. and fem., -es, neut. Oeo-yevys, es (Geos, 6, God, yevos, yeveos, 70, race) born of God TeLxo-peAns (Teixos, Telxeos, TO, Wall, péAos, péAeos, To, music) walling by music moAv-avOns (zoAvs, much, avOos, avOeos, ro, flower) blossoming If the final term is a neuter noun of the third declension with the stem in -at, nom. -as, the compound adjective as a rule ends in -ws (contraction of -aos for -aros) masc. and fem., -wv neut., or, rarely, in -os, -ov. peyado-Kep-ws, wv (péyas, peyddov, large, Képas, Képatos, 76, horn) large-horned moAv-Tep-ws (aodvs, much, répas, tépatos, 76, wonder) full of wonder povo-Kepat-os (p0vos, Single, xépas, ro, horn) with one horn 6p0o-Kep-os, ov (é6p00s, upright, Képas, ro, horn) with upright horns yAukvu-Kpe-os (yAukis, Sweet, Kpéas, Kpéws (Kpéaros) +d, meat) sweet-meated If the final term is a noun of the third declension with the stem in the vowels, or v (-ts, -vs, nom. m., f., -t, -v, neut.), it retains this form; rarely it terminates in -os, -ov. moAv-txOus (zodvs, many, ixOvs, ixOvos, 6, fish) abounding in fish; also zroAv-tx6v-os, ov , Greek and Latin in Biological Nomenclature 23 mukvo-Opus (aukves, thick, dpvs, Spves, 7, oak) beset with oaks peAav-dpv-os, ov (péAas, wéAavos, black, dpvs, oak) dark with oak leaves xpve-olus (xpvoeos, golden, dus, dfews, 7, appearance) looking like gold pvyo-rodis (Pvyos , fleeing, rds, 7dAEws, 7, City) fleeing from a city Il.1. If the last term is an adjective, the compound usually be- comes an adjective of two terminations, -os, -ov; rarely, there is no change. In case it is used substantively, it may appear in any gender at the coiner’s pleasure. The compound may, moreover, pass into a noun by the addition of a substantive suffix. oTEVO-pakpos, ov (oTevos, Narrow, paxpos, long) long and narrow peAavo-daios, ov (péAas, péeAavos, black, dads, dusky) dark gray Aevko-epvOpos, ov (Aevkds, white, épvpds, red) whitish red Erepo-yAavkos, ov (€repos, different, yAavkds, gray) with one eye gray Aevko-peAas, atva, av (Aevkés, white, wéAas, black) whitish black ofv-yAvkus, ea, v (dfs, Sour, yAvKUs, Sweet) sourish sweet €repo-wv-ta (érepd-pwvos, of different voice) difference of tone fydnpo-ovvyn (€nAjpwv, jealous) jealousy II.2. Ifthe last term is a verbal adjective (in -os, -ros, or -reos), it may retain the active ending, -os, -ov, or the passive end- ing’, -Tos, or -ys, -es, may be substituted for either. dia-cTpod-os, ov (diuactpépw, to twist about) twisted Tept-Tpor-os (mepitpo7rw, to turn round) turned round mrepi-pep-ns, €s (repupepw, to carry round) revolving dvc-pab-ns (dvopabéw, to be slow in learning) hard to learn Svo-rax-Tos, ov (tacow, to arrange) disordered, irregular G-eriOw-Tos (Aeriddoua, to be scaly) not covered with scales THE CONNECTIVE (THEMATIC VOWEL) The connective in Greek compounds was originally the final vowel of the stem, or, in imparisyllabics, the vowel of the geni- tive. The connective -o- was originally, then, characteristic of 343 24 eos Frederic E. Clements nouns of the second declension, and of many stems of the third. By analogy, it spread to stems of the first declension, and the remaining stems of the third, and, finally, even to verbal stems. The overwhelming predominance of the connective -o- makes it advisable to disregard the use of the thematic vowel of each declension as a connective in making new compounds, and may be considered sufficient warrant for its insertion in compounds already constructed upon the basis of another thematic con- nective. Such duplicates as Corynephorus and Corynophorus ~ should be avoided for the sake of uniformity in spelling, if for no other reason, but when they are the names of different genera, as in the present case, they are altogether unfortunate. More- over, alternative connectives of this sort will always furnish oc- casion for similar blunders on the part of those not thoroughly conversant with the principles of Greek word-formation. The con- nectives which may be properly used with the different classes of first terms of compound words are shown in the following list of examples. The diversity of classical usage in this matter is a cogent argument for the use of -o-as a connective in all cases where the first term is a nominal stem, if not, indeed, every- where that a connective is required. 1. First declension: stem in -a, nom.; -a, -7, fem.; -as, -ys, masc. Garaca(a)-o-pvdrov (OdrAacca, 7, Sea, PvAAOv, 70, leaf) Thalas- sophyllum kepar(n)-0-oTrypa (Kedadry, n, head, oriypya, 76, mark) Cephal- ostigma oKa-piry (oxida, y, Shadow, ¢éAos, loved) Sciaphile §yrn- Popa (AyAH, 7, nipple, dopa, y, carrying) Thelephora kXext(1)-0-purov (KA€rrys, 6, thief, durov, 76, plant) Clepto- phytum 2. Second declension: stemin -o, nom., -os, masc. and fem., -ov neut. ) aoKo-Aeris (doxos, 6, leathern bag, Aeris, y, scale) Ascolepis paBdo-Kpwov (paBdos, 7, rod, Kptvov, 7d, lily) Rhabdocrinum Badav (0)-n- opos (BaAavos, 7, acorn, opés, carrying) Balano- phorus podo-devdpor ( pddov, 76, rose, devdpov, 70, tree) Rhododendrum 344 t ; . _ bas . 4 . Soo hes tae tee? hn a Vi ak ee ee i eh Greek and Latin in Biological Nomenclature 25 3. Third declension: (a) Stem extending in an explosive, i.e., any consonant except ao, #, Vv, A, p, OF -par. padixo-vdXov ( padié, padixos, 7, branch, ines 70, leaf) Rhad- icophyllum KnALtO(0)-avlos (Kndis, KynAidos, 7, Spot, avOos, TO, flower ) Celi- danthus. Kepato-oTopa (Képas, Képatos, To, horn, oropua, 70, mouth) Cerato- stoma do7i8(0)-n-popos (doris, doridos, y, round shield, dopés, bearing ) shield bearing Kepas-opos (xépas, To, horn, popds, bearing) bearing horns pedt-KoKkos (péAt, wédiTos, To, honey, KoxKos, 6, berry) Melicoccus ai-roXos (al, aiyos, 6,7, goat, -rodds (-KoAéw, dwell) goat-herd (6) Stem ending in a nasal or a liquid (7, A, p). axtwvo-aTpoBos (aris, axtivos, 7, ray, oTpoBos, 6, whirling) Ac- tinostrobus daipovo-poy (daipwv, daiuovos, 6, divinity, poy, 7, bush) Dae- monorops Ov(0)-avOn (Ois, Owes, 6, 9, heap, dune, av6y, 7, bloom) Thin- anthe dxpo-Qerov (akuwv, axpovos, 6, anvil, Geres, placed) anvil-block ddo-craxus (GAs, dXcs, 7, Sea, ordxus, 6, spike) Halostachys Gd-Opdak (GAs, 7, sea, Opidak, 7, lettuce) Halithridax Onpo-povov (Oyp, Onpos, 6, beast, povds, slaying) Therophonum yaatpo-xetdos (yaorryp, yaotpds (-épos), 9, belly, xetAos, 76, lip) Gastrochilus yaorep(o)-avOos (yaornp, 7, belly, avOos, ro, flower) Gasteranthus mup-popov (mip, updos, To, fire, Popes, bearing) Pyrphorum mupt-proyos (zip, 76, fire, dAoyos, blazing) flaming with fire (¢) Stem ending in -par-, nom,, -wa, neuters ypappato-Onxn (ypappa, ypdpparos, ro, line, Oyxn, 7, box) Gram- matothece depparo-BrAacros (d€pya, Bépparos, ro, skin, BAaorés 6, sprout) Dermatoblastus pvpato-orpwpa (Pipa, Piparos, To, tumor, otpOpa, ro, bed) Phymatostroma 26 Frederic E. Clements oTop.(o)-appnva (oTopa, oTopaTos, TO, mouth, appnv, appev, male) Stomarrhena oTopa-iuvyn (oTowa, To, MOuth, Atuvy, 7, lake) salt water lake (d@) Stem ending in -es; nom., -os, -7s, EU ce0e mostly neuters. (e Bero-creppa (Bédros, BédXeos, To, dart, oréupa, to, wreath) Belos- temma éXo-urov (€Aos, eAeos, TO, marsh, puro, aa plant) Hasoeyene 6po-paky (dpos, dpeos, TO, MOuntain, Pak7H, 7, lentil) Orophace 6peo-do€a (dpos, To, mountain, doga, 7, glory) Oreodoxa 6pes-Buos (pos, To, Mountain, Bios, living) living in the moun- tains épect-Tpopos (pos, To, Mountain, tpodes, nurtured) mountain- nurtured épe-oxwos (dpos, 76, Mountain, oxds, shadowed) over-shadowed by mountains Ev-n-hopos (Ethos, Eipeos, To, Sword, Popds, bearing ) armed wath the sword Bere-n-hopos (Bédros, Bédcos, 76, arrow, Popds, bearing) bearing arrows ) Stem ending in -c or -v; nom., -ts, -vs, masc., fem., -, -v, neut. moALo-devdpov (mdAts, moAwos, woAEws, y, City, dévdpov, ro, tree) Poliodendrum mroAua-vopos (modus, 7, City, vowos, dealing out) civic magistrate oAt-7ropbos (mous, 9, City, topGos, destroying) destroyer of cities TLypo-edns (Typos, Tlypuos, 7, tiger, etdos, 70, form) spotted ég.o-cKopodoy (ddus, duos, dpews, 6, Snake, oxcpodoy, 70, garlic, Ophioscorodum 6t0-7r0A0s (dis, dios, 6, 4, Sheep, -aodos (-KoAéw, dwell) shepherd ixOvo-peOn (ixOvs, ixOvos, 6, fish, wé6y, 7, strong drink) os methe vekvo-oToAos (veKus, veKvos, 6, dead body, orodds, ferrying) ferry- ing the dead | vexv-n-ToAos (véxvs, 6, dead body, -zoAds, dwelling among) hay- ing to do with the dead Botpv-opos (Borpus, Borpvos, 6, cluster of grapes, dopos, beat ing) bearing grapes “fs Bov-7Aevpov (Bots, Boos, 6, 9, OX, wAevpov, To, rib) Buslemne . 346 Greek and Latin in Biological Nomenclature 27 4. Verbal stems. When the first term is a verbal stem, it enters into composition with a thematic -e (the form of the second person singular present imperative of -w verbs), or with a sigmatic stem, -ov, resembling the sigmatic stem of aorists. The influence of analogy has been felt here also, in that both connectives occasionally yield to the -o- of noun stems, and, more rarely, «and of the verbal stems interchange or assimilate. ; depe-Borpus (pépw, bear, Borpus, Borpvos, 6, bunch of grapes) bearing bunches of grapes Avor-OpE (Avo, loose, Opié, tpixes, 7, hair) with loose hair depeo-Bios (dépw, bear, Bios, 6, life) bearing life repoe-hovn (dépw, bear, pov, 7, death) Persephone, bringer of death mepoe-Todis (répOw, destroy, ros, 7, City) sacker of cities dpxt-Oadracao0s (dpxw, rule, Gadacoa, 7, sea) ruling the sea Auro-oxwos (Aeirw, leave, oxia, 7, Shade) shadowless pupo-xwdvvos (pixtw, throw, xivduvos, 6, risk) venturesome THE FIRST TERM The first term of a compound may be a nominal stem (noun, pronoun, or adjective), an indeclinable particle (adverb, prepo- sition, or inseparable particle), or a verbal stem. The form of the first term will be that of its stem if this ends in -o; if the stem ends in -a, -o- will be substituted as the connective, and if it ends in -1, -v, or a consonant,-o- will be added as a connective. The connective is omitted in the case of an indeclinable particle, and it is regularly elided before an initial vowel of the last term. In the following examples intended to show the form in which first terms of various categories should enter into composition, the effect of analogy is extended over all first terms of compound words which take a connective, with the exception of adjectives in -vs, -ea, -v, and verbal stems. Its use might well be extended to verbals upon the analogy of Aetrw, which regularly enters into composition in the form, Auo-, but verbal first terms are rare in scientific compounds, and are rather to be discouraged on ac- count of the alternatives to which they are certain to give rise. 347 28 Frederic E. Clements From the standpoint of the biologist, the application of the con- nective -o- might well have been made universal, but in the case of adjectives in -vs, the use of the thematic -vas connective is so invariable that the addition of an -o, as it is found in noun stems of the same sort, was felt to be unwarranted. I. Nouns. 1. First declension; nominative singular feminine, -a, -7; mas- culine, -as, -7s. The stems of this declension are all orig- inally in -a, which is often modified into -y. In feminines, the stem is identical with the nominative singular; in mas- culines, the stem is obtained by dropping the termination, -o, of the nominative. metp(a)-o-piry (wérpa, 7, rock, iAos, loved, loving) Petrophile nHEp(a)-(0)-avOos (juepa, 4, day, avOos, rd, flower) hemeranthus, -um 25 paxatp(a)-(0)-avOnpa (pdxapa, 7, dagger, dvOnpos, flowery) Machaeranthera 8 Lwv(n)-0-Op€ (Lovn, 4, girdle, Opié, 9, hair) Zonothrix 2 Bope(a.)-o-purov (Bopeas, 6, north wind, ¢vrov, 76, plant) Boreo- phytum irmot(1)-0-pvddov (immotys, 6, horseman, PvAAov, 76, leaf) Hip- potophyllum yew-rvbis (v9, yea, 7, earth, wvfis, 7, box) Geopyxis 38: yeo-, 1; 9 in yy- 2. Second declension. The stems of this declension terminate in -o, rarely in -w, and are obtained for composition by dropping o of the nom. sing. of masc. and fem. and v of the neuter. Buw-dvrov (Bios, 6, life, purdév, ro, plant) Biophytum 43 Lepup(o)-avOos (Lépupos, o, west wind, avOos, rd, flower) Zephyr- anthus TOTO[LO-YELT WV (roTapos, 0, river, yeiTwv, 0, 7; neighbor ) Potamo- giton 18 dporo dopos (dpdaos, 7, dew, popds, bearing) Drosophorus 8 1 The first number after a compound indicates the number of times the proper stem is found in composition in the Greek lexicon as the first term. Other numbers indicate the frequency of alternatives. 348 Greek and Latin in Biological Nomenclature 29 taro-cerpa (tados, 7, glass, cetpa, 7, band) Hyalosira 7 pyyo-rrepis (pyyos, n, beech, rrepis, 7, fern) Phegopteris 1 isate(o)-avOos (indriov, 76, outer garment, dvos, 7d, flower) Himatianthus 12 io-dpaBn (lov, 70, violet, 8pa8y, 9, sort of mustard) Iodrabe 4 guKo-poppy (aikov, 70, fig, poppy, 7, form) Sycomorphe 18 oor0-Onxn (doréov, darodv, To, bone, O7Ky, 7, box) place for bones 20; écTe0-, 6 Aayw-xethos (Aayds, 6, hare, xeidos, 7d, lip) Lagochilus 20; Aayo-, 6 Ta(w)-oupa (tads, 6, peacock, ovpa, 7, tail) Tatira . Third declension. These may be either consonant or vowel stems. The stem is derived most readily by dropping the ending, -os, of the genitive singular. A. Consonant stems. (1) Stem ending in an explosive, i.e., any consonant except o, ,v, A, p, and rt in -war, nom. -~a; nominative singular ending in a double consonant, y or €, or in o. (a) Stem in a labial, 7, 8, ¢; nominative in y (labial+o). pur-o-yovatioy (pip, purds, 7, rush, yovariov, to, small joint) Rhi- pogonatium prcB-0-xitwv (pray, HrEBOs, y, Vein, xiTaHv, 6, frock) Phlebochi- ton 13 KatnAup-o-poppy (KarnArul, Kkat7pdihos, 7, ladder, popdy, 7, form) Cateliphomorphe (6) Stem ina palatal, x, y, x; nominative in € (palatal +o). dXwrek-(0)-ovpos (dAdeE, GXdreKos, 7, fox, odpd, 7, tail) Alope- curus I proy-(0)-axavOos (PrAvE, Proyds, 7, flame, dkavOos, 6, spiny plant) Phlogacanthus 14 dvvx-o-meTadov (dvvé, dvuxos, 6, Claw, rérador, 79, leaf) Onycho- petalum 2 (c) Stem in a dental, 7, 6, 0, or vr, v6, xr; nominative in o, * rarely in p dwr-o- boos (das, durcs, To, light, PoBds, fearing) Photophobus 27:1 iwavr-o-xaitn (iuds, tuavros, 6, thong, xaéry, 7, long hair) Himan- tochaete 10 349 30 Frederic E. Clements ddovT-0-KvkAos (d80vs, ddovTos, 6, tooth, KvKAos, 6, ring’) Odonto- cyclus 18 vuKT-o-puKys (VE, vuKTos, 7, night, pwd«ys, 6, mushroom) Nycto- myces 28:33 in vu«kti- kepat-o-aTvAs (Képas, Képatos, T6, horn, orvdAis, 7, pillar) Cerato- stylis 16:3 -oTeat-(0)-oria (oréap, oTeaTos, Td, tallow, dzds, 6, juice) Steatopia 3 idat-o-popa (tdwp, vdaros, 7d, water, Popa, 7, a carrying) Hyda- tophora 22 kAevd-o-vna (KAeis, KAedds, 7, key, hook, viva, ro, thread) Clidonema 6 xAapivd-o-pwovas (xAapts, yxAapvdos, 7, mantle, povas, 7, unit) Chlamydomonas 8 Kopv0-(0)-atodov (Képus, Képvbos, 7, helm, aidAos, nimble) Coryth- aeolum EApwO-0-craxvs (€Apuvs, EApuvOos, 4, worm, ordxus, 0, spike) Hel- minthostachys 2 (2) Stem ending in a liquid, v, A, p; nominative in the same consonant, or o. (a) Stem in v, nominative in y, or oc. xLov-0-pirn (xLdv, XLOVOS, 7, Snow, PiAos, loving ) Chionophile 17 kAwy-0-oTaxus (KAW, KAwvos, 6, Shoot, oraxus, 6, spike) Clono- stachys Xnv-0-mrod.ov (Hv, XnVvos, 6, 7, FOOSE, 7odiov, To, small foot) Cheno- podium 13 deAdw-(0)-aorpov (dergis, deAfivos, 6, dolphin, aorpov, ro, star) Delphinastrum 3 xtev-(0)-odous (Kreis, KTevos, 6, comb, ddovs, 0, tooth) Ctenodus 2 (6) Stem in A, nominative in A, or o. GA-o-ductvov (GAs, dAos, 7, sea, dikrvoy, ro, net) Halodictyum 4:79 in dA- (c) Stem in p, nominative in p. avOp-o-rwywv (avyp, avdpes, 6, man, tweywv, 6, beard) Andropo- gon II9 aarep-(0)-ouparos (aornp, aorépos, 0, star, dudadds, o, navel) Asteromphalus 19 | 350° Greek and Latin in Biological Nomenclature C2 i yaotp-o-Kappy (ysoTHp, yaotpos, 7, belly, xapdy, y, dry scale) Gastrocarphe 1:16 in yaorpo- Onp-o-popdy (dyp, Onpds, 6, beast, opy, 7, form) Theromorphe 44 mup-o-Aeipiov (zip, mupds, TO, fire, Aefprov, 7d, lily) Pyrolirium 49:64 in -t-; II in p-; 4 in-7- (3) Stem in -par, nominative in -pa. ~ A special study of this class of imparisyllabics has been made for the sake of determining just what warrant ex- isted in Greek for making the stem of the oblique cases the invariable form for composition, These were thought to constitute a very fair criterion on account of their wide extension, and for the further reason that they furnish a large number of alternatives, since the nominative might readily be supposed to represent a first declensional stem ina. The Greek lexicon contains 1,782 neuters in -pa, largely secondary stems, though there are also many pri- mary ones, these being by far the most frequent in compo- sition. Of these 1,782 neuters, 231 are found in compo- sition or derivation as the first term, occurring altogether in 969 derivatives. Of the 231, 208 occur in derivatives only in the proper stem form in -yar, being used 555 times. Eleven words, dyadpa, glory (13:1), dopa, chariot (24:1), dépya, skin (9:1), Oadpua, marvel (14:13), Jedpnua, theory (2:1), képya, small change (3:1), «va, wave (19:11), K@pa, coma (2:1), dvouu, name (20:12), capa, body (48:5) and Aeypa, flame (12:4), show the alterna- tive stems, a@yaApar- and ayadp-, though the former is pre- ponderant, occurring in 166 derivatives, while the latter is found in only 5:1. Six words of this class, aiua,: blood (32:72), €pua, prop, (2:3), w@pa, drink (3:4), o7éppa, seed (16:22), oréua, mouth (3:24), and xeia, cold (1:4), occur more frequently in the shortened form, aiu-, the fre- quency being 129 to 57. Three only, Anya, bane (1), otddaypa, drop (1), and ¢paypua, fence (3), are found in- variably in the shortened form, while three, érdrwpa, mapéyxupa, and ofypa occur once in each form. To sum- 351 32 Frederic IE. Clements marize the foregoing: of 231 neuters in -wa, which furnish stems for compounds or derivatives, 208 always appear in the proper stem form, -war, 11 occur more frequently in this form, 6 more frequently in the shortened form, -», 3 always in this short form, while 3 occur once in either form. Of 969 words derived from these neuters, 781 show the proper stem in -yar, while 188 have the shortened stem in -p. Again, it must be borne in mind that, while these alterna- tive stems are a source of growth rather than a misfortune to the language, in nomenclature they must always lead to confusion, as analogy will soon: « or later produce doub- lets, such as aiparcoreppa and aipcowepwa, in the case of every stem which enters into composition. The marked preponderance of the proner stem in compounds of this group has been considered ample warrant for extending this stem to all compounds formed from neuters in -ya. If further warrant were needed, it is found in the fact that every neuter of this class shows the proper stem in the oblique cases, its disappearance in certain compounds be- ing due to the use of the shortened nominative form, a use arising to a large extent out of ignorance. aipat-o-xapis (aipa, atparos, TO, blood, yédpus, 7, grace) Haema- tocharis deppat-o-KuBy (Sepa, dépparos, 76, skin, xvBy, 7, head) Derma- tocybe Twpat-o-deppis (TOua, rwpatos, TO, drink, deppis, 7, leather coat) Pomatoderris Oavpar-o-rrepis (Outpa, Oariparos, TO, wonder, Trepis, N; fern ) Thaumatopteris oTEppat-o-xvoos (ao7épa, orépuatos, To, seed, yvoos, 6, foam) Spermatochnous oTopat-o-OyK.ov (oTopa, TTOuaTOs, TO, MOUth, AyAKx0v, 7d, little box) Stomatothecium : owpat-(0)-ayyeov (cGua, swopatos, To, body, ayyeiov, To, vessel ) Somatangium (4) Stem in -es, genitive -eos (—eoos), nominative usually in -os, mostly neuters. The form for composition is Obtained . by dropping -eos of the genitive, or -os of the nominative. 352 Greek and Latin in Biological Nomenclature 33 xep0-(0)-ovpa (xépdos, Kepdeos, 70, trick, ovpa, 7, tail) Cerdura 4 BeX-o-repovyn (BeXos, BéXeos, 76, dart, repovn, 7, point) Beloperone 7:1 in -en dyy-0-fopa (ayyos, ayyeos, TO, vase, opd, a carrying) Ango- phora_ 1 xEtA-o-oKupos (xelos, xetAeos, TO, lip, oxvdos, 76, cup) Chilo- scyphus. 2 av0-0-fuxos (avOos, avYeos, 76, flower, Pdxos, To, seaweed) Antho- phycus 46:2 pux-o-hutov (piKos, PvKeos, To, seaweed, urov, ro, plant) Phyco- oli ae Ook ae oan B. Vowel stems (in -c or -v). (1) Stem in -t, nominative in -1s, masc. and fem., -1, neut. _ The stem is obtained by dropping -os of the genitive, or -s of masc. or fem. nominative. 6yi-(0)-avOos (dys, dYos, dYews, 7 look, avOos, ro, flower) Opsi- anthus Ggi-o-Kapvoy (ddis, Opis, opews, 6, Snake, xdpvov, ro, nut) Ophiocaryum 20 merept-o-pvdrdrov (rérept, memépios, TO, pepper, PvAAov, Td, leaf) Peperiophyllum 1 in -o TpoTt-o-Aemis (Tpoms, Tpdm.os, 7, keel, Aeris, 7, scale) Tropiolepis gpvot-o-yoxis (pious, piovos, Piaews, 7, Nature, yAwyxis, 7, point) Physioglochis 14 (2) Stem in -v, nominative in -vs, (-avs, -evs, -ovs) masc. and fem., -v, neut. The stem is obtained by dropping -os of the genitive, or -s of the nominative. . dpv-o-rrepis (Spits, Spvds, 7, oak, rrepis, 7, fern) Dryopteris 11:4 in -v pv-(0)-ovpa (pis, pds, 6, mouse, otpa, y, tail) Myura 18:2 in puo- mitv-(0)-oxrs (ritus, rirvos, 7, pine tree, dys, 7, look) Pityopsis otaxv-0-Botpus (aTaxvs, otaxvos, 6, spike, Bdrpus, 6, bunch of grapes) Stachyobotrys 10:7 in -vy; I in -v Here are usually placed Bots, ypats, vats. These originally had the stem in a consonant, digamma, as Bog-, but the digamma was lost, leaving third declension stems in -o and 353 34 Frederic E, Clements -a. The proper stem is obtained by dropping -os of the genitive, though these stems are quite irregular in the mat- ter of composition. Bo-0-yAnvos (Bovs, Boos, .6, Ox, yAnvyn, 9, eyeball) ox-eyed 17: 96 in Bov-; 3 in Bo- ypa-o-Aoyia (ypats, ypads, 7, old woman, Aoyia, 7, speech) gossip 6 Anomalous nouns. "= ex) eee o0=0 nv = eu €=7Z y=ps =u infinal-os, o=oe 6=th yk =nc v=y [ov ov=u Kae yx =nch 0 =O w= yl aa yy =ng a= B=mM p—sEh; pp =1rh y=e v= ae ® =O =m in final -ov o_o el o,s=sS Medial ‘ (%) arising from word-formation is to be transliterated, thus preventing elision of a preceding vowel, unless its presence is already shown by aspirating the preceding consonant, as in épyjpepa. Latin usage is variable in this particular, since words already compounded in Greek, in which the aspirate was not visible, were transliterated into Latin as they stood, while in other words in which the presence of the aspirate was felt or known, the latter was transliterated. In scientific words it is important that the rough breathing be rendered by 4, not only in order that the terms of a compound may be readily recog- 378 Greek and Latin in Biological Nomenclature 59 ' nized, but also to avoid the possible confusion of two compounds otherwise exactly alike.t The following list will serve to illustrate the more frequent errors in transliteration, and their correction. Adenocaulon = Adenocaulum (or, much better, Adenocaulus, the last term being xavAds, 0, stalk); Lachnocaulum Agropyron = Agropyrum (vpds, 0, wheat) Aerophyton = Aerophytum (¢vrov, 76, plant); Petrophytum -Ampelodesmos = Ampelodesmus (éeopcs, 0, band) Amphicarpon=Amphicarpum (better, Amphicarpus from Kap7cs, 0, fruit) ! Acrospeira = Acrospira (o7eipa, 7, knot, coil) Amorpha = Amorphe (pop¢y, 7, form) Arachnion = Arachnium (dpaywov, to, spider’s web) Apios = Apius (amos, 7, pear, pear-tree) Aplopappus = Haplopappus (dzAcos, simple) Arctostaphylos = Arctostaphylus (better, Arctostaphyle from otapvAy, 7, bunch of grapes) Astrebla = Astreble (o7péBAn, 4, roller, orpeBdos, 7, ov, twisted ) Batodendron = Batodendrum (éévépov, ro, tree) ; Linodendrum, Phorodendrum, Rhododendrum, Toxicodendrum Blepharoneuron = Blepharoneurum (vedpor, ro, fibre, nerve) Brachychaeta = Brachychaete (airy, 7, hair) Cal.irhoe = Callirrhoe; Glycyrrhiza, Coralliorrhiza, ete. Chaetochloa = Chaetochloe (xA¢cy, 7, grass); Echinochloe, Erio- chloe, Helochloe, Leptochloe, Scolochloe Chamaecladon = Chamaecladus (KAdéos, 0, shoot) ~ Chamaenerion = Chamaenerium (vypuov, to, oleander) Chamaerhodos = Chamaerhodus (better Chamaerhodum, pcdov, To, rose) Chionyphe = Chionohyphe (4%, 7, web) Cheiranthes = Chiranthus (yep, 7, hand); Chiromyces Coilomyces = Coelomyces (xotXos, hollow ) Winné. Critica Botanica, 129. 1737. Dall, W. H. Nomenclature in Zoology and Botany, 55. 1877. Kuntze, Otto. Revisio Generum Plantarum, 3:354. 1893. Miller, Walter. Scientific Names of Latin and Greek Derivation. Proc. Cal. Acad. Sci., 1:127. 1897. BES. 60 Frederic E. Clements Cyperus = Cypirus (xvreipos, c, marsh plant) Corypha = Coryphe (xopydy, 4, head, top) Dasylirion = Dasylirium (Aeépuoy, 76, lily) Diospyros = Diopyrus (vpés, 6, wheat) Dolichos = Dolichus (d0Arx6s, long) Eleocharis = Helocharis (€Aos, ré, marsh) Elodea = Helodes (€A#édys, marshy ) Gyrotheca=Gyrothece (6yxn, 4, box); Heterothece, Tetra- gonothece Haplymenium — Haplohymenium (ipenov, 76, little membeane) Helicoon = Helicoum (ov, 70, egg) Hemicarpha= Hemicarphe (xéo¢y, 4, scale, better Hemicar- phus, xapdos, 70, scale) Hydrocleis = Hydroclis («Aes, 7, hook, key) Hydrodictyon = Hydrodictyum (éé«rvov, 76, net) Korycarpus = Corythocarpus (xépus, KépvOos, 7, helmet) Lecanidion = Lecanidium (Aexavidiov, 76, dish, pan) Lycopersicon = Lycopersicum (cepotxev, To, peach) Metroxylon = Metroxylum (€vAov, 76, wood); Stereoxylum Microthyrion = Microthyrium (6vpior, 76, little door) Opegrapha = Opographe (ypady, 7, drawing) Orophaca = Orophace (¢axy, 7, lentil) Potamageton = Potamogiton (yzirwv, 6, neighbor) Protalos = Protohalus (aXs, ddos, 7, sea) Prinos = Prinus (zpivos, #, evergreen oak ) Rhodospatha = Rhodospathe (ord6n, 7, broad blade) Sicyos = Sicyus (a/kvos, 6, common cucumber ) Spirodela = Spirodele (djAos, visible) — Steirochaete = Stirochaete (oreipa, 7, beam of a keel); Stironema Stenospermation = Stenospermatium (ozepparvor, ro, little seed) Symphoricarpos = Symphoricarpus (xapzos, 6, fruit) Symplocos = Symplocus (cvpzrAoxos, entwined) Syndesmion = Syndesmium (better, Syndesmus, ovvderpos, 6. band) 380 Greek and Latin in Biological Nomenclature 61 IV Of two or more similar terms, the earliest alone is valid, unless they show an essential difference in root, suffix, or prefix; differences of spelling. gen- der, or alternative termination are insufficient. Retroactively, the earliest name, if not already in the proper form, is to be corrected, while all others fall- “Nomina generica, simili sono exeuntia, ansam praebent con- fusionis.”” Critica Botanica 43. “Nomina generica ex aliis nominibus genericis, cum syllaba quadam in fine addita, conflata, non placent.” Ibid. 38. Similar generic names have long constituted a grave source of confusion in biology. Nearly every writer upon botanical nomenclature has appreciated this fact, and has suggested some method of obviating the difficulty. Linné* pointed out clearly the way by which all such duplicates and apparent duplicates might be avoided, but in the subsequent rapid development of taxonomy his precepts were lost sight of or ignored. The Paris Code, though silent on this matter, unintentionally aggravated the situation by the unfortunate reservations of Article 66. In passing, it should be noted how signally the purpose of this scholarly article has been defeated by the presence of an unim- portant exception. The provision that “every botanist is au- thorized to rectify the faulty names or terminations, unless it be a question of a very ancient name current under its incorrect form,’ obviously made exception only for names given by Aris- totle, Theophrastus, Dioscorides, Pliny, and other Greek and Roman writers upon plants. But this exception has since been persistently misunderstood, or purposely extended to cover any incorrect name of any degree of currency whatsoever, and has finally found expression in the absurd dictum that “the original form of a name is to be retained no matter how incorrect it may be.” This feeling seems to have had some influence upon the treatment of similar generic names in the Berlin Rules and in the Rochester Code. Though the statement of the rule is dif- ferent, the treatment is practically identical in both. According 1Critica Botanica, 39, 43. 62 Frederic E. Clements to the former*, “similar names are to be conserved, if they differ ever so little in the last syllable; if they only differ in the mode of spelling, the newer one must fall.” Also, “there are to be conserved Adenia as well as Adenium, Apios as well as Apium, Chloris as well as Chlorea and Chlora, Danae as well as Danais, Hydrothrix as well as Hydrotriche, Silvaea as well as Silvia, ete. ; we doubt that there is any scholar who will confound them. On the contrary, Tetraclis and Tetracleis, Oxythece and Oxytheca, Epidendrum and Epidendron, Oxycoccus and Oxycoccos, Aster- ocarpus and Astrocarpus, Peltostema and Peltistema are only different modes of spelling the same word, and the newer one is to be rejected if they name different genera.” The Rochester Rules’ provide that “Similar generic names are not to be re- jected on account of slight differences, except in the spelling of the same word; for example, Apios and Apium are to be re- tained, but of Epidendrum and Epidendron, Asterocarpus and Astrocarpus, the latter is to be rejected.” In both codes, it will be noticed that similar names are to be rejected only when the difference is merely one of transliteration of the ending, or, very rarely, of connective. A difference of gender termination or of alternative ending is considered sufficient to warrant retention, even though this difference results from incorrect formation, as in Hydrotriche. Both rules are equally far from any classical warrant, and, in consequence, neither code can furnish a logical or accurate basis for the treatment of similar terms. In formulating a rule for these, however, it is impossible to give serious consideration to the views of mere logophiles, who would make wholesale rejec- tions on the basis of slight or fancied similarities. Thus, it has been suggested that Micranthus and Micranthemum are so simi- lar as to warrant the rejection of one, while of Macranthe and Megalanthe, Glycyphila and Glycyphylla, one should be rejected because the first two are practically identical in meaning, and the last two in pronunciation!’ Between the two extremes there 1Vorschlage zur Erganzung der “‘ Lois de la Nomenclature Botanique.’’ Berlin, 1892. 2 Bull. Torr. Bot. Club, 19:290. 1892. 382 Greek and Latin in Biological Nomenclature 63 -is but one logical position, namely, similar terms are identical in nomenclature when as Greek and Latin words they exhibit no essential differences. Thus, Cerastium, Ceratium, and Ceratia are merely different forms of a Greek word «xepdéruov, and are -homonyms, while Lecane, Lecanarium, and Lecanidium are dif- ferent words, the last two being formed upon the first by the use of suffixes. Frequent affixation of the same stem should be carefully avoided, however, regardless of the validity of the re- sulting derivatives. Such Greek words as dvOos, ypadus, xepady, and their relatives, which are extremely frequent in nomenclature, will serve very well to show the difference between homonyms and similar yet valid terms. Besides many compounds, the lexicon shows twenty deriv- atives of the root av6-: of these, the following seventeen are suffi- ciently distinct to justify their use: avOos, 70, flower; av@7An, 7, pan- icle; avOewov, ro, flower; avOéucov, To, floweret; avOiov, ro, floweret; av@épixos, 6, flower of asphodel; dvOedv (avOdv), 6, flower bed; avOceuoons, flower-like; avOeuces, flowery; dvOeuwros, adorned with flowers; avOndov, n, bee; avOypcs; flowery; avOnpcrns,y, bloom; avOyors, 7», full bloom; dvOyriKes, blossoming; avvos, blooming; avOocvvy, 7, bloom. Av6y, 7, full bloom, flower, should be avoided in composi- tion, since it is identical with a@v6o0s when used as a first term, and is confusing asa last term; dv@eor is identical with av6/ov, and dvGenis too near avOenov to be fortunate. The root ypad- shows two series of de- rivatives, one based upon the root, and the other upon the stem Ypappat-. Of the latter, ypapypa., ro, letter, picture, ypapparetov, TO, document, ypappare(diov, ro, small tablets, ypapparevs, o, scribe, and ypapparixy, 1, written character, are different, while ypappy, 7, stroke, line, is to be regarded as a mere alternative of ypaypa. In the first series, ypady, 7, drawing, ypadetov, ro, pencil, ypadetduov, 70, pencil, and ypaduxcs, graphic, are distinct, but ypadés, 7, stylus, and ypaos, to, letter, are alternatives. Of the derivatives of xepaA7, 1, head, xedadis, n, little head, should be avoided, but the following are distinct; xedadwov, 70, little head; xedadcdvov, 70, little head; keharivn, 7, head of the tongue; xeadrkos, of the head; xepddauos, of the head; xepadauddys, chief. Nomenclature would, however, become very much involved for anyone but the philologist, if all 383 64 Frederic E. Clements the proper derivatives of such roots as the above were to finda place in it. Such a condition can be readily avoided if proposers ~ of terms will take the trouble to acquire a Greek vocabulary. I. Homonyms. These arise from alternative forms of the same root or stem, from mere differences of spelling, translitera- tion, gender or alternative ending, or from differences produced by erroneous connectives or terminations. Aceras Pers. 1807 Acerates Elliott 1817 Aceratium DC. 1824 Aceratia F. Mull. 1854 Acetabulum Tourn. 1700 Acetabula Fries 1822 Acetabularia Lamx. 1816 Acetabularium Endl. 1836 Achlys DC. 1821 Achlya Nees 1823 Adenia Forsk. 1775 Adenium Roem. & Schult. 1819 Adenogyne Klotzsch 1841 Adenogynum Rchb. & Zoll. 1850 Adenophorus Desvaux 1808 Adenophora Fisch. 1823 Apios Boerh. 1720 Apium Hoffm. 1814 Calanthe R. Br. 1821 (mel. Ca- lanthus ) Kalosanthes Haworth 1821 Calanthea DC. 1824 Calosanthes Blume 1826 Callitriche L. 1751 (cor. Calli- thrix ) Calythrix R. Br. 1819 Calothrix Ag. 1824 Cerdstium L. 1737. 7 (cotuaaes ratium ) Ceratia Adans. 1763 Ceratium Alb. & Schwein. 1805 Chamaedrys Tourn. 1700 Chamaedryon Seringe 1825 Chamaemelum Tourn. 1700 Chamaemeles Lindl. 1822 Chamaemela DC. 1837 Chlora Adans. 1763 Chloraea Lindl. 1826 Chlorea Nyland. 1854 Coleosanthus Cassini 1817 (Cor. Coleanthus ) ~Coleanthus Seidl 1817 Dasanthera Raf. 1819 (cor. Dasyanthera ) Dasianthera Presl 1831 Dermatocarpon Eschw. 1824 (cor. Dermatocarpum ) Dermatocarpus Miers 1852 Dermocarpa Crouan 1856 Desmanthus Willd. 1805 Desmosanthes: Blume 1825 Dicera Forst. 1776 (cor. Di- ceras ) 384 Greek and Latin in Biological Nomenclature Diceras Endl. 1840 Dictyanthes Rchb. 1837 (mel. Dictyanthus ) Dictyanthus Decaisne 1844 Drimys Forst. 1776 Drimia Jacq. 1786 Epiphegus Spreng. 1820 Epiphagus Rylands 1843 Eremanthis Cassini 1827 (cor. Eremanthus ) Eremanthus Lessing 1829 Eremanthe Spach 1836 Erythranthus Hanstein 1853 Erythranthe Bailion 1858 Eurotia Adans. 1763 Eurotium Link r809 Gamochilum Walpers 1839 Gamochilus Lestib. 1841 Glyciphylla Raf. .1819 (cor. Glycyphylla) Glycyphylia Steven 1834 Glyphia Cassini 1818 Glyphaea Hook. f. 1846 Gonatobotrys Corda 1839 Gonatobotryum Sacc. 1879 Gonyanthes Blume 1823 (cor. Gonatanthus ) Gonatanthus Klotzsch 1840 Grammocarpus Ser. 1825 (cor. Grammatocarpus ) Grammatocarpus Pres! 1831 Heterocladia Decaisne 1841 Heterocladium Schimp. 1852 Hippobroma G. Don 1834 Hippobromus Eck. & Zeyh. 1836 Holophyllum Lessing 1830 e 65 Holophylla G. Don 1837 Isomerium R. Br. 1830 Isomeria Pres] 1837 Lecanium Presl 1843 Lecania Massalongo 1853 Lepidocarpus Adans. 1763 Lepidocarpa Blume 1855 Lepidotis Palis. 1805 (cor. Lepidotus ) Lepidota Sterb. 1820 Lepidotus Fries 1836 Lepidotig Rchb. 1841 Lepidotum Dunal 1852 Lepisanthus Blume 1825 (cor. Lepidanthus ) Lepidanthus Nees 1830 Macranthus Poir. 1813 Macranthea Boiss. 1840 Macrantha Bunge 1843 Macropodium R. Br. 1812 Macropodia Fuckel 1869 Marainophyllum Pohl. 1825 (cor. Marantophyllum ) Marantophyllum Miquel 1855 Megalanthe Gaudin 1828 (mel. Megalanthus ) Megasanthus G. Don 1834 Microglossa DC. 1836 Microglossum Sacc. 1884 Microtea Swartz 1788 (mel. Microtes ) Microtis R. Br. 1810 Monochila G. Don 1834 (cor. Monochilus ) Monochilus Fisch. & Meyer 1835 66 Frederic E. Clements _Olhiganthes Cassini 1817 (cor. Oliganthus ) Oliganthos Barneoud 1845 Oligotrichum DC. 1805 Oligothrix DC. 1837 Pachypleurum Rchb. 1832 Pachypleuria Presl 1836 Pachypleura Jamb. & Spach 1842 Petrophile Knight & Salisb. 1809 Petrophila R. Br. 1800 Rytiphlaea Ag. 1817 _Rhytidophloeus ) Rhytidofloyos Corda 1845 Salpianthus Humb: & Bonp. 1808 (cor. Salpinganthus ) Salpixantha Hook. 1845 Schismus Palis. 1812 (mel. Schisma) Schisma DuMort. 1822 Schizanthus Ruiz & Pav. 1794 Schisanthus Haworth 1819 Schistanthe Kunze 1841 (cor. Sphaerophorus Pers. 1794 Sphaerophora Blume 1850 Sphaeroplea Ag. 1824 (cor. Sphaeropleum) Sphaeropleum Link 1826 Stilbe Berg. 1767 Stilbum Tode 1790 Tapinanthus Blume 1824 Tapeinanthus Herbert 1837 Tetrandra A. DC. 1845 (cor. Tetraner ) Tessarandra Lindl. 1847 Thrixspermum ‘Lour. 1790 (cor. Trichosperma) Trichospermum Blume 1825 Trachysperma Raf. 1809 Trachyspermum Link 1821 - Trichopteris Necker 1790 Trichipteris Pres] 1822 Trichosanthus Lo 1727. Gears _ Trichanthus) Trichantha Hook. 1844 Trichanthus Philippi 1857 Xanthoglossa DC. 1837 Xanthoglossum Lindl. 1852 II. Terms classically different, but so similar in form as to be unfortunate. There is not sufficient warrant for the re- jection of these, but their formation is to be avoided, if not, indeed, invalidated, for the future. Acarphaea Harvey & Gr. 1849 Acarpha Griseb. 1856 Chlora Adans. 1763 Chloris Swartz 1788 Danae Medic. 1787 Danais Vent. 1799 Galax L. 1753 Galactia P. Br. 1756 Galaxia Thunb. 1782 Gliocladium Corda 1840 Gloeocladia, J. Ag. 1842 (cor. Gloeocladium ) Glyphis Achar. 1814 Glyphia Cassini 1818 Hydrophila Ehrhart 1780 (cor. Hydrophile) Greek and Latin in Biological Nomenclature 67 Philydrum Gaertn. 1788 (mel. Philohydrum ) Isomerium R. Br. 1830 Isomeris Torr. & Gr. 1838 Ixianthes Benth. 1836 (mel. Ixianthus ) Ixanthus Griseb. 1839 Lepanthes Swartz 1799 (mel. Lepanthus ) Lepisanthus Blume 1825 (cor. Lepidanthus ) Rhaphidospora Nees 1832 Rhaphiospora Korb. 1855 Syncephalum DC. 1837 Syncephalis Van Tieghem 1875 Theriophonum Blume 1835 Therofon Raf. 1836 (cor. Therophonun ) Xanthiopsis DC. 1836 Xanthopsis DC. 1837 III. Similar terms distinct classically and nomenclaturally Actinostemon Klotzsch 1841 Actinostemma Lindl. 1847 Alectra Thunb. 1784 Alectryon Gaertn. 1788 _ Brachylobos DC. 1821 Brachylobus ) BrachylobiumC.A. Meyer 1841 Calopogon R. Br. 1813 Calopogonium Desvaux 1826 (cor. Ceramianthemum Donati 1750 Ceramianthe Rchb. 1831 (mel. Ceramianthus ) Cladodes Lour, 1790 Cladodium Bridel 1826 Diceratium Lagasca 1815 Diceras Endl. 1840 Eritrichium Gaudin 1828 Eriothrix Rchb. 1828 Glechoma L. 1737 (cor. Gle- chonoma ) Glechon Spreng. 1827 Haplocarpha Lessing 1831 Haplocarpaea Endl. 1838 Micranthus Wendland 1798 Micranthemum Michx. 1803 Stylidium Swartz 1807 Stylis Poir. 1817 Trachypodium Leman 1828 Trachypus Reinw. & Hornsch. 1829 V Terms are invalid unless properly spelled; retroactively, improper spellings are to be corrected. ; Apart from its application to improper formations, this thesis is of secondary importance. It is given place here merely to em- phasize again the fact that nomenclature in all its aspects must rest upon a classical basis, a repetition rendered imperative for the reason that many biologists and more than one code still re- 67 68 I'rederic E. Clements gard the Latin of Linné as the model. In Greek, a large num- - ber of incorrect spellings have arisen from the careless practice of dropping one or more letters at the end of a word, or from the arbitrary change of the termination. The names of The- ophrastus and Dioscorides, especially, have suffered mutilation, and should be restored to the original form, while the correction of later misspellings should be made upon the basis of the classi- cal form of the terms of the compound. In the rare cases in which the spelling of a Greek word has been changed in Latin, the Greek form should prevail. VI Terms are invalid If they exceed six syllables in length; retroactively, the correction of sesquipedalian words must never take place by contraction or mutilation. “Nomina Generica Sesquipedalia, enunciatu difficilia, vel naus- eosa, fugienda sunt.” Critica Botanica 133. The practice of biologists with respect to the formation of ex- tremely long terms has been so exemplary that the present rule scarcely requires postulation. Its justification may be found in the fact that inconveniently long words, more or less frequent a century ago, still appear occasionally, and that such words, if there were no definite sentiment or legislation against them, might again become frequent as the supply of primitives and short compounds becomes exhausted. It is more or less unsatis- factory to limit the length of a word by the number of syllables, ‘since these vary greatly in length in different stems, but this is undoubtedly better than limitation by the number of letters. — It is a question whether nomenclature would not gain more than it would lose, if the maximum length of words were placed at five syllables, though the number of changes necessitated would prob- ably render such a rule inacceptable. Naturally, the present rule should not be made operative in the case of names of groups above the genus. 68 Greek and Latin in Biological Nomenclature 6g VIE Hybrid terms are invalid: retroactively, Greek-Latin hybrids are to be cor- rected upon the basis of the Greek element, but all vernacular and personal hybrids fall. “Nomina generica ex vocabulo graeco & latino, similibusque, Hybrida, non agnoscenda sunt.” Critica Botanica 28. “Everyone is bound to reject a name in the following cases: (4) When it is formed by the combination of two lan- guages.”’ Paris Code, Article 60. 1867. “The possibilities of the field he has opened up for us are in- deed great, witness: Smithia, Smithago, Johnsmithotoma, Ig- smithia, Smithalga, Smithodendron. | dwell on this because it seems to me that botanical Latin is impure enough already with- out such gratuitous monstrosities.’ Pound. American Natur- alist, 26:147. 1892. “An unhappy feature of Dr. Kuntze’s work, and one in vindi- cation of which I can say nothing, is his method of constructing new names for genera. Perhaps in some distant century, when self-repeating history may have brought the return of times when scientists were mostly men of clear ethics, solid learnine, and re- fined tastes, some such reform in plant nomenclature as that which M. Saint-Lager in these times vainly advocates will be carried into effect. If, before the advent of that good time, Dr. Kuntze’s Radlkofertoma and Schweinfurthafra shall have become current for certain genera, they will be the first to be rejected.” Greene. Pittonia, 2:277. 1892. The indifference of many biologists to a classical standard for nomenclature reaches its logical culmination in the formation of hybrid words. Botanists especially are practically unanimous in condemning hybrids, but, in spite of this fact, carelessness and ignorance are steadily increasing the number of illegitimate words. It is unnecessary to prove that hybrids are as unfortu- nate in nomenclature as in philology, but it is necessary that particular attention be given to them in order that they may be avoided, or at least corrected. No biologist of any real attain- 69 70 T'rederic E. Clements ment can afford to stand sponsor for a hybrid name, when a trifling expenditure of time will yield a word of pure birth. For the sake of clearness, hybrids may be divided into two classes: (1) Greek-Latin hybrids, in which one element is Greek and the other Latin; (2) vernacular hybrids, in which one ele- ment is from a modern tongue, while the other is classical, usu- ally Latin. Each class shows hybrids in which both terms are independent words, and those in which one term is an affix. There is no essential difference between these as hybrids, but the distinction is an important one, because words of the second © group are rarely recognized as hybrids on account of the slight familiarity of biologists with classical methods of derivation. The matter presents indeed some difficulty for the philologist, because of the similarity of cognate affixes in Greek and Latin, and because of Greek affixes borrowed by Latin. On account of the difficulty of detecting them, hybrids of this sort are becom- ing more and more common. ‘The raising of hybrid sectional names in ev-, Wevdso-, -wdys, -ella, -astrum, etc., to the rank of generic names is contributing very largely to this result, as also the endeavor to honor a biologist by attaching all the Latin suf- fixes in turn to his name. There has been considerable discussion regarding the treat- ment of such hybrids as pseudorepens and Eucarex. The con- tention is made that these words are not hybrids, since these affixes were regularly used by Latin writers, but, as a matter of fact, they are not found in classical Latin outside of borrowed Greek words in which they are a proper affix. It has further been urged that such words are scarcely hybrids, for the reason that pseudorepens does not mean “false creeping,’ but merely refers to a species of Agropyrum, which is not A. repens. Such argument is mere sophistry, since every compound or derivative which contains a Greek and Latin element, whether independent word or affix, is a hybrid. The only possible exception is found in those rare Greek words which have become so completely domiciled in Latin that their origin is no longer felt. The correction of hybrids* is possible only when the word 1Since the above was written, three instances of a similar correction of 7O Sra Greek and Latin in Biological Nomenclature 71 arises from the combination of Greek and Latin forms, in which case the cognate or corresponding Greek form is used to replace the Latin element. In the case of vernacular hybrids, such sub- stitution is so rarely possible that it may be entirely disregarded, and all vernacular hybrids, the majority of which are personals, are to be summarily rejected. Such names fall not only be- cause they are hybrids, but also on account of the operation of rules I and VIII. Greek-Latin hybrids, which are current in nomenclature, are to be corrected and followed by double cita- tion of author and reviser, but hybrids proposed in the future, being invalid under the present rule, may be corrected or ig- nored at the will of the reviser, who alone is to be cited for the new name in either event. The following list will illustrate the various kinds of hybrids, as well as the method of correction, when this is possible. I. Greek-Latin hybrids in which both terms are independent words. Actiniceps — Actinocybe (dxris, dxrivos, 7, ray, xv¥By, 7, head) Aureobasidium = Chrysobasidium (xpvceos, golden, Bacidvov, 70,. pedicel) Baculospora — Bactrospora (Baxrpov, 70, staff, oopd, 7, seed) Botrypes = Botryopus (Bcrpvs, 6, cluster of grapes, zovs, odes, 6, foot) Callosisperma = Sclematosperma (o«Ajpa, oKdjpartos, 70, hard- ness, ovéppa, To, seed) _ Claudopus = Loxopus (Aoéés, slanting, crooked, zovs, 6, foot) Clavogaster — Rhopalogaster (forador, 76, club, yaoryp, 7, belly ) Clypeosphaeria = Peltosphaeria (7éAry, 7, small shield, o¢uipa, 9, ball) Fagopyrum = Phegopyrus (¢yyos, 9, oak, updos, 6, wheat) Fimbristylis = Lomatostylis (AGpa, Adparos, 70, fringe, oTvAis, 7), pillar) ~ Fusicolla = Chytocolla, (xur¢s, poured out, KéAda, 7, glue) Geminispora = Dissospora (8oaos, double, oop, 7, seed) hybrids have been found in Pfeiffer’s Nomenclator Botanicus 1:624, 1050, 1640. Catasetum Kunth is corrected to Catachaetum, Diastemella Oersted to Diastemation, and Loroglossum Rich. to Himantoglossum. 71 “iI i) Frederic E. Clements Gorgoniceps = Gorgocybe (Topy, cos, 7, the Gorgon, «vy, 7, head) Hemicarex = Hemidonax (jmu-, half, dovaé, 6, reed) Massospora = Mazospora (paa, 7, barley cake, ozopa, 9, seed) Muciporus = Myxoporus (pa, 7, mucus, zpos, 6, pore) - Nemacola= Nemocolus (véuos, 76, wooded pasture, -Kodds, dwelling) Nitophyllum = Phaedrophyllum (qa:dpes, bright, vAAov, 76, leaf) Nothofagus = Nothophegus (vé6os, spurious, ¢yyes, 7, oak) Nucleophagus = Caryophagus (xdpvov, to, nut, Payos, eating) Onychosepalum = Onychocalyx (évvé, ovuxos, 6, claw, KaAvé, 7, cup of a flower) Pachyfissidens = Pachyschizodon (zaxvs, thick, oyilodav, 64, split tooth ) Peltigera = Peltophora (7éArn, 9, shield, popd, n, a carrying) Phaioclavulina = Phaeocoryne (aus, dusky, xopvvy, 7, club) Pseudopeziza = Pseudopezis (wevdys, false, weéis, », stalkless fungus ) Radulotypus = Psectrotypus (~y«rpa, , scraper, tUros, 6, form) Retiporus = Dictyoporus (dikrvov, 76, net, mépos, 6, pore) Scirpodendron = Donacodendrum (éddvag, ddovaxos, 6, reed, devopov, To, tree) Septosporium = Schizosporium (oxi@a, 9, cleft, omcpiov, To» spore ) Verticicladium = Helicocladium (€Ac€, €Ackos, r, whirl, xAadiov, 70, - branch) If. Greek-Latin hybrids in which one term is an affix, Anthostomella = Anthostomatium Bisporella = Disporyllium Brizula = Brizyllium Chlorosa = Chlorotes Coryneliella = Corynisce Cyphella = Cypharium Dolicholus = Dolichidium Eucaprifolium = Euaegophyllum (at, aiyos, 6, 7, goat, PvAAov, 0, leaf) 72 ra ) — Greek and Latin in Biological Nomenclature 73 Eucarduus = Euacantha (dkav@a, 7, thistle) Fusidium = Atractidium (drpaxros, 6, spindle) Gaurella = Gauryllium Glossula = Glossidium Graphiola = Micrographium (cfr. Graphis, Graphium, Graph- idium, Graphyllium) Hormiactella = Hormiactinium (épmia, 7, fishline, axris, axrivos, 7, tay) Hypocrella — Hypocreatium Ilicioides = Dryodes (épis, dpvos, 7, oak) Juncodes = Thryodes (@pvov, ro, rush) Labridium = Chilidium (yeiAos, 70, lip) Lachnella = Lachnium Lachnellula — Microlachnium Lithophragmella = Lithophragmatium Lophiola = Microlophium Myriactula = Myriactinium Nasturtioides = Napyodes (varv, ro, mustard) Phaeodiscula = Phaeodiscium Pholiotella= Pholidotium (odAidwros, clad with horny scales) Polystomella — Polystomatium Pterula = Pteridium Rhodiola = Rhodarium Sphaerosporula = Sphaerosporyllium Stigmatella = Stigmatium Struthiola = Struthidium Tiarella = Tiaryllium Trichopeltulum = Trichopeltium Typhula = Typhidium Zomicarpella = Zomatocarpium (f@pya, Copartos, ro, girded doublet) III. Vernacular-classic hybrids in which one term is a personal name. Hybrids of this class lack even the excuse of ignor- ance. Nomenclature can show but one greater mon- strosity, namely, the mutilated vernacular compound. Such personals can not be corrected and must fall ir- 73 74. Frederic E,. Clements revocably. Kuntze has been censured unjustly as the originator of the personal hybrid, since the latter was already found in numerous examples, as he himself has shown. But he deserves to be severely censured for greatly extending its use. Such atrocities as Pseu- doleskia, Microschwenkia, Gerrardanthus, Pringleophy- tum, etc., were in existence before the Revisio, but they are lost sight of in the deluge of such foundlings pro- posed in the latter. The magnitude of Kuntze’s offense against a classical nomenclature may be seen from the fact that out of Iog generic names proposed by him, 67 are personal hybrids, and the remainder are almost entirely mutilations, such as Watsonamra, Clarkeinda, Schweinfurthafra, Itoasia, etc. In the first volume of the Revisio,* the author gives a variety of methods by which the same botanist may be “honored” ad nauseam without increasing homonymy. The whole treatment manifests not only an entire absence of linguistic taste, but also an abiding ignorance of classical philology. Kuntze elsewhere*® says apologetically, “Ich bin im Griechischen wenig erfahren.” It is to be regretted that this feeling did not restrain him from such mon- strous treatment of classical stems. The following lists, though by no means complete, will serve to illustrate the various kinds of vernacular hybrids, all of which are to be rejected. 1. Vernacular-Greek hybrids—personals. Bakeropteris Beccariodendron Balfourodendrum Beckeropsis Barleriacanthus Benthamidium Barlerianthus Blumeodendrum Barleriopsis Buforrestia Barleriosiphon Caloknightia Barleriotes Chamaesaracha Beccarianthus Chamissomneia 1Kuntze, O. Reviso Generum Plantarum, 1:51. 1891. 2Tbid., 3:214. 1893. 74 Greek and Latin in Biological Nomenclature 75 Christiastrum Cordierites Cyphokentia Diserneston Doellochloa Doniophytum Dyerophytum Ellisiophyllum Englerophoenix Epizeimeria Epibrissonia Eugeniastrum Eugenioides Fritschiantha Gayophytum Gerrardanthus Glaziostelma Grayemma Hackelochloa Halterophora Harveyastrum Henningsocarpur Henningsomyces Huegeliroea Kentiopsis Kuhniastera Kuntzeomyces Leioclusia Lenzites Ludwigiantha Lyonothamnus Macounastrum Macowanites Mannoglottis Marilaunidium Melioschinzia _Microkentia 75 Microschwenkia Microweissia Montagnites Neobrunia Neocontarinia Neograyia Neohuttonia Neopeckia Neoskofitzia Neowashingtonia Nesogordonia Oliverodoxa Orchidofunkia Osbeckiastrum Palaeogrewia Parabesleria Parabouchetia Parapottsia Phaenohoffmannia Pleomassaria Porteranthus Preussiaster Pringleophytum Prockiopsis Protohopea Protoventuria Pseudehretia Pseudobarleria Pseudogunnera Pucciniopsis Pycnoseynesia Radlkofertoma Rhabdoweissia Roeperocharis Sarcolippia Schmitzomia Schroeteriaster 76 Sibbaldiopsis Silvianthus Siphoneugenia Smithantha Stahlianthus Stanhopeastrum Sternhopeastrum Thalianthus Absolmsia Agardhina Algogrunowia Algorichtera Arcangelina Balfourina Bartramidula Baumanniella Beccariella Benthamistella Berkeleyna Bisboeckelera Brocchinia Caruelina Cohnidonum Cookeina Crepinula Delpinoina Detonina Dillwynella Drudeola , Eremicella Errerana Fabreola Flueckigera Forsteronia Freynella Friesula Frederic E. Clements . Vernacular-Latin hybrids—personals. 76 Thileodoxa Thouanidium Tulasnodea Urbanodendrum Uroskinnera Weinmannodora Wittmackanthus Zieridium Fuckelina Gerrardina Gibberinula Greeneina Grisebachiella Harziella Hemsleyna Hendersonula Hodgsoniola Hofmeisterella Hookerina Hostana Jacksonago Julella Karstenula Kickxella Koehneago Knyaria Kuetzingina Latzinaea Magnusina Massariella Massarina Montagnula Mohlana Mortierella Munkiella Neilrichina EY, Greek and Latin in Biological Nomenclature Nicholsoniella Nylanderaria Nymanina Octavianina Oliveriana Oudemansiella Patouillardiella Peckiella Peckifungus Penzigina Peyritschiella Pfeifferago Phillipimalva Pringsheimina Richterago Saccardinula Saccardoella Scopulina Stephanina Thozetella Triumfettaria Urbanisol Velloziella Vernonella Voglinoana Warscewiczella Weddellina Wettsteiniella Wildpretina Wingina Winterella Winterina Zukalina Vernacular-classic hybrids—impersonal. Calamovilfa Camphoromyrtus Galedragon Gelatinosporium Iguanura Liquidambar Obaejacoides Sphaeropezia Tacsonia Talinastrum Talinellum Tamarindus Toluifera Vauanthes Vernacular names are invalid; this rule 1s retroactive. “Nomina generica primitiva nemo sanus introducit.” Botanica 22. “Nomina generica, quae ex Graeca vel Latina lingua radicem non habent, rejicienda sunt.” Ibid. 48. 77 Critica “Not to draw names from barbarous tongues, unless those names be frequently quoted in books of travel, and have an agreeable form that adapts itself readily to the Latin tongue, and to the tongues of civilized countries.” 28. Paris Code, Article 78 Frederic E, Clements The vernacular name has long been the refuge of the unlet- tered or indifferent systematist, and will doubtless continue to be while there are biologists of this kind. The arguments against the use of vernacular terms are so obvious and cogent. that they would not be dwelt upon were it not for the contra- dictory provisions of the Paris Code. As in so many other questions of nomenclature, Linné’s pronouncement should have been regarded as final by the framer of the Code. But, as in more than one place, the Code admits a fatal exception. It is absurd to base biological nomenclature in any degree upon books of travel, and it is futile to think that an author who speaks any vernacular tongue whatever, no matter how crude and un- couth, would find it either harsh or disagreeable. Some biolo- gists have endeavored to improve vernacular names by shorten- ing them or by adding a Latin suffix, but such a remedy is worse than the original trouble. Correction by translation, as Chenan- thus for Gansblum, is occasionally possible, and in such cases might be more fortunate than the rejection of a name. The fundamental fact still remains, however, that nomenclature is already essentially classical, and should in the future be made completely so. Vernacular names have no place in it. This condition can be made to prevail only by rejecting all such names whether past or future. Anagrams, if they be considered words at all, are vernacular, since they are neither Greek nor Latin. They are the ultimate product of puerility or illiteracy in nomenclature. Such a series as Filago, Gifola, Ifloga, Logfia, and Oglifa throws a clear light upon the good sense and linguistic taste of the authors con- cerned. One might better make names after the fashion of Car- roll, or take names from the “hog-Latin” of childhood. All other mutilations, like anagrams, are unpardonable offenses against nomenclature, and are to be summarily rejected. I. Anagrams. Alibum (Liabum) Behuria (Hubera) Amida (Madia) Beriesia (Siebera) Anogra (Onagra) Blitrydium (Tryblidium) Baziasa (Sabazia) Galpinsia (Salpingia) 78 Greek and Latin in Biological Nomenclature 79 Gandriloa (Oligandra) Gelfuga (Fluggea) Gifola (Filago) Gosela (Selago) Ifloga (Filago) Lagatea (Galatea) Lebidiera (Briedelia) Narthecium (Anthericum) Neoceis (Senecio) If. Vernacular mutilations. Andreoskia Beccarinda Berkleasmium Bolusafra Brittonamra Cavanilla Clarkeinda Cosimibuena Dickneckera Durandeeldea Elidurandia Fregirardia Gomortega Gonzalagunia Hallomuellera Hasskarlinda IX Nepera (Spennera) Norysca (Ascyron) Obaejaca (Jacobaea) Oglifa (Filago) Parosela (Psoralea) Phledinium (Delphinium) Ranugia (Anguria) Trelotra (Rottlera) Trilisa (Liatris) Tsidrogalvia Itoasia Kinginda Kurzamra Kurzinda Lippomuellera Maximowasia Meyerafra Muelleramra Razumovia Ridleyinda Schinzafra Sebschauera Schweinfurthafra Watsonamra A name is not valid unless its etymology and application are clearly indi- cated: this rule is not retroactive. “Nomina generica, quae Characterem essentialem, vel faciem plantae exhibent, optimae sunt.” “Botanists who have generic names to publish show judgment and taste by attending to the following recommendations: 2) To give the etymolo of each name.” g a) Sy Article 28. 79 Critica Botanica 97. Paris Code, 80 Frederic E. Clements The desirability of being able to know the etymology and ap- plication of each generic and specific name is obvious, but the rule given above will work advantageously in other matters also. An author who cites accurately the derivation of a proposed name will be much less apt to err in its construction, while the necessity for indicating its application will bring about greater accuracy in the choice of characters. Desirable as it might be, it is futile to demand that names show a proper degree of rele- vancy, or to reject them because they are more or less inap- plicable. In matters of taste, it is both possible and highly de- sirable to have a standard, but it is idle to expect that it will be either appreciated or followed by the majority. Since names are to be rejected if improperly constructed, it is imperative that the exact etymology be given in each case, in order that their valid- ity may be readily ascertained. A name then would stand or fall by its given etymology. It is extremely unsatisfactory to say of a name, for example, “from the Greek for flower;” the exact form of the Greek or Latin stem employed should be given. xX The termination of family, ordinal, class, and branch names shall be uni- form within each group: tribes shall terminate in -inae, families in -aceae, orders in -ales, classes in -eae, and branches in -phyta. “The names of divisions and subdivisions, of classes and sub- classes, are drawn from their principal characters. They are expressed by words of Greek and Latin origin, some similarity of form and termination being given to those that -designate groups of the same nature.” Paris Code, Article 18. The designation of all groups of the same rank by means of a common suffix is at present merely a convenience, but with the increasing minuteness of systematic work and the growing ten- dency toward segregation, it will soon become a necessity. Sub- divisions and superdivisions will need to be set off from tribes, fami- lies, orders, and classes, and the terminations for the latter must be definitely fixed in order to secure a basis for distinguishing the next 80 Greek and Latin in Biological Nomenclature 81 group above and below. The number of possible divisions above the genus is fifteen, which makes it impossible that each should re- ceive a distinct suffix. The most satisfactory method, then, will be to fix the designations for the five main groups, and to indi- cate sub and super divisions by prefixes, or by slight variations of the proper suffix. A further reason for this is found in the fact that cognate suffixes can alone be used, since generic names are either Greek or Latin, and that proper cognate suffixes are few. In fact, they are practically exhausted by the five principal groups, -alis, -ales, being, indeed, very hard to justify as a ter- mination for Greek stems. It should be noted that -phyta is merely the neuter plural of the Greek word, ¢vurov, 70, plant, and can be attached only to Greek stems. The following examples will illustrate the operation of the above rule. Protophyta: Schizophyceae: Nematogenales: Nostocaceae: -Au- losirinae ; Phycophyta: Chlorophyceae: Conjugatales : Zygnemataceae: Me- socarpinae Carpcphyta: Ascomyceteae: Discomycetales: Pezizaceae: Sarco- scyphinae - Bryophyta: Hepaticeae: Jungermanniales: Jungermanniaceae: Aploziinae Pteridophyta: Filiceae: Filicales: Polypodiaceae: Onocleinae Spermatophyta: Angiospermateae: Glumales: Graminaceae: Fes- tucinae 81 82 Frederic E. Clements XI In proposing generic names, the following rules are to be observed: (1) The name shall be a Greek substantive, i. e., not a simple adjective. (2) A single generic name may be founded upon the name of a botanist, Such names are only to be formed by adding -ia to cognomina ending in acon- sonant and -a to cognomina in a vowel or -r, except in the case of names al- ready Latinised, in which case the termination is first dropped. (3) Personal generic names shall be bestowed only in recognition of emi- nent services in botany. (4) Anagrams and geographical names are invalid. (5) Double generic names are invalid. Generic names should in the future be formed exclusively from Greek, as simple Latin nouns suitable for plant names have been practically exhausted, and the formation of compound terms in Latin is awkward. Greek nominal stems of all sorts, simple or compound, with the exception of simple adjectives, such as paxpés, peyas, etc., are readily available. The proposal of generic names in honor of rulers, patrons, collectors, friends, and relatives should be severely discountenanced. Furthermore, duplicates of the same personal, as Saccardaea, Saccardia, Pasaccardoa, Sac- cardoella, Saccardinula, and Beccaria, Beccariella, Beccarianthus, Beccardinda, and Beccariodendron must be regarded as invalid, because their terminations are no longer significant endings, but miere variations, and also because they are hybrids. Anagrams, as has been pointed out before, fall because they are vernacular, or mutilated, or both. Geographical names are almost invariably vernacular also. Double generic names, such as Dens-canis and Bursa-pastoris are compounded syntactically and are hence in- valid, while others, such as Genisto-Spartium and Lilio-Narcis- sus are mere hybrids. page ee Ne ee Greek and Latin in Biological Nomenclature 83 XIl In proposing specific names, the following rules are to be observed: (1) The name shall be a Greek or Latin adjective, referring to a characteror function of the plant, orto its habitat. (2) Reduplicative specific names are to be avoided. (3) Comparatives, superlatives, and geographical adjectives are invalid; rot ‘retroactive. (4) Personal adjectives and genitives are invalid; not retroactive. (5) The specific name is invalid if the same as the generic name; retroactive. None of the above rules are of primary importance, but their observance will materially improve the nomenclature of species. They represent the best usage at the present time, but need to be emphasized in order that they may be more generally followed. A specific name should not only mean something, but should also have a direct and evident application to some characteristic of the plant or habitat. In this connection, the necessity for the rules is obvious, though there will doubtless be dissent from the treatment of geographical and personal names. In support of the position taken on geographical names, it is sufficient to cite the names ‘‘canadensis, carolinianus,’’ ‘‘pennsylvanicus,’’ ‘‘vir- ginianus,” etc., of Linnaeus, Gronovius, Elliott, and others, for species found the country over, and the names ‘““coloradensis,”’ ‘Soensis,’’? “missouriensis,” etc., of more recent writers for spe- cies which completely ignore the political limits of their native states. Asclepias syriaca L. is a classical example of the value of geographical names for species. The logical outcome of geo- graphical names is seen in such absurdities as Crataegus raleighensis and Panicum auburne, and, when combined with a ‘proper degree of illiteracy, in such nomenclatural atrocities as Crataegus colorado and C. shallotte. The genus Crataegus fur- nishes convincing proof that nomenclatural and taxonomic in- competence go hand in hand. The practice of naming species after persons has absolutely nothing to commend it. As a rule, personal specific names are the result of a mistaken desire to honor some one, or of*mere laziness. The day is long past in which a biologist can be hon- ored by attaching his name to a species, and the honoring of other persons is not the province of nomenclature. It can not >? 66 ° ug 84 | Frederic E. Clements be gainsaid that the use of personal names for species does ob- viate the necessity of knowing the species of a genus sufficiently- well to avoid homonyms, but it is clear that such knowledge might at least make for more thorough systematic work. With respect to doublets, it is greatly to be regretted that the original rule of the Rochester Code was not permitted to stand. The Madison amendment has not only resulted in numerous absurd one-word binomials, but has actually weakened the cause of priority by making the latter override all considerations of ac- curacy and taste. 84 Greek and Latin in Biological Nomenclature 85 BIBLIOGRAPHY Allen and Greenough. Latin Grammar. 1893. Andrews, Lewis, and Short. A new Latin Dictionary. 1882. Barnhart, J. H. Family Nomenclature. Bull. Torr. Bot Club, 2227... 1895. Buttmann, Philip. A Greek Grammar. 1851. Clements, F. E. A System of Nomenclature for Phytogeog- raphy. Engler’s Jahrb., 31:b. 70. 1902. Curtius, George. Principles of Greek Etymology. 1875. Dall, A. H. Nomenclature in Zoology and Botany. DeCandolle, Alphonse. Laws of Botanical Nomenclature. 1868. Dieterich, Karl. Untersuchungen zur Geschichte der Griech- ischen Sprache. 1898. Gray, Asa. Some Points in Botanical Nomenclature. Az. Journ. Sct., 26:417. 1883. Goodwin. A Greek Grammar. Harms, H. Die Nomenclaturbewegung der letzten Jahre. Eng. Bot. Jahrb., 23:b. 56. 1897. Henry, Victor. A Short Comparative Grammar of Greek and Latin. 1890. Kuntze, O. Revisio Generum Plantarum, I, II. 1891. tid: Ti 1° <1893. Ibid., III, 2. 1898. Liddel and Scott. A Greek-English Lexicon. Linné, Carl von. Critica Botanica. 1737. Philosophia Botanica. 1751. List of Pteridophyta and Spermatophyta. 1893-4. Miller, Walter. Scientific Names of Greek and Latin Deriv- ation. Proc. Cal. Acad. Sct., 1:115. 1897. Pfeiffer, Ludwig. Nomenclator Botanicus. 1873. Proceedings of the Madison Botanical Congress. 1894 Smith and Hall. A Copious and Critical English-Latin Dic- tionary. 1899. Yonge, C. D. An English-Greek Lexicon. 1899. 85 At foot CONTENTS OF THE UNIVERSITY STUDIES, Vot. II No. 1 Additional Notes on the New Fossil, Daemonelix By ERWIN HINCKLEY BARBOUR On the Decrease of Predication and of Sentence Weight in English Prose By G. W. GERWIG Mirabeau, an Opponent of Absolutism By F. M. FLING No. 2 History of the Discovery and Report of Progress in the Study of Dae- monelix By ERWIN HINCKLEY BARBOUR Notes on the Chemical Composition of the Silicious Tubes of the Devil’s Corkscrew, Daemonelix By THOMAS HERBERT MARSLAND . On the Continuity of Chance By ELLERY W. DAVIS The Bacon-Shakespeare Controversy—A Contribution By CARSON HILDRETH Generalization and Economic Standards By W. G. LANGWORTHY TAYLOR F No. 3 Topical Digest of the Rig-Veda Spanish Verbs with Vowel Gradation in the Present System By A. H. EDGREN The Oath of the Tennis Court By F. M. FLING No. 4 Influence of the Breton Deputation and the Breton Ciub in the Revolu- tion ( April—October, 1789) By CHARLES KUHLMANN The Mercantile Conditions of the Crisis of 1893 By FRANK S. PHILBRICK Single numbers of the STUDIES may be bought for $1. 00 Roar. ie JACOB NORTH & CO., FRINTERS, LINCOLN, 82 Vo... III. Aprii, 1903 COMMITTEE OF PUBLICA Son v L. A. SHERMAN C. E. BESSER VAL MUSE H.B.WARD W.G.L. TAYLOR H. H. NICHOLSON T. L. BOLTON R. E. MORITZ F. M. FLING, Eprror CONTENTS I; THE DEGREE OF ACCURACY OF STATISTICAL DATA Carl C. Engberg. ; : , j ; 87 II. THe ANoMALOUS DISPERSION AND SELECTIVE ABSORPTION OF FUCHSIN W. B. Cartmel. Sar aeety : : ‘ 101 III]. MALLOPHAGA FROM BIRDS OF CosTA RICA, CENTRAL AMERICA M. A. Carriker, Jr. ; 3 ; : 123 LINCOLN, NEBRASKA Entered at the post-office in Lincoln, Nebraska, as second-class matter, as University Bulletin, Series 8, No. 9 UNIVERSITY STUDIES VoL. Il APRIL, 1903 No. 2 I—The Degree of Accuracy of Statistical Data BY CARL C, ENGBERG I This paper is written as-a protest against the unnecessary re- _finement of statistical computations as carried out by the biome- ‘tricians of to-day. These practices are well illustrated by the case of the college freshman, who in his zeal and desire for abso- lute accuracy used = to fifteen decimal figures in the determina- tion of the size of the micrometer divisions in his microscope, although the change in focus necessary even for the other eye of the same observer will, in many cases, alter the size of the object observed by as much as half a division. While this is an extreme example, it is not much worse than the performances of all inexperienced computers, or even than those of many distin- guished mathematicians who are experienced computers, but who have had their practical sense killed by impractical theories. In order to give greater weight to my remarks, I shall discuss cases taken from the works of prominent biometricians, espe- cially those of Professor Karl Pearson, the originator and developer of the science. 87 Distribution of 8,680 cases of enteric fever received into the Met- ropolitan Asylums Board Hospitals, 1371-93. Karl Pearson, Carl C. Engberg Il Mathematical Theory of Evolution. 186, pp. 390-91. NUMBER OF Phil. -Trans. A., vol. NUMBER OF he CASES ey CASES ’ Under 5 266 40-45 163 5-10 1143 45-50 98 10-15 2019 50-55 40 15-20 1955 55-60 14 20-25 1319 60-65 8 25-30 857 65-70 4 30-35 503 70-75 1 35-40 299 | The distribution here given for the 13 cases above 60 is what © Professor Pearson “considers” the most probable distribution for _ those years. It seems that the hospital authorities “lumped” their old patients, a practice fatal to accurate statistics. For convenience, we shall take five years as the unit. then get for the first four moments about the vertical through —2.5 years :+ V1==4.294, W=22.341, v3—=137.051, 4967. 682. Transforming to moments about the centroid vertical by means of the relations. pa==0 pa=w—v1 1/6 p3—v3— 3v1¥2-+ 2v3 I pa=v4—Aviv3 + 6v1?v2— 318-1211" ++ at 5 we get: H2=4.071, ps=7.599, p4=69. 314. 1JIn practical computations, a vertical near the mean is used, but as I had worked out these moments for another purpose long before this paper was ai abe , ow es thought of, I preferred to use them here rather than to do the whole work — over again. 88 a s . The Degree of Accuracy of Statistical Data 3 2 Putting a= and eas we find: pea pe Bi=.856 and Bo=4. 182; whence the critical function 3B1—2/2+6=. 203>0, and the equation of the theoretical frequency curve takes the form 2 eet eB a) my =| ms Jo egers I a2 . As the formulae used in determining the constants in the above equation are the best argument for excessive accuracy, I shall _ give them. SGC Ga== Bie E) ~ 38i—2B2+6 9 4+ Bir+2)?/(r+1) Sea ela my—my —I and mms —I where 72; and mz’ are the roots of 7’?—rm'+-e=o. M2 ay} ag—=b and —— ag m2 1 1 a (mbm 1) V mk ms Yas — ae) o— b 1 2777717722 = athe number of cases. Skewness= 14 VA an ‘= ee A, 89 etal) 4 Carl C. Engberg im We then get for the values of the constants: f %==—6S276 g== "i Go5 ete API ATO 3 FSA TA | mj==~ 2.757 my—64.02 : | Mm—= 3.056 ag—=7 1.08 2 Yo—1883 The centroid vertical is at 18.97 years. F Carrying the work out to six decimal figures, Professor Pear- son finds for these constants: / P2==4.070554, Ha=7-598196, py—=69. 379605 | 381 —2B24-6=. 1935 . F=f 2r2 BORe d= °398643 ee €= 259.78912 A= >,40egz2 | b==" 77.28312 : y=" 279201 ma=67.49351 Wi AL Og 508 a9=74.20511 Vo=1890. 83 The centroid vertical is at 18.9691 years; i. e., .29382 unit — from 15-20. To compute the run of a disease to a twenty-thou- ; sandth part of a year is rather fine work, especially when five years is the unit. . A comparison of the two sets of results shows a considerable difference in the values of the constants. Theoretically the latter is the more correct set of values; practically the former is the better. The equations given above apparently necessitate a high degree of approximation, for high powers of the moments and other constants are involved, and the successive powers of an approximate decimal fraction are correct to fewer and fewer decimal places; so that, if we want the values of the constants correct to two or three decimals, we must start with at least six decimal figures in the values of the moments. This is true if we are dealing with six-place data, but suppose we have before us only three-place data? We can then at best make a guess at the fourth figure, but can tell absolutely nothing about the following / go ey y pe The Degree of Accuracy of Statistical Data 5 figures. Here, then, we are compelled to follow a rule which — ought to be more widely observed; namely, use the degree of approximation warranted by the data, and let the answer take care of itself. Against this we have the contention that our data are not merely three- or four-place data, but are of any desired degree of approximation, for Vs ’ and the division here indicated may be carried out as far as we please. Further, the best theoretical curve is necessarily defined as the one which fits the given observations best. Under these pleas have been committed many outrageous crimes against com- mon sense laws in computation, and by the greatest of masters. It will be seen later, however, that a very slight change in the data or even the slightest change in the unit of groupings, in most cases affects the valtie of the moments in the third or fourth place. Under these circumstances, we ought to get sensibly as good a fit with three or four figures as with five or six, but with only a fraction of the work. Further- more, neither curve can coincide with the polygon of observa- tions, and as they differ somewhat in shape, in this place the one, in that the other may be the better fit. If the degree of approxi- mation warranted by the data has been used, the chances are that, on the whole, the one curve will be as good a fit as the other. To determine the degree of accuracy of the above data, I let the number of cases from 15-20 years be 2018 instead of 20109, a very insignificant change. Computing the v's we get: Vi=4.294, v92=22.343, v3=137.072; va—967. 866. These give for the moments about the centroid vertical * pa=4.072,°s=7-595, H4—=69. 347.7 This change is well within the probable error of the number of cases for the given period, and hence this set of values of the 1Had the actual distribution of the 13 cases above 60 been in any way different from the one assumed by Professor Pearson, a much greater change _ in the moments would have occurred. gt 6 Carl C. Engberg v's and ws is as probable as the previous one. Thus we can, by very slight manipulations of our data, obtain a new set of con- stants which differ considerably from the old set, but which must give rise to a curve equally as good as the first one. To carry out the computations to six or more places, when an exceedingly slight accidental change in the data will make a proportionately large change in the constants, is a pronounced case of “saving at the spigot and spending at the bung.” In fig. I are drawn the curves obtained, using six and three decimal figures respectively. A comparison of the two curves with the frequency polygon shows the one to be as good a fit as the other. Ill The distribution of dorsal teeth on the rostrum of 915 specimens of Palaemonetes Varians from Saltram Park, Plymouth. Professor Karl Pearson, The Mathematical Theory of Evolution. Phil. Trans., A., vol. 186, pp. 403-4. TEETH CASES AQT Whe (Se) =] bo The centroid vertical lies .314 of a tooth beyond 4, i. e., at 4.314 teeth, The moments about the vertical through 4 are: W==. 3137, W=.8426, 3=.4973, v4==1.9705. These give to the moments about the centroid vertical the following values: P2=.9109, pa=-—.234, pi= 2.6259; whence Pi=.0724, Be=3.1647, 92 The Degree of Accuracy of Statistical Data 7 and 282—3B1—6 =. 1122>>0, or we have a curve of type IV. Proceeding to determine the other constants, we find: F==TEL. 8: y=110.4 (v is positive since ps3 is negative. Professor Pearson gets these interchanged), a=7.155 m= 56.99. Distance of origin from centroid vertical—=7.057. log yo= 18.0822. Thus the equation of the curve is V=7o costs%ge-10.40, S71 55 ta 0. Professor Pearson’s constants, obtained by the aid of seven- _ place logarithm tables are . pa=.910906 Pi 072222 pa=. 233908 Bo=3.164684 p4=2.625896 2B2—3Bi—6=. 112702 F=11I1.398 a=7.16613 v=109.047 m=56.699 Distance of origin from centroid vertical—7.o149, log vo=18. 4431056, whence the equation of the curve is Y=Vo COSTS. G 108.0479, x=7.16613 tan 6. The two curves are drawn in fig. 2. A comparison of them with the frequency polygon shows that not enough is gained in accuracy by carrying more than four decimal figures to warrant the extra expenditure of time and labor. Thus this problem also verifies the conclusions reached above: ‘that, in the fitting of a theoretical curve to the observations, it is the height of folly to waste time and energy on needless and 93 8 Carl C. Engberg meaningless refinement of the computations; and that, as a rule, in spite of the high powers of the constants, three or four decimal places will be sufficient to give results as reliable as the data warrant. BA in arn The equations ee Y=Yo COS 2m @ e-V0 aa tan have a rather forbidding appearance, and the mere sight of them is often sufficient to deter a man from going farther with the work. If handled properly, however, they are not so bad as they look. Taking logarithms, we have log tan 6=log 1+—log a log y=log y.-+ 2m log cos 6—vM6 log e, where M is the factor which converts from circular to radian: measure. We now take a card and write on it the values of log a, log Tae log (vy loge), and log 2m, thus: log a= , log (vA7 log e)= log 7/5 », log: 2m The proper use of this card will enable us to plot the curve in- one-fifth of the time required did we use no labor-saving devices. ‘ Similar methods may be employed in computing other forms of probability curves. : 94 , The Degree of Accuracy of Statistical Data 9 Vv Another matter that should be spoken of in this connection is the method of computing the v's, i. e., the moments about an arbitrary vertical. Professor Pearson, in one of his articles, pub- lishes a table of the first six powers of the integers from I up to 30, and refers to this table several times. Professor Davenport, in his Statistical Methods, copies this table, and says: “This table is useful in calculating moments.” This is not true, except in very rare instances, and then only to a very insignificant ex- tent; but the statement coming from such authorities is likely to unduly influence inexperienced computers and cause them to waste much valuable time. Instead of multiplying the ordinates by the successive powers of the corresponding abscissas to obtain the moments, it will be found much better to proceed as in the. following example: ss o oa | 2s Mi=Sryr Noe=>r.7 N3=Sr.yrr? | Ma=Sr.yr.P ea 98 4 ~The Degree of Accuracy of Statistical Data 13 CASE I Observed | Calculated > Group frequency | frequency My—My! — | (Wty—Mty/)?/ my it 1 1 0 0 2 8 2 —1l a! 3 50 48 —2 ae ‘4 120 123 3 a 5 195 200 5 1 6 211 200 —ll1 6 7 150 155 5 2 8 100 104 4 2 9 40 38 —2 ol 10 2 1 —l 1.0 otsls... y 877 877 0 K2= 2-5 ' Mr. Elderton’s table for y’==2.5 and z=10 gives P=.98 about. CASE ‘II Group enone face Mr—My! | (My—M;’)? My 1 3 3 0 0 2 25 24 —l 0 3 400 385 —15 6 4 1000 1025 25 6 5 1997 1970 —27 4 6 2095 2015 —80 3.2 7 1431 1475 44 1.3 8 955 990 30 1,2 9 375 395 20 1.0 10 20 np A) —1 1 Totals .... 8301 8301 0 N2=8.4 Mr. Elderton’s table gives P’=.49 about. _ Thus, although the second set of observations has a less per- centage error in frequency than the former, its probability is only half as large. This difficulty is partially obviated by divid- ing x’? by the square root of the ratio of the two frequencies and entering the table with this new value of x’. This, however, gives no absolute measure of the probability. When, therefore, 99 14 Carl C. Engberg the rapid growth of the numerator, as compared with the denominator of the fraction (a2,—m,')? My. ; for large values of 2, is sufficient even to destroy the value of the table, it is sheer folly to fool with ten-place logarithms. Mr. Elderton would have conferred a much greater benefit on bio-— metricians had he put less time on the “small differences in the eleventh place,” and more on the securing of an absolute measure of the probability.? This paper has not been written in a fault-finding spirit by a detractor of the new science of biometry, but by a teacher of the science, in the hope of doing something to help reverse the popular process of swallowing the camel, but straining at the gnat. 14 measure of this probability, which takes no account of the variability — of an object, is of no value whatsoever. I0o SS a PLAT He g@ 6 decimals--------- Curve obtained usin 74.20511 79291 ek ( 9 , J x 3.07801 +. quation: y==1890. 83 [ I E ay wD ete ANE | ~ | H : Ge) | S & Gh Fy o No) eats 9 oe) 5 | g ies ra} cH i v phe aale f = a So) Yo hne Be emh blove) (op) 5 ao \O ~ co ao) ° Co) | Re es + (ss) pC eiaeotayiafta = ge tepeite 5 ae ae OP SG =) ki ba] oO 2) Sree! PLATE Ti ‘6b10°L [eoIVIaA-plosjUI. OF ULSLIC) ‘(Soil by gi Bo[ ‘9 we} £1991 LF" psn apace S09 4 uonenbyY sjeulloap 9 SuIsn ‘paute}qo dAIND *LGo°L yeotAdA-pl0.1}UID 0} ULSLIO ‘ZZBO'SI=" SO] gots Sire Le penis Abe SOo Ct SOE — — — gjeundap > Sutsn poulezqo aaing The Dispersion and Absorption of Fuchsin 1 Il.—The Anomalous Dispersion and Sclective Absorption of Fuchsin* BY W. B. -CARTMEL Although fuchsin is the substance in which anomalous disper- sion was first observed, and although it shows this phenomenon -much more decidedly than any other substance upon which it has been found possible thus far to make anything like reliable measurements, its optical constants have not yet been fully deter- mined except by indirect methods. Very complete absorption curves have indeed been given for solutions of different concen- tration, but not for solid fuchsin. It was therefore thought that it might be of interest to deter- _ mine the absorption and dispersion directly, both of these upon the same identical fuchsin; and as a very good determination of the dispersion curve has already been given by [Pfluger,* who measured the deviation produced by a thin wedge of solid fuchsin, it was decided to redetermine this, using interferential means. This would present the advantage of a redetermination by a dif- ferent method, and, furthermore, the methods of absorption and dispersion could both be made upon the same fluid. Films were therefore prepared in the usual way by dipping glass plates into an alchoholic solution of fuchsin and allowing the alchohol to evaporate. The fuchsin upon which the first experiments were made was some that had been purchased for general laboratory purposes, and it was found that the dispersion and absorption had values very much lower than those given by Pfliiger. Some fuchsin of the same kind as that which Pfluger had used was therefore imported from Kahlbaum in Berlin, and 1Read before the Washington meeting of the American Association for the Advancement of Science. *Wied, Ann., vol. 65, p, 203, 1898. Io! 2 W. B. Cartmel upon this were carried out the experiments which form the basis of the following paper. As already pointed out by Wood?, the great difficulty in deter- mining the dispersion of strongly absorbing substances by inter- ferential methods is that the ray which passes through the sub- stance is so reduced in intensity that it is not capable of causing interference when it meets the undiminished light of the other ray. This may be easily conceived to be the case with fuchsin, when we consider that a film of fuchsin a wave-length thick transmits only five parts in one hundred million of the incident light within the absorption band. Experiments were, therefore, made with the object of reducing, if possible, the intensity of the light in one of the paths of the interferometer, without producing any change in its optical length, in order that the fuchsin might _be placed in the more intense beam. However, none of the _various plans tried were adopted. It was then decided to use a form of interferometer in which the light does not return upon itself, and for two reasons: First, because in traversing the film twice the diminution in intensity is squared while the retardation is only doubled; and, second, because the enormous reflection from the surface of the film obscures the fringes in the ordinary form of interferometer, but in the type used the reflected light | does not reach the observer’s eye at all. After a number of trials to determine the best adjustment, it was found possible, by using this type of interferometer and mak- ing the films sufficiently thin, to obtain distinct fringes through- out the spectrum. An unsymmetrical arrangement was first tried, in which the partly silvered plates were very lightly silvered, or not silvered at all, and the fringes observed at a direction at right angles to that at which the light entered the instrument. ~In this way fringes were obtained from beams of unequal intensity, though the method was finally abandoned for the following, which is more satisfactory. The partly silvered mirrors of the interferometer are silvered so as to reflect and transmit equally, in order that the light in the two paths may be of equal intensity, and then the fuchsin film is 1Phil. Mag., vol. 1, p. 43. 1901. 102 . The Dispersion and Absorption of Fuchsin 3 introduced in one of the paths and an absorbing screen in the other. Good fringes may now be seen because the intensity of the light in the two paths is reduced. The retardation may be measured by the shift of the fringes on the removal of the fuchsin film. But the presence of the absorbing screen causes the fringes to be indistinct when the fuchsin film is removed, so the absorb- ing screen should have only half the absorption of the fuchsin film, in order that the fringes may be seen with equal distinct- ness whether the fuchsin film is in or out of the interferometer, and the absorbing screen in both upper and lower halves of the other path, so that with the interferometer adjusted for vertical fringes two sets were seen, one above the other, but one set dis- placed with respect to the other. Sunlight from a slit S, fig. 1, was brought to a focus by means of the lens L, so that an image of the slit fell upon the glass plate upon which the fuchsin had been deposited, and thus light of very great intensity was concentrated upon a strip of the film only a millimeter wide, a portion narrow enough so that the thickness could be determined definitely. With a wider film only an average thickness could have been obtained. The light, after leaving the interferometer, was brought to a focus upon the slit of a small spectroscope by means of the lens, L’. By observing through the eye-piece of the spectroscope, spectral bands could be seen. At this point it may be as well to mention that it takes very careful adjustment of this type of interferometer to be able to observe the bands either with a spectroscope or a telescope. Even with the naked eye it was found that the bands might, under certain circumstances, be seen with the eye in one position, but with the eye nearer or farther from the interferometer they were invisible. Or again, by moving the eye nearer or farther away from the interferometer, a position might be found in which two sets of bands could be seen, crossing one another at right angles, though this only occurred when the instrument was very care- lessly adjusted. The following method of adjustment enabled very satisfactory fringes to be produced. A telescope having a fairly well corrected objective was focussed as carefully as pos- 103 The Dispersion and Absorption of Fuchsin 5 sible for parallel rays, and then an object at least two or three hundred meters away was observed by means of this telescope in such a way that light from the distant object reached the tele- scope directly and at the same time by reflection from two of the mirrors of the interferometer. If the mirrors were not quite parallel, two images of the object could be seen, so that all that was necessary to make the mirrors parallel was to adjust them till the two images seen in the telescope coincided. In this way, by comparing one mirror with another, all the mirrors of the instrument could be brought into almost perfect parallelism. The final adjustment was made by observing with the naked eye the reftection of a pointed object held near the instrument in such a way that reflections of the object reached the eye by way of the io paths of the instrument, and the images thus seen brought into coincidence by moving one of the mirrors parallel to itself by a screw motion. When the images are thus brought into co- incidence, the two paths of the instrument are equal and the colored fringes of white light may be seen. A simple adjustment of one of the mirrors by trial will widen or narrow down the bands at will, or rotate them through any azimuth, though they can not be so widened or narrowed or rotated through any azi- muth if the mirrors are not nearly parallel. After having arranged the interferometer so that the fringes were all that could be desired, no difficulty whatever was experi- enced in seeing them with a telescope, but there was still trouble in seeing spectral bands when the apparatus was set up, as shown in the figure. The difficulty was found to be in the lenses, L and L’, which, though achromatic and well corrected, were of very short focus, but on changing these for lenses of 25 cm. focal length, the spectral bands were so bright and clear that they could evidently be very much dimmed by the introduction of the fuchsin and still be visible. In order to determine what kind of spectroscope would give the best results when viewing very faint bands, a trial was made using various spectroscopes. The conclusion reached was that a spectroscope of very low dispersive power, having a low power eye-piece, was the most satisfactory, and therefore that kind was used in this work. 105 6 ‘ W. B. Cartel: ° ; iy The bands as viewed with the system just described presented the appearance represented diagrammatically in fig. 2, and showed the anomalous dispersion of fuchsin in a general way at a glance. Fip 2. Air bands. Fuchsin bands. Violet. Red. In discussing these bands we will, for the sake of clearness, call those bands affected by the fuchsin the fuchsin bands, and the others the air bands. The bands as shown in fig. 2 corre- spond to a film too thin to produce a whole band retardation, so, at the two points where two sets of bands coincide, the index of refraction is unity. To determine the index of refraction in any other part of the spectrum, one of the mirrors is moved parallel to itself by means of a screw till a fuchsin band appears in the required portion of the spectrum, unless indeed one is already there. Now, by means of a compensator, an air band is brought into coincidence with the fuchsin band, and the amount of re- tardation introduced by the compensator is equivalent to the retar- dation produced by the fuchsin. In the case of a film thick enough to produce more than one band displacement, the fraction of a band is measured by means of the compensator, and the whole number of bands added to this. It was found that when an attempt was made to produce an arbitrary shift of a definite fraction of a band, either by moving one of the mirrors or rotating a compensator, the bands were disturbed by the mere touching of the mirror or compensator. A very thin mica compensator, thin enough so that it had to be 106 The Dispersion and Absorption of Fuchs 7 rotated through about 20° to produce a shift of one band, was therefore used, instead of the usual thick glass ones, and with this no such disturbing effect was observed, since an infini- tesimal movement of this would not visibly affect the bands. To get around the difficulty of mica having three different indices of refraction, the piece of mica was cut so that one of its axes of elasticity was the axis of rotation. By using plane polarized light, and making the plane of polarization parallel to the axis of rota- tion of the mica, the retardation introduced by the mica when at different angles was always proportional to just the one index of refraction. It was found incidentally that-the introduction of the Nicol prism, N, fig. 2, made the bands much -more distinct. On rotating the Nicol prism through a right angle, the bands becaine more confused than with the Nicol out: The compensation C, fig. 1, was made by cutting a thin piece of mica in two and attaching one of the pieces to a device by means of which it could be rotated through an angle and the angle read off. These two pieces of mica were placed in the same path as the fuchsin, the fixed piece and the fuchsin in the lower half of the path, and the movable piece in the upper half, the line of separation of the mica precisely on a level with the edge of the fuchsin; in both upper and lower halves of the other half was a very thin film of fuchsin that served as an absorbing screen. When both pieces of mica were at right angles to the beam of light, their effect was small, because they affected the air bands alike, being of equal thickness. It could be insured that they were equally thick by removing the fuchsin and observ- ing whether the upper and lower spectral bands coincided when both pieces of mica were in the same plane. It is usual to calibrate a compensator by turning it through different angles, and noting the angles corresponding to dif- ferent numbers of bands displacement. However, for this work, as it was always less than a single band displacement that had to be measured, it did not help to know the angle corresponding to one band displacement or two bands displacement, because the greatest change of curvature of the calibration curve comes between zero and one. 107 W. B. Cartinel \ cae A formula was therefore used which may be deduced as fol- lows: If a piece of mica of thickness ¢ is traversed by a ray which meets it at an angle of incidence 7 as shown in te sketch, the length of the path of the ray will be increased by (u#—2), if the index of refraction of the mica is represented by . Also t >, SEC a VYa— sin’r d=! cos (t—r) pl—d=/[p—cos (7—r)] 108 ‘ a ee ee ssigahasiplid’ ete The Dispersion and Absorption of Fuchsin 9 which reduces to t (V p2—sin*2—cos 2) The introduction of the mica at right angles to the ray will shorten its path by (u—1)¢ The number of wave-lengths retar- dation produced by a piece of mica in the path of the ray when the mica is moved from a position at right angles to the ray, through an angle i, is the difference of these two quantities, and we have at once nh=1(V p?—sin*Z —cos 7)—(p—1)¢ In using a compensator with this type of interferometer, there is a shift of the bands due to the lateral displacement of the beam of light, so it is usual to use a double compensator, one piece in each path, thus overcoming the difficulty by moving both pieces together and giving an equal lateral displacement to each beam. The compensator that was used could be turned through 40°, pro- ducing a retardation of three or four bands, before any effect due to this lateral displacement could be observed, and as the greatest shift that had to be measured was less than a band, it was decided to use a single compensator for the sake of con- venience. Using the above formula, the values of 1 that were observed for one, two, and three bands displacement were found to be consistent. Assuming » to be constant and equal to 1.58 the value of t was found from the formula, and then by substituting back this value the various values of 7A, corresponding to different values of i, were found and a curve of A and i plotted. From this the fraction of a band displacement corresponding to any value of i for any part of the spectrum could be determined. There would be a slight error due to the variation of the index of refraction of the mica for light of different wave-lengths, but as this was less than I per cent from each end of the spectrum to the other, the compensator could be calibrated for one part of the spectrum, and a linear correction made in any other part. As a matter of fact, it was calibrated in the red end of the spec- trum, and as the ‘index of refraction of fuchsin is nearly unity 109 10 : W, B. Cartmel at the violet end of the spectrum, and as (u—1r) is the quantity that is measured, it does not matter very much whether the cor- rection is made or not, considering that there are much graver errors which are unavoidable. | Knowing A, the retardation produced by the fuchsin in light’ of any wave-length, we may determine p, the index of refraction’ of fuchsin for light of that wave length, by the formula mA=(p—1)t ; provided we know #, the thickness of the fuchsin film. This was determined by a method which will be described later. The mea- surements were made-upon a film whose thickness was 580 pp, with the exception of those values falling between A=460 pp and A=590 pw, which were made upon a film 192 pp. thick. The results are plotted in plate I, and in the following table they are compared with the values obtained by Pfliiger* by the prism method, and by Walter? by the method of total reflection ; | Pe INTERFERENCE |. x PFLUGER WALTER METHOD A Siete 2.019 2.055 a sae é 2.086 2.12 706 2.31 meee 2.15 B Hops 2.161 2.22 Li 2.34 exe 2.2 &, BEES 2.310 2.35 634 eats 2.412 2.48 D 2.64 2.684 2.70 539 1,95 Racks 1.98 E Shit 1.912 1.85 FF 1.05 1.074 1.05 Sr 0.83 eat 0.82 455 ites 0.847 0.83 G 1.04 0.95 1.00 425 aah 1.00 1.11 fT cas 1.32 sees The curve is uncertain within the absorption band and perhaps also to some extent in the violet, but in the red end of the spec- 1Zoc. cit. 2 Wied. Ann., vol. 57, p. 396. 1896. II@ Ase Dibaba tit aS a im tapes» Lhe Dispersion and Absorption of Fuchsin II trum it is established beyond a doubt that the values given by Pfluger do not apply to the fuchsin upon which my measure- ments were made. The fact that Pfluger obtained the same values using prisms of different angles seems to show either that the two methods lead to different results, or that the com- position of the two samples of fuchsin was different. There are _ a number of red triphenyal menthane dyes of distinctly different compositions, all of which go under the name of fuchsin. Wernicke’* obtained the dispersion curve of a fuchsin by the prism method and found the indices to range from 1.31 to 1.90. The ome on which the writer’s first work was done had indices ranging from less than unity to about two. Kundt experimented on a fuchsin having two absorption bands in the visible spec- trum, However, there is no doubt but what the fuchsin in ques- tion was made by the same process as that which Pfliiger used, though, having been purchased seven years later, it was probably made in a different batch, and hence there is a possibility of the composition being slightly different. The agreement with Walter in the red is extremely satisfac- tory. The two curves run uniformly parallel to one another, this writer’s being a little higher than Walter’s, which may be aceounted for by an error in the thickness of a systematic error of some kind. A slight change in the thickness would make the two curves almost identical. The thickness of the films was determined as follows: A por- tion of the film is washed away with alcohol, leaving a clean, sharp edge. In general, when a film is thus washed away, there is a concentration of fuchsin at the edge, due to a flowing of the solvent. This is of course undesirable and was eliminated by using a piece of blotting paper with which to clean away the fuchsin. This piece of blotting paper was itself cut to a clean sharp edge and slightly dampened with alcohol. The glass near the film was wiped with it, a new edge of the blotting paper was prepared, dampened, and the operation was repeated until a sat- isfactory edge of fuchsin was obtained. , Now a plate of glass was placed upon the glass upon which 1Pagg. Ann., 155. 1875. 12 W. B. Cartmel the fuchsin had been deposited. There were thus two layers of air of different thickness, one between the fuchsin and the glass, and the other between the two pieces of glass. The difference in the thickness of the two films of air gives the thickness of the fuchsin. The thickness of the air films may be obtained very accurately by the interference of their films, by a method first used by Wernicke’ and modified by Wiener.? White light was allowed to fall on the two air films, and the reflected light was examined with a spectroscope, by means of which interference bands could be seen in the spectrum. The arrangement of the apparatus is shown by fig. 3. The glass plates enclosing the air film are shown at G. P is a totally reflecting prism, one edge of which forms an edge of the slit of the collimator. A Rowland grating, R, sent a spectrum into the observing telescope, and the distance from band to band was measured by means of the micrometer eye-piece. The two glass plates, G, were so arranged before the slit that the film of air between the fuchsin and the glass were before the lower part of the slit, and the film of air between the two bare plates before the upper part. By a suitable mechanical contri- vance, the plates were adjusted so that the bands seen at the eye- piece were vertical. On examining the two sets of bands, it was noticed that while those corresponding to the film between the two glass surfaces were regular and became uniformly narrower toward the violet, those corresponding to the air film between the glass and the fuchsin showed an irregularity. For instance, when there was a film of fuchsin whose thickness was about a fourth of a wave-length of red light, the bands of one set were displaced with respect to the other set by half a band in the red, and this displacement regularly increased, till a point in the blue-green was reached where it was a whole band. At this point there was a sudden change of half a band, and then a regular change took place, going from short to shorter wave- 1Pogg. Ann. Erghd. 8, p. 65. 1878. 2Wied. Ann., vol. 31, p. 629. 1887. II2 The Dispersion and Absorption of Fuchsin 13 lengths, till in the violet there was again a whole band displace- ment. This brings up the question of phase change. There was evi- dently about half a band phase change in the blue-green, but in the violet end the red fuchsin reflects like a transparent body, R Fig 3. | - % ss Cae | and it is likely that the difference of phase change between it and the glass is zero. Pfliiger has gone into this point very thor- oughly, and says that from the experiments of Wernicke it is safe to assume that the phase change by reflection from fuchsin 113 14 ; » .. W. B. Cartmel. is the same as-that from glass for wave-lengths longer than 640 wp. The measurements were, therefore, made with the light comprised between the B and C lines. Sunlight was used and a number of gelatin screens interposed at S, fig. 3, so as to keep out light of those wave-lengths not needed, because the films bleach easily. ; The thickness of the fuchsin may be determined from. the twe sets of interference bands in several ways, but the method involved in the following formulas, due to Wiener, was the one used: We have | 21d, = (m+ 1) Amt when ¢ is the thickness of the film between the two pieces of glass, A,, the wave-length corresponding to the center of one band, and 4,,4, that of the next band towards the violet. Tak- ing the one of the bands due to the film of air between the glass and the fuchsin, which falls between the m” and (m-+-1)* of the other bands, and calling it ’,, we have: / | 2f=mMN in where ?’ is the thickness of the film of air corresponding to this case. From these, two equations follow at once: yao Am—V' mn ; A mH Xm — rm 2 which gives the thickness of the fuchsin ¢—t’ directly. The thickness of the air films was regulated so that about five or six interference bands fell between the B and C lines. Thus there were several sets of bands upon which independent mea- surements could be made. : Some idea of the accuracy of the measurement can be obtained from the following set of measurements made upon one of the films, in which is included every measurement that was made upon that film: : 114 exist couecs {hay iy Po g ees Se ee “wo ee The Dispersion and Absorption of Fuchsin ie HALF THE WAVE- THICK- f MEASURED x LENGTH [Oca rep NESS 661 36 625 661 46 615 667 57 610 607 60 606 672 70 602 672 67 604 677 92 585 677 89 588 660 37 623 660 48 612 665 67 608 665 60 605 671 70 601 671 72 604 676 89 587 676 82 594 This makes tle probable error of a single observation 8 yp and of the mean 2 pp, which is about the order of accuracy that inter- ference methods usually give. There is some little uncertainty sometimes as to whether the shift refers to a fraction of a band or to one or two whole bands plus a fraction of a band. In the above instance, the shift is about eight-tenths of a band, and the total displacement one and eight-tenths, giving a thickness of about nine-tenths. If the dis- placement had-been assumed to be two and eight-tenths bands, the resulting thickness would have been one and four-tenths wave- lengths. To decide this point definitely a number of films were made of different thicknesses. The first of this series was made from a saturated solution, and the rest of the films were made, each succeeding one thinner than the one which pre- ceded it, by continually diluting the solution from which they were deposited. Starting from the thinnest (which was 320 pp thick), the thickness of each succeeding one was measured till the thickest was reached (617py). The continuity of the series furnished data from which the thickness was deter- mined definitely, and observation upon the films with a spectro- photometer confirmed the conclusions. 115 16 W. B. Cartmel The same trouble was experienced in the measurement of the retardations with the interferometer, but when the thickness was definitely known a comparison with a film sufficiently thin to show bands all through the spectrum cleared up this point, for with the very thin film the retardation was never so much as a whole wave-length, for reasons already mentioned. The matter might have been cleared up by viewing the fringes of white light in the usual way and observing the shift of the central fringe produced by the fuchsin, but this was found impracticable on account of the selective absorption of the fuchsin fiim, which made the central fringe look just like any other fringe. This plan would have, however, furnished a very decisive solution of the difficulty with regard to the thickness, for by making the plate upon which the fuchsin was deposited the back mirror of an interferometer, and comparing the fringes due to the reflection from the bare plate with those due to the reflection from the top surface of the fuchsin, the shift of the central fringe could have been easily determined. The photometric measurements were made ae means of a Brace spectrophotometer. The methods of adjusting the instrument have been already given by Tuckerman. * The extremely great absorption gave rise to some difficulties which do not occur in ordinary photometric work. For instance, a little of the red light (for which fuchsin is transparent) reached the eye as stray light by reflection in the telescope tubes and the comparison prism. This, too, in spite of the fact that the tele- scope tubes were well blackened and diaphragmed. This small amount of red light was more intense than the green light that was being measured, since the green light had been cut down a thousand times or more by the absorption of the film. This was remedied by using a screen having absorption bands in the red and the blue, but which transmitted green quite freely. In this way green light alone entered the instrument when mea- surements in the green were being made. Fig. 4 shows the general arrangement of the apparatus. A beam of sunlight is brought to a focus on each of the totally re- 1A4strophysical Journal, p. 145. Oct., 1902. 116 aia ee tds Rete am ee Se TS - The Dispersion and Absorption of Fuchsin 17 flecting prisms p and fp’ by means of a split lens S. Then prisms reflect the light on to the mirrors 7 and m’, and these direct the light on to the collimator slits. Part of the light from the right-hand collimator reaches the observing telescope by s Feo 4. reflection from a silvered strip cemented in the comparison prism P. The rest passes above or below the strip and is lost. On the other hand, of the light which comes from 117 ‘18 - -W. B. Cartmel the left-hand telescope, only that which passes above or be- ~ low the silvered strip reaches the observing telescope, with the result that in the eye-piece may be seen a field illuminated . from the left-hand collimator, with the exception of a central strip illuminated from the right-hand collimator; and when the inten- sity of the illumination from both these sources is alike, the field appears uniform. This uniformity may be brought about by varying the width of either of the slits s or s’. If the fuchsin film is placed before the slit s it will cause the center of the field to seem dark, and a match may again be produced by narrowing the slit s’, and if we know the original width of the slit, the ratio Fig 5, of the change in width to the original width gives very closely the proportion in which the introduction of the fuchsin has di- minished the intensity of the light. If, instead of diminishing the intensity at s’ by narrowing the slit, a revolving sector is used which cuts down the light by the proper amount, the error due to the lack of proportionality between intensity and slit-width does not enter in. However, it is found more convenient to use a sector cut into an arbitrary number of parts, as shown in fig. 5, and as with this we can, in general, only make an approximate match, the varying of the slit-width is used as a fine adjustment. This gives practically as good results as can be obtained with a variable sector. A sector of eighty notches was used except for 118 The Dispersion and Absorption of Fuchsin 19 the measurements within the absorption band, in which case a tin disc, as shown in fig. 6, was used, having a slit cut in the edge 0.7 mm. wide and another, about a cm. from the edge, I.5 mm. wide. This disc was rotated very rapidly before the collimator slit by means of a small motor, and thus the intensity of the light incident upon the collimator slit was diminished by 99.856 or 99.682 per cent, according to which of the slits of the disc was used. Since the edges of the slits of the disc are parallel, care had to be taken to have the collimator slit at a definite distance from the center of the disc. This was done by using the disc close up against the collimator slit; and then, by holding a sharp edge on a line B or C scratched on the disc, the shadow of this edge would fall on the center of the collimator slit, when the center of the disc was at a distance OC or OB from the ray entering the collimator. The use of this tin disc, which cut down the intensity of the light so greatly, made it possible to match very accurately, al- though the light had been reduced by the fuchsin to such an enormous degree. Since the slope of the absorption curve was very steep, it was necessary to have the collimator slit s, quite narrow, not more than half a millimeter wide at most, in order that the readings should be made with sufficiently homogeneous light; otherwise the absorption as observed would have too low a value. With the slit s half a millimeter wide, s’ would have been less than a five-thousandth part of a millimeter wide, when the adjustment was made using no sector, and accurate settings of the slit at this width would have been impossible. This fea- ture of the photometer, together with the fact that a bare flame or direct sunlight can be used instead of a flame or sunlight behind ground or opal glass, which would cause a loss of 95 per cent to 98 per cent, and the absence of Nicol prisms, which would cause a further loss, makes it singularly well adapted for work on strongly absorbing substances. The diminution in intensity caused by transmission through the fuchsin film is due to two factors, reflection from its surfaces and absorption within the film. To determine the absorption it is necessary either to eliminate the reflection by measuring the dif- 119 20 W. B. Cartmel ferential absorption of two films of different thickness or by making a separate determinative of the reflection and subtracting it. Preparations were made to measure the reflection, but this part of the work was not finished, though I expect to continue this work and measure the reflection. Not having the reflection measurements, [ must be content with computing them. The effect of an error in the reflection upon the final values of the absorption coefficients will be greatest within the absorption bands, and the 10 per cent change in the reflection only changes the absorption coefficient I per cent. For computing the reflection from an absorbing substance into air, Cauchy’s theory leads to the following formula: (u—1)?-+R? Conn while the electromagnetic theory leads to pa GTR) +126 pe (i-F AR") 2p which may be seen to be identical with the previous formula when we remember that in Cauchy’s theory the quantity k is identical with wk of the electromagnetic theory. For the reflection at the interface between an absorbing substance and a transparent sub- stance whose indices of refraction are w and yw’ respectively, the electromagnetic formula becomes (ue) ek Cam Pek According to this formula, the reflection is the same on either side of the interface between the media, while with Cauchy’s formula the reflection on the two sides will, in some instances, be different, which would lead to the loss by reflection being differ- ent, according to whether the light went through the glass plate and then through the fuchsin or through the fuchsin first, an effect which the writer has been unable to observe. The values of the reflexion as computed in this way are plotted in plate III. 120 The Dispersion and Absorption of Fuchsin 21 If m, 72, and 7, represent the reflection at the fuchsin-air, _ fuchsin-glass, and air-glass surfaces respectively and / the orig- inal intensity of the incident light, the effect of the reflection is to make the transmitted light have the value (1—7) (1—72) (1—73) Lo, neglecting the effect of multiple reflections which does not enter in within the absorption bands. For those colors for which the film is transparent, the effect of multiple reflection is easily al- lowed for to the degree of accuracy possible when working with strongly absorbing substances. Knowing the thickness of the film, we may compute the absorp- tion coefficient wk from the following formula: | I Ee sal a RG Fp a Ct) eS in which ‘s is the ratio of the intensities as determined by the oO spectrophotometer, ¢ is the thickness of the film, and A the wave- length in vacuo. The photometric measurements were all made upon a film 192pp thick. The transmission of this film is plotted in plate II. Using these values, the curve given in plate I was computed. In the following table these values are compared with Pfliiger’s and Walter’s results: Values of pk A _|PFLUGER| WALTER |CARTMEI, 589 0.79 0.792 .86* 527 1.22 1.419 1.354* 486 0.98 1.168 TOT 455 0,43 0.533 0.56* Walter used Cauchy’s formulas for the elliptic polarization which is associated with metallic reflection, and as this could only * By interpolation. 12] 22 W. B. Cartmel be applied to those radiations which are strongly absorbed by the fuchsin, he obtained only the four values given above. In order to make a comparison with Walter’s work, Pfluger measured the same four quantities directly, using the difference in transmission of two films of different thickness. In conclusion, my best thanks are due Professor Brace for help and encouragement throughout the work, and for the excel- lent laboratory facilities afforded me at the University of Ne- braska, where the work was done. I must also thank Professor B. FE. Moore for various courtesies extended to me, and for sug- gestions in regard to the photometric measurements. a 122 > tea wy ‘ ye : b 4 7 q 2 rf ORME MTN Pipa Oye hy: - rane Pee ie) fe) ° cay GESSERSEERezEERmUn EY acbu range UNARNE SERRATE ?= 5: _ Absorption Coe + i oa % be saeeau ‘ wa 8 6aeK enna Gisacance 5 2 apewaua , SUPSaeSeow Ss Hestgeseageee ieee ff Cale Sez ates st bras ed ne ie 5s e+ 4 bc, bl ba _* 00L 009 00s 00+ ood 009 00S 00% *S1OJOU|||{WOD] A] U] SY}TUI|-9ARM % Ol 1%oSe Yo 0% 0S % oe % SL *OILJANS SSVIH-U|SYON} Woy UONIOJOY "g BAND *QOLJANS AIL-U|SYON} WO} UO}PO|JOY "y BAND *YOLYI SUOIIIW ZGI"0 UJSYON} JO WIL} B JO UOJSS|WSUR | Se Vas Rarer & : fe f , | | a CESS Se ee oe %o 001 Ill ALVW Id - I] HLW Id I11.—Mallophaga from Birds of Costa Rica, Central America BY M. A. CARRIKER, JR. INTRODUCTION The specimens of Mallophaga, upon which the contents of this paper are based, were collected by the author during the summer -of 1902, while engaged in a search for natural history specimens in different regions of Costa Rica, Central America. As will be seen from the text, the birds were collected in three principal localities, namely: Juan Vinas, on the Atlantic slope, at an alti- tude of approximately 3,000 ft.; the volcano Irazu, situated on the continental divide, the majority of the specimens being found between the altitudes of 8,000 and 10,000 ft., although some were collected at the summit, which attains an altitude of 11,198 ft.; Pozo Azul, on the Rio Grande de Pirris, about thirty miles from the Pacific coast, with an altitude of not more than 300 ft. The birds collected are, with the exception of a few duplicates at the University of Nebraska, now in the possession of the Car- negie Museum, Pittsburg, Pa., and I am greatly indebted to Mr. W. E. Clyde Todd, curator of birds and mammals in that insti- tution, for his kindness in giving me the correct determination of the bird hosts, the host list given in this paper having been arranged and corrected by him. I also extend thanks to Mr. C. F. Underwood, of San Jose, Costa Rica, for his kindness in giving me determinations for my specimens while in that coun- try, and for the privilege of examining birds, collected by him, for Mallophagan parasites. Lastly, my thanks are due Professor Lawrence Bruner, of the University of Nebraska, under whose direction this paper was written, for his many valuable suggestions during the course of the work. The types of all new forms described in this paper are in the collection of the author, with co-types of most of them in the 123 2 M.A. Carriker collection of the Department of Entomology in the University of . Nebraska. A few words in regard to the distribution of Mallophaga col- lected would probably be interesting. It was found that birds belonging to genera not strictly tropical, and found principally in the higher altitudes, were much more often infested with the parasites than those strictly tropical and inhabiting the lower altitudes. This statement, however, is only true in a general sense, but when exceptions were encountered it was usually found that the parasites belonged either to the genus Menopon or Colpocephalum, and were present in great numbers, these two genera, especially the latter, having been taken in large numbers on strictly tropical birds. The list of Mallophaga collected on Tinamus robustus is a very interesting one, including two new genera and several new species, all of which are very aberrant forms of the genera in which they have been placed. Other very interesting forms are Colpoce- phalum extraneum sp. nov. and mirabile sp. nov. collected on Nyctidromus albicollis and Zarhynchus wagleri, in that they pos- sess a very marked mesothoracic suture and have the metathorax enormously developed posteriorly. It seems to me that many of the present genera need a thorough revision and that some of them should be split up into two or more genera or sub-genera, Colpocephaluin, Menopon, and Physostomum especially needing it. BIBLIOGRAPHY Burmeister, H. Handbuch der Entomologie. Berlin, 1832-33. Carriker, M. A., Jr. Some New Mallophaga from Nebraska Birds. Jour. N. Y. Ent. Soc., vol. X, pp. 216-229, 1902. Denny, H. Monographia Anoplurorum Britanniae. London, 1842. Geer, Charles Baron de. Méimoires pour servir a Vhistoire des insectes. Stockholm, 1752-78. Gervais, Paul. Histoire naturelle des insectes apteres. Paris, 1847. 124 a wp ee Mallophaga from Birds of Costa Rica, Central America 3 Giebel, C.G. IJnsecta Epizoa. Leipsig, 1874. Kellogg, V. L. List of Biting Lice (Mallophaga) Taken from Birds and Mammals of North America. Proc. U.S. Nat. Mus., vol. XVII, pp. 39-100. — New Mallophaga I. Contributions to Biology from the Hopkins Seaside Laboratory of the Leland Stanford Junior University, no. IV, 1896. Kellogg, V. L. New Mallophaga II. Contributions to Biology from the Hopkins Seaside Laboratory of the Leland Stan- ford University, no. VIII. Kellogg, V. L., and Chapman, Bertha L. Mallophaga from Birds of the Hawaiian Islands. Jour. N. Y. Ent. Soc., vol. X, pp. 151-69. — Mallophaga from Birds of Pacific Coast of North America. Jour. N. Y. Ent. Soc., vol. X, pp. 20-28, 1902. — New Mallophaga III. Contributions to Biology from the Hopkins Seaside Laboratory of the Leland Stanford Junior University, no. XIX, 1899. Kellogg, V. L., and Kuwana, S. I. Mallophaga from Alaskan Birds. Proc. Acad. Nat. Sci. of Phil., pp. 151-59, 1900. — Mallophaga from Birds of the Galapagos Islands. Proc. Wash. Acad. of Sci., vol. IV, pp. 457-99, Sept., 1902. Nitzsch, Chr. L. Die Familien, und Gattungen der Thierinsecten (Insecta Epizoica). Mag. der Ent. von Germar und Zinken, t. III, pp. 261-316, 1818. Osborn, H. Insects Affecting Domestic Animals. U. S. Dept. Agri., Bull. 5, New Ser. — Mallophaga Records and Descriptions. Ohio Nat., vol. II, pp. 175-78. — Mallophaga Records and Descriptions. Ohio Nat., vol. II, pp. 201-04. Piaget, E. Les Pediculines, Leide, 1880. — Les Pediculines, Supplement, Leide, 1885. —— Quelques nouvelles pediculines, Tidschr. voor Ent. XXXI, pp. 147-64, 1888. —— Quelques nouvelles pediculines, Tidschr. voor Ent. XX XIII, Pp. 223-59, 1890. Taschenberg, E. L. Insektenkunde. Bremen, 1879. 125 - c \ pet ere ee oe eS ee | | on fa 4 M. A. Carriker DESCRIPTIONS AND DISCUSSIONS Docophorus bisignatus N. Giebel, Insecta Epizoa, p. 106, pl. IX, fig. 9. Piaget, Les Pediculines, Sup., p. 11, pl. II, fig. 1. 4 Numerous individuals of both sexes collected on Guara alba at Poza Azul, Costa Rica, June, 1902. This species is ready recognized by the bilobate, cly peal, signature. Docophorus platystomus N. Nitzsch, in Burmeister, Handbuch d. Ent. 1839. Il, pp. 420. Y Giebel, Insecta Epizoa, p. 69, pl. IX, fig. 1. Denny, Anoplurorum Brittaniae, p. 108. Piaget, Les Pediculines, p. 17, pl. I, fig. 1. Osborn, Notes on Mallophaga and Pediculidae. Canad. Pent.) 3a4. Xk Vile pe sloy. a Teer Affecting Domestic Animals, p. ones Numerous specimens collected on Buteo borealis costaricensis, volcano Irazu, April;.on Buiteo abbreviatus, Pozo Azul, June, 1902. While these specimens do not exactly agree with Piaget’s plate and description in some particulars, they can be referred to that species. So Se Docophorus platystoma umbrosus var. nov. Mate.—Body, length 2.22 mm., width 1.02 mm.; head, length .86 mm., width .81 mm. Much larger than Piaget’s measurements for platystomus, especially the head, which is much narrower in proportion at the temples, with the clypeus very much narrower and deeply emarginate at the tip; the whole head is darker, also the lateral markings of thorax and abdomen; the pustules in the lateral fasciae are very clear and prominent, not obscured as in t'atystomus, while the great number of dorsal. abdominal hairs of Piaget’s specimens are wanting, there being but a scattering median row on each segment. Two males collected on Leucopternis semiplumbea, at Pozo Azul, Costa Rica, June, 1902. 126 Mallophaga from Birds of Costa Rica, Central America 5 Docophorus transversifrons sp. nov., pl. I, fig. 1 Mare—Body, length 1.76 mm., width 62 mm.; brownish golden; head large with broad truncate clypeus, and abdomen subclavate, with narrow, light smoky brown lateral bands and golden brown transverse bands. Head, length .66 mm., width .58 mm.; deep golden brown, with long, broad, truncate clypeus and constricted, rounded tem- ples; occiput convex; clypeal angles with two hairs; one mar- ginal and one submarginal hair at clypeal suture; two submar- ginal hairs just before trabeculae; trabeculae rather small for the genus, pointed, golden; antennae short and stout, first joint thickest, second longest, third and fourth subequal, shortest, and all uniformly golden brown; eye convex, colorless, with a hair; temples evenly rounded, with two rather long hairs and several short bristles; narrow, dark brown, antennal bands, broken at clypeal suture, not reaching clypeal angles, and with bases swol- len and bent inward at trabeculae; paler brown, internal, sinu- ated bands inside of antennal bands; clypeal signature large, broad pointed posteriorly, slightly emarginate at sides, scarcely reaching end of clypeus, and whole deep golden brown; mandi- bles heavy, chestnut; pale brown occipital bands running from sides of occiput to bases of antennal bands; temples narrowly edged with brown, and whole region brownish. Prothorax small, quadrangular, with slightly diverging sides; posterior margin bluntly angulated; posterior angles with one stout hair, lateral bands of deep brown, pale at the margins and curving around on posterior border nearly to middle of segment ; interior of segment pale golden brown; brown coxal bands vis- ible. Metathorax short, transverse, with strongly convex, diverg- ing sides, and flatly angulated posterior margins, with three hairs on each side; lateral submarginal bands of deep brown; whole segment golden brown. Legs rather short, stout, with well-de- veloped tibiae and aborted tarsi; uniformly concolorous with body, with the exception of pale brown semiannulations at tips of femora and tibiae. Abdomen short, subclavate, with lateral margins of segments ‘slightly convex, and rounded, projecting, posterior angles fur- 127 6 M. A. Carriker nished with two hairs in segments four to seven, four on each side of the eighth; one hair on posterior margin of segments three to seven just within the lateral bands; a median row of long slender hairs across the middle of segments two to six; narrow lateral bands of translucent brown, broken at the sutures and projecting into adjacent anterior segments; clear pustules at spiracles on segments two to seven; transverse bands of golden brown, separated by clear sutures and broken medially; apical segment clear, with brown tip; genitalia deep brown, short, stout, with inward curving points. | FEMALE.—Body, length 2.15 mm., width .83 mm.; head, length .71 mm., width .66 mm.; larger and lighter than male, with more oval abdomen; lateral bands narrower and transverse bands shorter and more widely separated; apical segment large, rounded, clear, with a median, transverse blotch. Two males and one female collected on Micrastur guerilla, at Pozo Azul, Costa Rica, June, 1902. A rather unique form for a Raptorial bird, but one readily recognized by the long, broad, truncate clypeus. 7 Docophorus californiensis Kell. Kellogg, New Mallophaga II, p. 483, pl. LXVI, fig. 6. Numerous males and females collected on Melanerpes formi- civorus and Dryobates villosus jardinti, volcano Irazu, April, 1902, and on Chloronerpes yucatanensis and Melanerpes aurifrons hoffmani at Juan Vinas, Costa Rica, March, 1902. Docophorus bruneri sp. nov., pl. I., fig. 2 FEMALE.—Body, length 1.19 mm., width .42 mm.; clear umber brown, with large head having clypeus pale golden and temples deep brown; abdomen elliptical, with narrow, pitchy, lateral bands and median, transverse, brown bands. Head, length .42 mm., width .42 mm.; very large in proportion to the body, with broad, truncate-conical clypeus, slightly swol- len at clypeal angles; one hair in clypeal angles, one marginal and two submarginal ones on sides; narrow golden brown. an- tennal bands; clypeal signature large, pale, pointed posteriorly, 128 Mallophaga from Birds of Costa Rica, Central America 7 with dusky lateral lobes projecting posteriorly, and the whole separated from adjoining parts by a narrow clear line; trabeculae large, curving, bluntly pointed, concolorous with front of head; antennae long and stout, with first two segments longest, and equal, and last three shorter, and subequal, all with semiannula- tions of brown and ground color same as temples; eyes large, protruding, clear, with a very fine bristle; temples evenly rounded with two short pustulated hairs and two bristles; occiput convex, heavy, pitchy brown, occipital bands paler and narrower ante- riorly, running forward past the posterior root of mandibles to ‘bases of the antennal bands; mandibles large, chestnut; whole temple from occipital bands to margin deep, clear brown. Prothorax small, short, quadrangular, with concave anterior margin, convex sides and posterior margin; posterior angles with one hair; narrow, dark brown, submarginal, lateral bands; brown coxal bands visible; metathorax larger than prothorax, pentag- onal, with convex diverging sides and sharply angulated pos- teriorly ; posterior angles with one long hair and two short ones ; posterior margin with five long hairs on each side; dark brown lateral bands, paler at margins; meso-coxal bands visible. Legs stout, with swollen femora and tibiae, concolorous with thorax, with a darker spot at the tips of femora and inner base of tibiae. Abdomen short, elliptical, with posterior angles acute and slightly projecting, furnished with one hair in fourth segment and two in segments five to seven; narrow pitchy brown lateral bands on segments one to seven, separated at the sutures and projecting into adjacent anterior segments; large clear pustules at the spiracles on segments one to seven; a long pustulated hair arising from the posterior margin of segments one to seven, just within the lateral bands; a second long pustulated hair on pos- terior margins of segments one, two, three, half way in toward middle of abdomen; a row of six long, pustulated hairs on the median posterior portion of segments four to six; eighth segment with two long hairs on each side; ninth segment very short, indented at tip; clear brown transverse bands, interrupted medi- ally, on segments one to seven, and separated by clear sutures; segments eight and nine entirely brownish; median, ventral trans- 129 8 M. A. Carriker verse bands, of dark brown, separated by broad clear sutures, except between segments five and six. Matre.—Body, length 1.00 mm., width .42 mm.; head, length .37 mm., width .37 mm.; markings same as female; genital hooks short and stout, tip of abdomen clear. Numerous individuals of both sexes collected on Menacus can- daei, at Juan Vinas, Costa Rica, March, 1902. A species ren- dered striking by the large head, pale clypeus, and dark brown temples, and narrow lateral bands of abdomen. Docophorus underwoodi sp. nov., pl. I, fig. 3 FEMALE.—Body, length 2.09 mm., width .89 mm.; clear, with bands and spots of deep, clear brown and pitchy. Head, length .70 inm., width .62 mm.; conical, end of clypeus emarginate, clear, with one hair in rounded angles; sides con- cave, with a short hair at the suture and two submarginal ones back of suture; trabeculae large, bent, bluntly pointed and pale brown; antennae medium, first segment largest, colorless, second a trifle smaller, and last three smaller and subequal, and last four deep brown, with narrow clear bases; eye large, clear, with a hair; temples evenly rounded, narrowly margined with brown, with two pustulated and two non-pustulated hairs; occiput con- vex; clear brown antennal bands, broken at the sutures and not reaching clypeal angles; internal narrow bands of clear brown, following margin of ocular band, running backward half way to the mandibles, then bending inward a short distance, then straight back, and finally spreading out to bases of mandibles; area be- tween antennal and internal bands a pale brown; clypeal signature long, narrow, dark at tip, emarginate, with posterior portion gradually tapering to a point at mandibles; mandibles large and_ heavy, chestnut ; heavy, pitchy occipital bands running from sides of occiput to bases of antennal bands; a pitchy ocular band from eye to junction of antennal and occipital bands, a small, dark brown occipital signature. / Prothorax quadrilateral, with rounded anterior and posterior angles and flatly convex, slightly diverging sides; dark brown lateral, marginal bands and a pitchy line across postero-lateral 130 Mallophaga from Birds of Costa Rica, Central America 9 portion of segments. Metathorax scarcely longer than prothorax, pentagonal, with convex, widely diverging sides and angulated posterior margin; posterior angles with one short hair, and pos- terior margin with a row of ten short, pustulated hairs on each side; deep clear brown lateral bands, widening posteriorly; a pitchy brown band around posterior border, slightly submarginal in median portion ; meso-coxal bands visible. Legs stout, femora and tibiae swollen at the tips; clear, with bases and tips of femora annulated with pitchy; a pitchy band along posterior border of tibiae, and tips annulated with brown. Abdomen broadly oval, with lateral margins of segments more or less convex and the rounded posterior angles protruding, with one hair in segments two to eight and one in anterior portion of segments five to seven; a row of hairs across the middle of seg- ments one to seven, pustulated along posterior margin of trans- ' verse bands; narrow, pitchy, lateral bands on segments one to eight, separated by a small clear spot in posterior angles; clear brown, lateral transverse bands, extending inward one-third the width of abdomen, narrowing inwardly, with a large, round, clear area just within the lateral bands and a darker oval spot near the inner ends; a large, round, ventral, brown patch cover- ing the apical portion of the abdomen; two median brown spots on the eighth segment and the tips of the ninth brownish. Matre.—Body, length 1.61 mm., width .62 mm.; head, length .63 mm., width .55 mm.; very similar to female, except the shorter, rounder abdomen; genitalia short, stout, deep brown. Numerous males and females collected on Psilorhinus mexi- canus, at Juan Vinus, Costa Rica, March, 1902. This form is of the type of D. rotundus Piag. and corvi Osb., but is easily distin- guished from corvi by the slightly fuscus color and dark brown, instead of pitchy markings, and from rotundus by the absence of lateral angles in the prothorax and by the double occipital bands ; also female is much larger. Docophorus communis N. Nitzsch, Germar’s Mag. of Ent., 1818, vol. III, p. 290. Kellogg, New Mallophaga II, 486, pl. LXVI, fig. 7. Large numbers of this type of Docophorus collected on a great 131 10 M. A. Carriker variety of hosts (see host list). I have made no attempt to sep- arate them into varieties from lack of time and material, much as they need it. As Mr. Kellogg says, this group needs revision badly, but it is a difficult undertaking, with much time and material as a requisite. Docophorus cancellosus sp. nov., pl. I, fig. j FEMALE.—Body, length 2.51 mm., width .83 mm.; head and thorax translucent, smoky brown; abdomen clear, with pitchy lateral, and deep brown, transverse bands; head conical with narrow emarginate clypeus, and squarish temples. Head, length .74 mm., width .66 mm.; conical from trabeculae forward, with sinuate sides and narrow emarginate front; two short fine hairs before trabeculae, which are of medium size, bluntly pointed and golden brown; antennae rather small, first joint thickest, second joint longest, last three shortest and sub- equal, all uniformly golden brown; eye large, clear, prominent, with a short hair in the middle and one at posterior margin; temples squarish, with rounded angles, furnished with several short hairs; occiput slightly concavo-convex ; heavy, deep brown antennal bands, widening and darkening to pitchy at bases, which bend inward almost to bases of mandibles; paler brown bands inside of antennal bands; clypeal signature very indistinct ; man- dible very heavy, chestnut; temples narrowly edged with pitchy brown and whole templar regions deep smoky brown inward to the broken, occipital bands, which do not reach the occipital mar- gin; short faint ocular bands running inward from eyes to occipital bands; a small, brown, occipital signature. Prothorax small, quadrilateral, narrowed in front, with con- cave interior margin, convex diverging sides, and concave pos- terior margin; posterior angles rounded, with a short hair; deep brown, lateral bands, curving around on posterior portion of seg- ment; prominent brownish coxal bands; metathorax larger than prothorax, with convex, diverging sides and angulated posterior margin; posterior angles with one hair, and posterior margin with five hairs on each side, three near the angles and two farther in; broad lateral bands of brown and a broad, pale brown, trans- verse band in median portion; coxal bands prominent. Legs 132 Mallophaga from Birds of Costa Rica, Central America 11 short and stout, deep smoky brown, with pitchy edging on ante- rior margins of femora and both margins of tibiae. Abdomen rather long for the genus, elongate elliptical, with convex lateral margins to segments and projecting rounded pos- terior angles, furnished with one hair in segments four and five, two in six and seven, one in eighth, and one short one on each tip of the short, indented ninth; a long hair on posterior margins of segments two to six, just within the lateral bands; two short ones in the median portion of posterior margin of segments one to six, and four in the seventh; two in anterior angles of eighth; broad pitchy lateral bands in segments two to seven, separated by clear posterior angles and projecting into anterior adjacent segments; heavy, dark brown, transverse bands on segments one to six, separated by broad clear sutures; clear pustules at spira- cles in segments two to sever, with pale spots running inward from them for a short distance; pale brown, median, ventral bands connect the ends of the dorsal bands and form a dark spot on segment six; tip of ninth pale brown. Mace.—Body, length 1.93 mm., width .81 mm.; head, length -71 mm., width .65 mm.; abdomen much shorter than female and orbicular, with longer, narrower, transverse bands, narrow clear sutures, and the whole median portion obscured by a brown, ven- tral patch; ninth segment larger, protruding with rounded tip furnished with fourteen long hairs; genitalia short, compact, and very dark brown; tip of ninth segment with a crescent-shaped band. . Two males and two females collected on Rhamphastos tocard at Pozo Azul, Costa Rica, June, 1902. A very striking species, easily recognized by the dark conical head, with emarginate clypeus, and by the heavy bands of the abdomen. Nirmus fuscus epustulatus var. noy. Femave.—Body, length 2.15 mm., width .52 mm.; pale yellow golden, with the usual markings of fuscus in the form of heavy complete antennal bands, narrow templar bands, lateral bands on thorax and abdomen, and the heavy, median, transverse bands of the abdomen, . 133 I2 M. A. Carriker Head, length .58 mm., width .44 mm. This variety of fuscus is recognized at a glance by the absence of clear pustules and fewer hairs on the transverse abdominal bands, there being but four on each segment, while fuscus has eight on segments one to five, and six on segments six and seven. With the exception of the smaller size, there are no other appreciable differences. Six females collected on Accipiter bicolor, at Juan Vinas, Costa Rica, March, 1902. Nirmus curvilineatus Kell. and Kuw. Kellogg and Kuwana, Mallophaga from Birds of. the Galapagos Islands. Proc. Wash. Acad. of Sci., vol. IV, p. 470, pl. XXIX, fig. 4 This species, described from specimens collected on Nesopeli« galapagoensis and Oceanites gracilis in the Galapagos Islands, was taken in large numbers from several individuals of Buteo borealis costaricensis, on the volcano Irazu, April, 1902. The taking of this species on a Buieo complicates still more the already confusing state of its distribution. It seems to me that the present host is more typical of the three, and it leaves room for the query as to whether Mr. Kellogg’s specimens might not have straggled from Buteo galapagoensis. My specimens agree perfectly with Mr. Kellogg’s description and plate in every detail. Nirmes atopus Kell. Kellogg, New Mallophaga III, p. 18, pl. I, fig. 4 Numerous males and females collected on Piaya cayana mehleri, one male and female on Myiarchus lawrencei nigricapillus, and one male and female on Stelgidopteryx ruficollis uropygialis, at Juan Vinas, Costa Rica, March, 1902. I am inclined to think that the specimens taken from the last two hosts were stragglers, although there is no direct proof but, since it was described by Mr. Kellogg from a Piaya and I took it in such large numbers on a different variety of the same host species, it is probably confined” to the Cuculidae. 134 ' Mallophaga from Birds of Costa Rica, Central America 13 Nirmus rhamphasti sp. nov., pl. II, fig. 1 Mate.—Body, length 1.40 mm., width .48 mm., short and robust, with pale testaceus head and thorax, heavy antennal bands broken at the suture; abdomen oval, clear, with clear brown lateral bands. = Head, length .44 mm., width .43 mm.; front sharply conical, sides slightly concave, clypeus narrow and emarginate, with one small hair at the suture; trabeculae small, pale golden; antennae short, of median thickness, pale golden, second segment longest ; temples large, expanding laterally and posteriorly, with one hair ; eye prominent, colorless, with a very short bristle; occiput con- vex ; antennal band broad, smoky brown, curving inward slightly at base, inner margin sinuate, broken by clear clypeal suture, beyond which they are pale testaceus; clypeus and oval fossa clear, except a dusky submarginal band across the tip; mandibles heavy, deep chestnut; a small black ocular fleck and a short brownish ocular blotch; region between pale occipital bands clear ; a rather large brownish occipital signature; whole head, except oval fossa and part between occipital bands, an even testaceous. Prothorax quadrangular, sides flatly convex, without hairs; rather broad, brownish, lateral bands, extending around on pos- terior portion ; coxal bands visible, interior of segment same color as head. Metathorax larger, pentagonal, posterior margin angu- lated on abdomen, with four slender hairs on each side and two in the rounded posterior angles; sides flatly convex, widely di- verging; broad lateral brown bands, curving inward slightly in median portion of segment. Legs short and robust, pale golden, narrowly edged with testaceus on anterior margins of femora and tibiae. Abdomen, oval, with round, projecting, posterior angles, bear- ing one hair in segments three and four, two in five to eight, and six on the posterior margin of the ninth; rather heavy, lateral bands of smoky brown in segments one to seven, separated by clear posterior angles, and the portion extending into the adja- cent anterior segment, darker; eighth segment with pale brown spot in lateral portions and ninth with posterior half brownish; a _ median, longitudinal, dusky area, scarcely broken at the sutures 135 14 M. A. Carriker and darker and wider on segments five and six; genital hooks very short, slender, and widely curving; a hair on the posterior margin of segments three to six, just within the lateral bands; four short hairs on the median portion of the posterior margins of segments two to six, and two on the first. Two males collected on Rhamphastos tocard, at Pozo Azul, Costa Rica, June, 1902. This species resembles Piaget’s coniceps more than any other form of the genus. Nirmus hastiformis sp. noyvy., pl. II, fig. 2 FEMALE.—Body, length 1.73 mm., width .53 mm.; uniformly golden brown, with rounded forehead, encircled by the antennal bands, and spear shaped abdomen, having narrow lateral bands, scarcely darker than the heavy transverse bands. Head, length .47 mm., width .45 mm.; front converging, slightly concave sides and evenly rounded clypeus, with a few very short, fine hairs; trabeculae small, pointed, pale golden; antennae of medium size, joints subequal, pale golden; temples slightly expanded laterally and posteriorly, with one long hair and several short bristles; occiput convex; eye small, obscured, with a fine, short bristle; mandibles heavy, bright chestnut; an- tennal bands, scarcely darker than the golden brown temples, encircle the forehead, but are broken by a clear, diagonal line at the clypeal suture; region in front of the mandible clear, as is the portion enclosed by the occipital bands, and also a small area at the bases of the antennae; a black ocular fleck, and a golden brown, spear-shaped occipital signature. ‘ Prothorax quadrangular, broader than long, with sides flatly convex and posterior angles with one hair; lateral bands scarcely darker than the interior of the segment. Metathorax pentagonal, posterior margin obtusely angled on abdomen, with four slender hairs on each side, and two spines in the rounded posterior angles ; sides slightly convex and diverging; lateral bands and interior portion of segment the same color as in the prothorax. Legs stout, though little swollen, clear golden brown. Abdomen rather long, spear-shaped, with rounded, widely pro- jecting, posterior angles, furnished with from one to three hairs in segments four to seven and one in the anterior and posterior 136 Mallophaga from Birds of Costa Rica, Central America 15 angles, and four on the posterior margin of the eighth; narrow, deep golden brown lateral bands on segments one to seven, widely broken by clear posterior angles, and slightly projecting into the adjacent anterior segments; eighth segment very long, as long as the first, with nearly straight, converging sides and completely obscured by the transverse, golden brown band; ninth segment short, and almost transverse posteriorly, uniformly pale golden, with one short hair on each side of the tip; segments one to seven almost completely obscured by deep golden brown, transverse bands, darker medially and in posterior segments, and separated transversely by clear sutures; a clear pustule at the spiracles on segments two to seven and a single long hair on the posterior margins of segments two to six, just within the lateral bands. A single female collected on Trogon caligatus, at Juan Vinas, Costa Rica, March, 1902. This species is easily recognized by the rounded front, with concave sides, completely encircled by the antennal bands; by the length and shape of the eighth seg- ment of the abdomen and the presence of the broad, transverse abdominal bands. Nirmus parabolocybe sp. nov., pl. II, fig. 3 FemMALe.—Body, length 1.66 mm., width .40 rmm., almost entirely obscured by blotches and bands of smoky brown, with narrow pitchy bands on head, thorax, and abdomen; slender, legs short and stout. Head long, with narrowly parabolic front, bare, with a slight colorless protuberance at the tip of the clypeus; trabeculae small, colorless; antennae short, slender, second segment the longest, whole pale brownish; eye small, clear, with bristle; temples nearly square, narrowly margined with pitchy; flatly rounded angles, bearing one slender hair; occipital margin truncate ; nar- row, pitchy, submarginal, antennal bands ending at the clear oral fossa, and with the bases curving diagonally inward at the trabeculae for a distance of one-fourth the width of the head; short, deep brown ocular bands running transversely inward from the eye to base of antennal bands; mandibles small, chestnut; a large smoky brown occipital signature; whole interior of head except oral fossa evenly obscured with smoky brown, 137 16 M. A. Carriker Prothorax, small quadrilateral, anterior margin slightly con- vex, posterior transverse, sides convex; posterior angles with one weak hair; deep, almost pitchy brown, lateral bands, slightly sub- marginal, curving around on posterior portion of segment, but separated medially ; coxal lines show faintly; interior of segment paler than the head; a small clear pustule in the posterior angles. Metathorax much larger than prothorax, pentagonal, sharply angulated on abdomen ; sides convex, diverging ; posterior angles with two slender hairs and two more on each side of posterior margin ; a pitchy spot in the anterior angles, a large pitchy blotch running inward from the posterior angles, narrowing and nearly meeting in center ; remainder of segment same color as head. Legs very short and stout; concolorous with head, margined anteriorly with pitchy and femora semiannulated at tips. Abdomen long, subclavate, pointed abruptly by the last two segments ; posterior angles rounded and protruding with a single weak hair in segments three to six, three in segment seven, and three on each side and two on posterior margin of eighth seg- ment; ninth segment small, colorless, indented at tip; segments one to seven with deep, smoky brown lateral bands having a pitchy hue through the middle and extending anteriorly into adja- cent segments; smoky brown transverse bands on segments one to eight, separated by broad clear sutures and having in segments two to seven, two median longitudinal rows of darker, quadri- lateral blotches, joined on segments six and seven; large clear pustules on segments two to seven, just within lateral bands. Mare.—Body, length 1.34 mm., width .32 mm.; head, length .32 mm., width .25 mm.; differs from the female only in size and in the color and shape of the eighth and ninth abdominal seg- ments, the eighth being clear, except for a narrow brown band around the anterior margin, while the ninth is very small and obscured with brownish; genital hooks small, short, and blunt, being only the length of the eighth abdominal segment. Numerous males and females collected on Muscivora tyrannus and Tyrannus melancholicus, at Juan Vinas, Costa Rica, March, 1902. This is quite a distinct form, though resembling somewhat in shape of head and abdominal markings N. angustifrons Car. 138 ee Mallophaga from Birds of Costa Rica, Central America 17 Nirmus marginellus N., pl. II, fig. 4 Giebel, Insecta Epizoa, p. 147. Piaget, Les Pediculines, Sup., p. 21, pl. III, fig. 1. _ FemMALE.—Body, length 2.13 mm., width .74 mm.; head, length 60. mm., width .60 mm. Mate.—Body, length 1.90 mm., width .72 mm.; head, length .56 mm., width .56 mm. This species is of the interrupto fasciati type, and is distinguished by the shortness or rather by the width of the head and body, the head being an almost perfect equilateral triangle, while the abdomen is a perfect oval, a shape unusual among this group; the head is much darker than the body, being’ a translucent reddish brown, with heavy chestnut antennal bands ; abdomen very pale with reddish fulvous lateral bands and very faint golden transverse bands. Numerous males and females collected from Momotus lessont, at Juan Vinas and Pozo Azul, Costa Rica, March and June, 1902. From Giebel’s descr >tion alone the species is unrecognizable, but Piaget has thoroughly established it from specimens collected. on Prionites momota. My specimens agree quite closely with his descriptions and plate with the exception of the head measurements of the male, my specimen being smaller than the female, while he gives the head measurements the same for the two sexes. It is possible that he may have made an error in this point. Nirmus francisi sp. noy., pl. II, fig. 5. FeMALEe.—Body, length 2.19 mm., width .63 mm.; absolutely’ colorless, with bold markings of pitchy, and smoky brown on head, thorax, and abdomen. Head, length .59 mm., width .52 mm.; sharply conical from trabeculae forward, with sides concave, and the narrow tip emar-' ginate; three short, fine hairs along sides; trabeculae medium, : colorless, equal to the last three, which are deep brown; antennae slender, first two joints longest, eye prominent, colorless, with a short bristle, and slightly obscured by a pitchy ocular blotch; temples almost square, with rounded sides and angles, colorless with one hair in the angle; occipital margin nearly truncate, occi- 139 t8 M.A. Carriker put slightly convex; marginal, antennal bends, broken at the: clypeal suture, run backwards from the sides of the clear oral fossa nearly to the trabeculae, then straight inward to the bases of the posterior roots of the mandibles; the portion in front of the clypeal suture is clear brown, back of that, brown, with a pitchy stripe along inner portion, and pitchy after leaving the margin; occipital bands of deep brown, fading outwardly, and not reaching occipital margin, join the posterior ends of antennal bands ; mandibles chestnut; a smoky brown occipital signature. Prothorax small, quadrilateral, with anterior and posterior mar- gins straight, lateral margins convex, and one weak hair in pos- terior angles; pitchy, lateral, submarginal bands, curving part way around on posterior margin, and having a short projfection into the anterior angles of the metathorax; brown coxal bands show through. Metathorax larger than prothorax, pentagonal, sides convex and widely diverging, with posterior margin sharply angulated on abdomen; a series of six slender hairs on each side of posterior margin; heavy pitchy transverse bands in the region of posterior angles, with a projection into the first segment of abdomen; a narrow, submarginal, chestnut band along posterior margin. Legs short and stout, with pitchy spot at base, and semiannulation at tip of femora, and brownish spot at outer side of tip of tibiae. Abdomen narrowly oval, with rounded, projecting, posterior angles, furnished with one hair in the third segment, and three in four to eight, with two others in anterior angles and four on posterior margin of eighth; ninth segment small, indented at tip, with two chestnut blotches in sides; segments one to eight with narrow, lateral, pitchy bands, separated by clear posterior angles, and projecting slightly into anterior adjacent segments ; two me- dian longitudinal rows of deep, smoky brown blotches in anterior portion of segments, with a latero-posterior projection which reaches the lateral margins of abdomen in segments five to eight ; spots joined in the eighth segment and connected in remainder by a pale brown, ventral, transverse band, which is expanded into a continuous pale blotch on segments six to seven: MaAte.—Body, length 1.63 mm., width .57 mm.; head, length 140 ~ Mallophaga from Birds of Costa Rica, Central America 19 57 mm., width .48 mm.; abdomen shorter, nearer a perfect oval ; ninth segment flatly rounded, with a fringe of long hairs; eighth segment with only a narrow brown band around anterior margin, lateral bands. absent; genital hooks short, stout, and curving at tips. Numerous males and females collected on Zarhynchus wagleri at Juan Vinas, Costa Rica, March, t902. This is a very striking and beautiful form, something like the type of N. ornatissimus Gieb!, a type found on many Icteridae. Nirmus melanacocus sp. nov., pl. II, fig. 6 FEMALE.—Body, length 1.65 mm., width .46 mm., very pale testaceous, almost clear, except the head, which is translucent brownish, with pitchy antennal and templar bands; abdomen clavate, with pitchy lateral bands and brownish, median, trans- verse bands. Head, iength .38 mm., width .39 mm.; almost an equilateral! triangle, with slightly convex sides and truncate occipital mar- gin; clypeus narrowly truncate; trabeculae small, colorless; an- tennae of mediuin size, pale testaceus, second segment the long- est, third and fourth equal; eye flatly convex, colorless, with a small bristle; angles of temples almost square, with one slender hair; heavy pitchy brown antennal bands not quite reaching the clear oral fossa, with the fossa bending inward slightly at the trabeculae; mandibles long and slender, chestnut; temples mar- gined laterally with narrow pitchy bands, connected with the antennal bands by pale brown bands; whole head, except oral fossa and region enclosed between the pale occipital bands, an even clear brown; a large, deep brown occipital signature. Prothorax small, quadrangular, with convex sides and a single weak hair in posterior angles; broad’ pitchy brown lateral bands, curving around on posterior margin; coxal bands visible. Meta- thorax larger, pentagonal, with convex, diverging sides, and sharply angulated posterior margin, having five weak hairs on each side; a pitchy brown spot in anterior angles, connected with lateral bands of prothorax; pitchy brown transverse blotches in region of posterior angles ; coxal lines visible, interior of segment I4I 20 M. A. Carriker clear. Legs short and stout, narrowly margined with brownish on femora, concolorous with head. Abdomen clavate, widest at sixth segment; posterior angies of segments rounded and protruding, with one hair in third, and two in segments four to seven; eighth segment with three hairs on each side; ninth small, clear, and indented at tip; segments one to seven with pitchy lateral bands, slightly paler along outer margins, separated by clear posterior angles and projecting into adjacent anterior segments; segments one to eight with broad, median, transverse bands of pale, smoky brown, separated by clear sutures except between segments six and seven; the band m the eighth segment longer and narrower, nearly reaching the lateral margins of segment. A single female collected on Piranga bidentata sanguinolenta taken on the volcano Irazu, February, 1902. This form resembles N. ampullatus Piag., and is distinguished from that species chiefly by the narrower clypeus, heavier anten- nal bands, and smaller legs. Nirmus pseudophaeus sp. nov., pl. ITI, fig. L FrMALe.—Body, length 2.18 mm., width .60 mm.; almost exactly the shape of N. fuscus, pale translucent, smoky brown, much darker on head; antennal and temporal bands on fee much as fuscus, but jarerel bands of abdomen wanting. Head, length .60 mm., width .44 mm.; shape and markings practically the same as in fuscus, except that they are a smoky brown instead of golden brown, while the interior is paler and also smoky brown; temples with one long hair. Prothorax quadrilateral, with sides slightly sinuate and one short hair in the posterior angles; very pale, indistinct lateral bands. Metathorax larger, with slightly expanded anterior an- gles, sides rounding and diverging, posterior margin truncate, with a short, median, pointed projection on abdomen; posterior angles rounded with one slender hair; four more slender sub- marginal hairs along lateral portion of posterior margin; a pale brown, median, anterior blotch, about in the region of the meso- thorax; a darker band across posterior portion of segment, more 142 rye! ae Sa ’ _Mallophaga from Birds of Costa Rica, Central America 21 intense in region of posterior angles; posterior margin with a clear border. Legs stout, with femora narrowly edged with dusky anteriorly. Abdomen oval, widest at fourth and fifth segments; posterior angles acute, but scarcely projecting, with one hair in third and fourth segments, two in five and six, three in seven, two on each ~side, and two longer pustulated ones on posterior margin of eighth, and a pustulated one in the anterior angles of the ninth; lateral bands absent; clear, pale, brown bands across the posterior portion of segments one to eight, darker, and extending to an- terior margin of segment in the median portion, and separated transversely by broad clear sutures; ninth segment small, rounded, indented, at the tip and pale brown; two long dorsal hairs in the median anterior portion, and a transverse row of six across the median portion of first segment; segments two to seven with a median transverse row of nine slender hairs, all with only very faint, small pustules or none; clear pustules in regions of spiracles, on segments two to seven. A single female collected on Pezopetes capitalis, on the vol- cano Irazu, February, 1902. The finding of this fuscus-like species on a Tanagradae is about as inexplicable as the taking of N. curvilineatus on Neso- pelia galapagoensis, but there was absolutely no chance of strag- gling from a Raptor, since none were killed in that locality on that trip. The present species is easily recognized by the fuscus- like head, the absence of lateral abdominal bands, and the pale continuous transverse bands. Nirmus brachythorax ptiliogonis var. nov. Nirmus brachythorax Gieb., Insecta Epizoa, p. 134. Nirmus brachythorax Piaget, Les Pediculines, p. 150, pl. XII, fig. 8. FEMALE.—Body, length 1.61 mm., width .38 mm. These specimens really come much nearer to Piaget’s N. brachythorax cedrorum than to brachythorax, but since that form is a variety, I will make this another variety, but wil! compare it with the var. cedrorum. 143 22 M. A. Carriker No measurements are given for the variety, but this form is much larger than N. brachythorax cedrorum, having a broader head, occipital margin squarely truncate, temples more nearly square, and angles less rounded; abdomen paler, lateral bands pale fulvous instead of deep brown and median bands barely noticeable. Mare.—Body, length 1.33 mm., width .40 mm.; head, length .33 mm., width .33 mm. Very abundant on Ptiliogonys caudatus, on the volcano Irazu, Costa Rica, February and April, 1902. Nirmus caligineus sp. nov., pl. III, fig. 2 IFEMALE.—Body, length 1.83 mm., width .59 mm.; rather short and: stout, pale golden brown, with darker, smoky brown head, smoky brown lateral and transverse bands on abdomen, separated by clear sutures. Head, length .53 mm., width .51 mm.; sharply conical, with convex sides, bearing four fine, short hairs, and narrow truncate clypeus; trabeculae small, colorless; antennae of medium size, tinged with brownish, second joint the longest; temples broad, narrowly margined with pitchy brown, with evenly rounded an- gles, bearing one long, pustulated hair; eyes small, with a short bristle; occipital margin concave, occiput slightly convex; rather narrow, smoky, antennal bands ending at the clypeal suture; paler internal bands enclosing the oral fossa, running backward from the lateral angles of the clypeus, converging slightly in middle and spreading out in front of the mandibles; mandibles large, chestnut; narrow occipital bands, but slightly darker than the clear brown temples, running from the posterior roots of the mandibles to the occipital margins; bases of antennal bands con- nected with occipital bands by a very faint band; small, semi- clear areas at the bases of the antennae, between the antennal and internal bands, and between the occipital bands; a brownish, oval, occipital signature. ; Prothorax small, quadrangular, with anterior margin concave, posterior flatly rounded, and sides flatly convex and diverging ; posterior angles with one short hair; submarginal, deep brown, 144 + % % Mallophaga from Birds of Costa Rica, Central America 23 lateral bands curving around on the posterior margin; narrow brown lines cutting across from the middle to ends of lateral bands; brown coxal lines visible; interior of segment golden brown, margins smoky brown. Metathorax much larger than prothorax, pentagonal, posterior margin sharply angulated on abdomen, sides nearly straight, widely diverging and posterior angles broadly rounded, with two hairs; five hairs on each side of posterior margin; a pitchy brown spot in anterior angles; a deep brown lateral band, broadening rapidly posteriorly, with a -pitchy projection from posterior angles towards middle of seg- ment; meso-coxal lines visible; interior of segment golden brown. Legs short and stout, concolorous with body. Abdomen elliptical, with rounded, projecting posterior angles bearing one hair in the third segment, two in the fourth, fifth, and seventh, four in sixth, two on each side and four on the posterior margin of eighth; narrow lateral bands of smoky brown on segments one to seven, broken by clear posterior angles and projecting into anterior adjacent segments; broad, brownish transverse bands on segments one to eight, darker on posterior segments, separated by broad, clear sutures and having two me- dian longitudinal rows of larger quadrilateral spots on segments two to seven; eighth continuous; ninth segment small, deeply indented at tip, and with brown spots on each side; a long hair on the posterior margin of segments three to six; a median row of six hairs along posterior margin of segments one to six and two on segment seven. ~ Marre.—Body, length 1.49 mm., width .55 mm.; head, length .49 mm., width .46 mm.; slightly darker than female with the abdomen oval, ninth segment rounded, with about ten fine hairs on posterior margin. One male and one female.collected on Merula grayi, at Juan Vinas, Costa Rica, March, 1902. This form resembles Kel- loge’s N. simplex, but differs principally in the longer and nar- rower clypeus and the duskiness of the head and thorax. 145 24 M. A. Carriker Lipeurus longipes tinami, var. nov., pl. III, fig. 3 Lipeurus longipes Piaget, Les Pediculines, p. 329, pl. XXVIII, fig. 3. MAa.er.—Body, length 1.90 mm., width .35 mm.; head, length .48 mm., width .40 mm, This species is easily recognized by the conical head with truncate front and heavy antennal bands; by the short weak anterior legs and enormously developed middle and posterior pairs, the posterior ones being as long as the abdomen; pro- and metathorax quadrangular, with blackish lateral bands; abdomen slightly clavate, with deep brown lateral bands and pale brown transverse bands. The present variety differs from longipes chiefly in the size and shape of the head, the head measurements for longipes being -44 mm. by .33 mm., while the body is practically the same; this form has the head in front of the antennae wider and shorter, with the front broader and perfectly truncate, while it is emargin- ate in lomgipes. FEMALE (not seen by Piaget).—Body, length 2.15 mm., width .48 mm.; head, length .52 mm., width .44 mm.; whole head darker, especially the temples; antennae scarcely different from the male, except in the absence of the appendage on the third segment and the last three segments being subequal; anterior pair of legs and thorax the same, except a trifle larger; two posterior pairs of legs slightly smaller; abdomen much larger, slightly clavate, but not pointed so abruptly; lateral bands nar- rower and darker; in addition to the continuous pale transverse bands, which are about the same as in the male, are heavy, translucent, brownish, quadrilateral bands, extending inward from the lateral bands a trifle more than one-third the width of the abdomen; on the last segment this band is continuous; apical segment tapering gradually to the slightly indented tip, which is furnished with two long hairs on each side, two short ones at the tip, and two short ones on each side in the anterior portion ; a large median ventral blotch of pale brown extends from the middle portion of segment five to middle of segment eight. Two males and two females collected on Tinamus robustus, at Pozo Azul, Costa Rica, June, 1902. 146 Mallophaga from Birds of Costa Rica, Central America 25 Lipeurus longisetaceus Piag. Piaget, Les Pediculines, Sup., p. 57, pl. VI, fig. 4. MAteE.—Body, length 1.87 mm., width .33 mm.; head, length .40 mm., width .25 mm. FEMALE.—Body, length 2.41 mm., width .32 mm.; head, length .56 mm., width .25 mm. This species, described by Piaget from specimens collected on Tinamus solitarius, is distinguished by its exceedingly slender form and nirmoid appearance; the head is conical with straight sides and rounded front and concave occiput; the clypeus is dis- tinctly set off and pale reddish brown, with two slender fleshy appendages on the dorsal surface; the antennal bands are heavy, while the temples are margined with blackish; mandibles large; whole interior of the head clear; thorax and abdomen with heavy lateral bands, interior clear. Two females and one male collected from Tinamus robustus, at Pozo Azul, Costa Rica, June, 1902. Piaget gives such a clear description and figure that a draw- ing in this paper is unnecessary. Lipeurus postemarginatus, sp. nov., pl. III., fig. 4 FEMALE.—Body, length 2.21 mm., width .59 mm.; clear, elon- gate; head broadly parabolic in front with clear brown temples, antennal bands encircling front and pitchy at base; abdomen spindle shaped, with pitchy brown lateral bands and clear brown transverse bands. Head, length .51 mm., width .41 mm.; slightly conical, broadly parabolic in front, with five short hairs on each side, trabeculae small, clear; antennae rather short and slender, first segment thickest and subequal to second and fifth, third and fourth shorter, equal, the whole pale fulvous; one short hair in front of the trabeculae; eye prominent, colorless, with a short bristle; temples slightly expanded, evenly rounded, with one long hair and several short bristles and narrowly margined with deep smoky brown; occiput concave; whole temple uniformly clear brown inward to the clear occipital bands, and separated from the base of antennal bands by a transverse clear space; mandibles ‘147 26 M. A, Carriker large, chestnut; the broad antennal bands completely encircle the front, broken at the clypeal suture, front part deep fulvous, sides darker, with the rounded bases pitchy; a paler internal band, parallel to the lateral portion of the antennal band; interior of head clear. Prothorax small, quadrilateral, with sides convex and diverg- ing; anterior margin slightly concave, posterior truncate; lateral bands of brown, widening posteriorly, with a pitchy blotch in region of anterior angles; a single hair in the posterior angles; coxal lines showing through. Metathorax much larger than prothorax, quadrilateral, anterior and posterior margins truncate; sides convex, diverging, and slightly sinuate in region of mesothoracic suture; broad lateral bands of deep smoky brown, with a curving pitchy blotch in the median portion of the sides; a large, median, ventral dusk spot; two short hairs in the posterior angles and four very long, stout hairs on each side of the posterior margin, adjacent to the lateral angles. Legs rather small and weak, pale fulvous with a few short hairs. Abdomen large, spindle shaped, widest at the fourth and fifth segments; posterior angles rounded, protruding, furnished with one short hair in segments one to four, one short and one long one in five to seven; eighth segment conical, with a conspicuous indentation at the tip, one long hair in the anterior angles and two large, pustulated hairs in the median portion; narrow pitchy brown, lateral bands broken by clear posterior angles; a clear pustule in the middle cf segments two to seven, just within the lateral bands; clear, pale brown transverse bands, paler on first three segments, extend inward from the lateral bands nearly to the center of the abdomen, broken by broad clear sutures; pale brown ventral bands connect the ends of these heavier dorsal bands; a long hair arises from the posterior margin of the trans- verse bands just inside the lateral bands, on segments three to seven; a row of six short dorsal hairs across the middle of seg- ment five, and two in the middle of segments six and seven; a heavy and continuous transverse band on segment nine, with another paler band encircling the posterior emargination. 148 Mallophaga from Birds of Costa Rica, Central America 27 Mare.—Body, length 1.46 mm., width .38 mm.; head, length -38 mm., width .33 mm.; differs greatly in size from the female, shape of abdomen, size of legs, and antennae. . Head markings the same; first segment of antennae greatly swollen and lengthened, remaining segments subequal, but slightly larger than in female and third having a slight protu- berance; two posterior pairs of legs larger than in female, the posterior pair being as long as the abdomen; abdomen short, thick, almost parallel sided and abruptly pointed by the last three segments, which are much aborted; transverse bands continuous and paler. One male and one female collected from Ortalis cinereiceps at Juan Vinas, Costa Rica, March, 1go2. This form is of the type of mesophelius and intermedius, re- sembling most intermedius, from which it is distinguished by the much shorter metathorax, with truncate posterior margin, by fuller and rounder temples, and by paler abdomen. Lipeurus assesor Gieb., pl. III, fig. 5 Giebel, Insecta epizoa, p. 207. Piaget, Les Pediculines, p. 294, pl. XXIV, fig. 3. FeMALE.—Body, length 3.45 mm., width .63 mm.; head, length .76 mm., width .50 mm. This species is easily distinguished by the long, almost straight sided head, with rounded front, and concave occiput; the peculiar curving marks on each side of the front, the black ocular blotches and dusky temples in strong contrast to the clear interior of the head; the thorax has heavy, lateral, blackish bands, while each segment of the abdomen is furnished with broad, black, lateral blotches and dusky transverse bands. A single specimen collected from Gypagus papa at Pozo Azul, Costa Rica, June, 1902. Although Piaget gives only a drawing of anterior half of head and last two segménts of abdomen, his description is good and, together with Giebel’s, leaves no doubt but what this specimen can be referred to Giebel’s species, 149 28 M. A. Carriker Goniocotes eurysema sp. noy. p’. III, fig. 6 FEMALE.—Body, length 1.93 mm., width .85 mm.; clear ful- vous, with head large, temples produced laterally and angulated behind ; abdomen oval, with pale brown lateral bands and narrow, pitchy, submarginal bands. Head, length .53 mm., width .76 mm.; broadly and rather flatly rounded in front, with slight depressions at the point where the antennal bands touch the margin, trabeculae absent; antennae short, slender, first two segments the longest and equal, last three shorter, equal; temples expanded laterally into a rounded protuberance, bearing two long hairs; posterior portion slightly extended and sharply angulated ; occiput deeply re-entering, con- vex; a broad brownish band running around the front of the head; short, pitchy bands run from the base of the mandibles to the anterior margin at the depression; just within these bands are chestnut-colored protuberances, running a short distance backward from the frontal band; pitchy ocular blotches ; a slightly dusky, narrow band runs around the temples from the eye to the posterior angles; heavy pitchy occipital bands run backward from the posterior roots of the mandibles to the sides of the occiput, thence spreading out into a narrow occipital margin; regions between ocular blotches and antennal bands, dusky brown. Prothorax ‘small, quadrilateral, with anterior and posterior margins slightly concave; sides nearly straight, diverging; pos- terior angles produced backward into a slender rounded protu- berance, bearing one long hair; heavy, pitchy, submarginal bands run backward from the anterior angles to the anterior angles of metathorax ; a portion of the coxal lines showing through; me- dian portion of segment clear, region between lateral bands and margin dusky. Metathorax short, transverse, with lateral mar- gins convex and posterior margin slightly angulated on abdo- men; heavy, curving, pitchy bands running backward from an- terior angles, extend half way across the first segment of the abdomen; lateral margins with two long hairs; a transverse chestnut band across the anterior portion, and another, broken in the center, across the posterior portion of the segment. . Legs short and stout, typical of the genus, with pale fulvous edgings on the anterior faces of femora. 150 Powe - Mallophaga from Birds of Costa Rica, Central America 2g Abdomen oval, slightly clavate; posterior angles rounded, pro- jecting, with from one to three short hairs; segments one to seven with rather broad, smoky fulvous, lateral bands, whose anterior portion extends inward, forming a rounded, backward- curving protuberance, almost obscured by the narrow, pitchy, transverse bands, widening and fading inwardly; eighth seg- ment clear, rounded and indented, bearing one and two short hairs on each side of the tip. Mate.—Body, length 1.38 mm., width .64 mm.; head, length .43 mm., width .62 mm.; differs from female in having long pitchy ocular blotches extending backward from the eye; a shorter, thicker abdomen, more abruptly rounded posteriorly, with the last segment smaller, less protruding, with a fulvous band around the flatly rounded posterior margin; genital hooks long, slender, and almost parallel. Numerous males and females collected on Odontophorus gut- tatus, on the Volcano Irazu, Costa Rica, April, 1902. It is of the general type of G. major Piag., but it is easily distinguished from that species by the heavy occipital bands and the wide diverging prothorax. ORNICHOLAX noy. gen. Body short, compact, head large, and with the general appear- ance of Goniocotes; antennae small, without appendages, and similar in the two sexes; trabeculae large, triangular, movable; prothorax small, short; mesothorax large, broad as head, and separated from metathorax by a very distinct suture; metathorax much narrower than the mesothorax and plainly divided into two lobes by a longitudinal, clear suture; abdomen of both sexes with lateral bands, with but eight segments and with the seventh much aborted; legs very short and stout; dorsal surface of the thorax and abdomen thickly and deeply punctured. Found as yet only on Tinamus robustus, but is.probably com- mon to the Crypturi. Ornicholax robustus n. sp., pl. IX, figs. 1-l¢ Two male specimens and one female taken on Tinamus robus- tus, at Pozo Azul, Costa Rica, June, 1902. This is a strikingly T51 30 M. A. Carrtker different form from anything so far described, enough so, indeed, to make it the type of a new genus. FEMALE.—Body, length 2.85 mm., width 1.35 mm. With the exception of the central portion of the abdomen, it is uniformly pale brown, with darker reddish-brown markings. Head, length 86 mm., width .go mm., somewhat shield-shaped, rounded in front with four short hairs on each side; sides nearly straight; temples obtusely angled, with two longish hairs; occipital mar- gin convex, with two stiff hairs on each side; occiput deeply concave; trabeculae nearly twice the length of the first segment of the antennae, triangular with a darker band along the lateral margin; antennae short, concolorous with head, first segment longest, second and fifth, and third and fourth equal; antennal bands extending in a curve from the anterior point of the trabec- ulae to the base of the mandibles; a narrow, transverse serrated band across the posterior portion of the head; a dark occipital signature between the occipital margin and the transverse occip- ital band; occipital bands extending from the base of the pos- terior root of the mandibles to the transverse occipital! band; a narrow transparent lobe extends along the lateral borders of the temples. Prothorax short, narrow, with lateral angles produced to a blunt protuberance, furnished with a short stiff spine; with the exception of the posterior median portion it is completely encir- cled by a broad, reddish-brown band. Mesothorax broad, lateral portion expanded anteriorly ; lateral angles blunt, furnished with two stout hairs; postero-lateral por- tion slightly concave, with two stout hairs towards the angle; the portion touching the prothorax broadly bordered, while two longitudinal curving bands extend across the segment from the ends of the lateral bands of the prothorax, enclosing a median clear spot; posterior margin transverse, slightly concave. Metathorax scarcely wider than the prothorax, somewhat tri- angular in shape, but completely divided longitudinally by a clear suture; without hairs and with a broad, dark, slightly curving band extending across the anterior portion and con- tinuous with the lateral bands of the abdomen. 152 Mallophaga from Birds of Costa Rica, Central America 31 Abdomen short, oval, with the first segment much wider than any of the others, and the seventh aborted (not extending to the lateral margin of the abdomen) ; lateral margins of the segments convex and the posterior angles of first and second segments with one short hair; the third, fourth, and fifth with two hairs, the sixth with three, and the eighth with a fringe of about twelve hairs on each side; tip of eighth segment slightly indented; a broad submarginal, darker band completely encircling the abdo- men, with the enclosed portion much clearer than the remainder of the body; a stout, slightly pustulated hair on the posterior margin of segments one to five, just inside the lateral band; seventh segment with a fringe cf about twelve short, weak hairs along the posterior margin. Legs short and stout; tarsi aborted ; tibiae almost as large as femora; claws long and stout. The male is slightly smaller than the female, measuring, length 2.41 mm., width 1.14 mm. Head, length .77 mm., width .83 mm. ; the hairs of the temples, thorax, and abdomen longer and stouter than in the female; abdomen more nearly orbicular; seventh segment of the abdomen appearing as two lobes, one on each side of the eighth, which is narrower than in the female, with sides deeply emarginate and tip deeply concave; the fifth seg- ment with three hairs in the posterior angle; the sixth with four, the seventh with six, and the eighth with three short hairs in each angle and four slightly longer, submarginal onés on each side of the middle; the genitalia are simple, curving slightly inward, and about half the length of the eighth segment. KELLOGGIA nov. gen. Body short, compact, and with the general appearance of Gon- iocotes, with the exception of the thorax; head of medium size, thorax smal!; antennae small, without appendages and similar in the two sexes; trabeculae entirely absent; whole thorax small, much smaller than the first segment of the abdomen; meso- and metathorax separated by a distinct suture; metathorax narrower than mesothorax, and completely divided into two lobes by a longitudinal suture; abdomen differing greatly at the tip in the sexes; female with seven segments, male with eight but with 153 32 M. A. Carriker the seventh aborted; lateral bands present in both sexes; legs short and stout; dorsal surface of thorax and abdomen coarsely punctured. Found as yet only on Tinamus robustus. Kelloggia brevipes n. sp., pl. I, figs. 2-2c Two adult females and five adult males taken on Tinamus robustus at Pozo Azul, Costa Rica, June, 1902. FEMALE.—Body, length 2.22 mm., width 1.05 mm.; with the exception of the central portion of the abdomen, it is a uniform testaceous, with darker smoky brown markings on head and thorax. Head, length .73 mm., width .73 mm., triangular and about equilateral, narrowly rounded in front, without hairs; sides of head from antennae to temples perfectly straight, with two short marginal and one submarginal hair; temples rounded with two long, stiff hairs; occiput deeply concave with three short stiff hairs on each side; antennae short, simple, and inserted near the front of the head, first joint largest, second and fifth equal, and third and fourth equal; no trabeculae; mandibles prominent, dark colored; antennal bands running in a slight curve from the base of the mandibles to the frontal margin at the sides, thence faintly around the front of the forehead; occipital bands faint, curving backward from the posterior root of the mandibles to the ends of the prominent occipital signature. Prothorax short, thick, quadrangular, with an emargination at the anterior angles, and the whole segment deeply inserted under the occipital margin, posterior angles drawn out to a blunt spine, armed with a short, stout bristle; margin encircled by a deep band, narrower along the posterior margin; coxal lines showing through. _Mesothorax much broader than prothorax; lateral portions slightly expanded forward and with an emargination just in front of the lateral angles; heavy dark submarginal bands ex- tending longitudinally across each side of the segment, with inner margins curving and extending slightly into the metathorax; lateral angles armed with a long, stout bristle and two stout 154 Mallophaga from Birds of Costa Rica, Central America 33 hairs; the median posterior margin rounded; the postero-lateral margins concave, with two stout hairs; metathorax much nar- rower than mesothorax, extending completely over the first seg- ment of the abdomen ; divided into two distinct lobes by a longi- tudinal suture ; posterior tips of lobes bluntly pointed. Abdomen short, thick, narrowing gradually from first seg- ment to the rounded point; first segment much broader than the others; segments two to six projecting under adjacent anterior segments at their lateral portions ; segments one to six with broad lateral bands, seventh entirely colored, while the median portion is almost entirely uncolored; segments one to five with one long and one short hair in the posterior angles, segment six with three hairs and segment seven with four weak, slender hairs on each side of tip; a long, stout dorsal hair on each side of the posterior margin of segments one to five, nearly to the inner margin of the lateral bands; dorsal surface of thorax and abdo- men strongly, though not closely, punctured. _Legs short and stout, concolorous with body. The male differs slightly from the female, especially in the tip of the abdomen. _ Measurements of male, length 1.76 mm., width .88 mm.; head, length .6 mm., width .65 mm.; the hairs of the thorax and abdomen are longer and stouter; sixth_segment of abdomen very wide (longitudinally), almost as wide as the first; seventh seg- ment much aborted, appearing as two lobes, one on each side of the eighth, which is deeply inserted into the sixth; segments four to seven with three hairs in the posterior angles; eighth with one very long and two short hairs at each angle, and the tips slightly indented. Goniodes minutus sp. nov., pl. IV, figs. 1 and 2 Mate.—Body, length .96 mm., width .4g mm.; short and ro- bust, head large, translucent golden brown throughout, with darker smoky brown frontal band, and lateral bands on thorax and abdomen. - Head, length .28 mm., width .39 mm.; quadrilateral, front broad and flatly rounded; sides nearly straight, slightly diverg- ¢ 155 ae M. A. Carriker ing with three short bristles; temples bluntly rounded, with one long stout bristle and a spine; occiput concave, with two short bristles on each side; a narrow dark brown band around front, with eight small lobate posterior projections; antennae of me- dium size, but first joint greatly swollen, ovoid, second much smaller, third emarginate on one side and pointed, with fourth arising from the emargination, fourth and fifth larger than third, subequal ; mandibles rather small, chestnut. Prothorax large, almost bell-shaped, projecting under the head, with anterior and posterior margins concave, posterior angles rounded, with a short spine; wide lateral bands of reddish brown. Metathorax about as long as prothorax and wider; with broadly rounded posterior margin; anterior angles rounded, with two stout bristles and a spine; three stout hairs in the latero- median portion and two shorter ones on each side of median posterior border; wide lateral bands of reddish brown, coxal bands visible. Legs short and stout, concolorous with body. Abdomen broadly oval, with lateral angles scarcely protrud- ing, and furnished with a stout spine; three stout hairs in the angles of the sixth segment and three on each side of the tip of seventh; two short hairs on each side of the middle portion of the posterior border of segments one and two, and one hair on the posterior margin of segments one to three just within the heavy lateral bands of reddish brown; broad transverse bands of golden brown, separated by narrow, clear sutures, wider be- tween segments four and five and five and six; segment seven entirely the color of lateral bands and deeply inserted into the sixth; genital hooks short, stout, and simple. FrEMALE.—Body, length 1.11 mm., width .52 mm.; head, length .25 mm., width .43 mm.; temples more projecting, occip- ital margin less concave; antennae simple, first, second, and fifth joints longest, subequal; third and fourth shorter, equal; abdomen narrower, apical segment protruding, rounded at tip; markings and color same as in male. Numerous males and females collected on Tinamus robustus at Pozo Azul, Costa Rica, June, 1902. A very distinct species, having little resemblance to any species heretofore described. 150 Mallophaga from Birds of Cesta Rica, Central America 35 Goniodes laticeps Piag., pl. IV, fig. 5 Piaget, Les Pediculines, p. 259, pl. XXI, fig. 6. This striking form is easily recognized by the broad head- shield and the peculiar large posterior tibiae, edged with chestnut and fringed with fine hairs on both sides. The female was not _ seen by Piaget, and a drawing of it is given here. Numerous males and females collected on Tinamus robustus, at Pozo Azul, Costa Rica, June, 1902. Goniodes aberrans sp. nov., pl. IV, figs. 4 and 5 Mare.—Body, length 2.00 mm., width .68 mm.; abdomen spindle shaped, pointed posteriorly, with heavy transverse bands of smoky golden brown; head deeply constricted back of the antennae, and temples enormously developed posteriorly into a slender, almost pointed process. Head, length .57 mm., width .7o mm.; front narrow, flatly rounded, with two median, submarginal hairs; antennae placed at the very front of the head, with first joint greatly swollen, second smaller, truncated-conical, third slender, with fourth aris- ing near its base, fourth and fifth slender subequal; a marked constriction behind the antennae, at which point the head is scarcely wider than the length of the first-segment of the an- tennae; a long, stout bristle just in front of the constriction and a shorter one on the first segment of the antennae; behind the constriction the sides of the head diverge widely with convex margins to the bluntly pointed temporal angles, which are fur- nished with two long, stout bristles; whole posterior margin of head deeply and regularly concave, with two short marginal hairs, and one submarginal on each side of the occiput; a deep chestnut, occipital border, curving forwards at the ends; narrow chestnut antennal bands curving inward from sides of front to bases of mandibles; mandibles small, chestnut; whole head and antennae an even, slightly smoky, golden brown. Prothorax large, flatly dome-shaped, with the whole border anterior to the posterior angles forming a regular, almost hal!- circle; posterior angles almost right angles, furnished with one stout bristle; posterior margin flatly convex; narrow, deeply sub-— 157 io 36 M. A. Carriker marginal, lateral bands of dark brown; whole of lateral regions. slightly darker than median portion, which is smoky golden brown. Metathorax wider and shorter than prothorax, with rounded anterior angles and transverse posterior margin, slightly angulated medially on abdomen; three long stout bristles in the 3 region of the slightly obtuse posterior angles, and one shorter . one farther in towards the middle; narrow, semicircular bands start from the median, anterior portion and pass outward and backward across the segment and half way across first abdominal segment; whole segment uniformly smoky, golden brown. Legs extremely short, femora swollen, but concealed under body, only the short parallel sided tibiae projecting. Abdomen spindle shaped, and pointed posteriorly, with pos- terior angles scarcely projecting and furnished with two stout hairs in segments one to five, sixth with four, and seventh with three on each side of tip; six hairs along the posterior margin of segments one to five; (the ventral portion of the abdomen is fur- nished with hairs very similar to the dorsal portion, and the body is so thin and translucent that care must be taken not to confuse the ventral with the dorsal hairs) all the segments with heavy, smoky, golden brown transverse bands, darker in posterior por- tion and separated transversely by rather broad, clear sutures; seventh (apical) segment deeply inserted into the sixth, with pointed tip and sides slightly concave; genitalia long, with pos- terior half slender and tapering to a point, with the slender portion projecting from abdomen. FEMALE.—Body, length 2.11 mm., width .73 mm.; head, length .63 mm., width .79 mm.; body about the same shape as male; head without the lateral constrictions, front more rounded; an- tennae shorter, simple, and length of joints much as in Nirmus; a larg2, fleshy ovipositor with numerous stout hairs along sides and on tip protrudes from the tip of the abdomen. Two females and one male collected on Tinamus robustus, at Pozo Azul, Costa Rica, June, 1902. This is a very distinct and striking form of Goniodes, and I am a little doubtful as to whether it really belongs there. The spindle shaped abdomen, with continuous transverse bands and the peculiar sexual organs are very aberrant for this genus. 158 Mallophaga from Birds of Costa Rica, Central America 37 Goniodes longipes Piag. Piaget, Les Pediculines, p. 253, pl. XX, fig. 7. A single female specimen of a Goniodes was collected on Odon- tophorus guttatus, volcano Irazu, Costa Rica, April, 1902, in company with several individuals of an undescribed species of the same genus. There cart be no doubt about its being this form, for it agrees exactly with Piaget’s description and plate, ‘although the specimens from which he described it were taken on quite a different host. Laemobothrium delogramma sp. nov., p!. IV, fig. 6 FrEMALE.—Body, length 10.00 mm., width 2.65 mm.; deep smoky brown, with pitchy markings on head and thorax, pitchy edgings on legs and pitchy lateral bands on abdomen; abdomen with a series of dorsal hairs on each segment arising from clear pustules. Head, length 1.84 mm., width 1.65 mm.; slightly conical, trun- cate clypeus and sides straight, interrupted by the antennal swell- ings; palpi projecting by last two segments; temple produced to a blunt point posteriorly; occiput transverse with numerous ‘short bristle-like hairs, two larger ones at the angles; four long and numerous short hairs on the surface of the antennal swell- ings; one hair at the eye; a fringe of short hairs on the sides of the temples; temples with three rather long hairs, two of which are pustulated, and several short ones; a pitchy brown. blotch running back from the angles of clypeus to the antennal swelling and half way around its inner border; a second band curving inward and backward from the posterior portion of the antennal swelling, to the occipital margin; antennal swellings obscured with brown; mandibles large, tipped with black, temples mar- gined with deep brown; a medium oval ventral spot of brown, the unmarked portion of the head clear, pale brown. Prothorax shield-shaped, with prominent lateral angles, antero- lateral portion emarginate, and postero-lateral margin slightly angulated in median portion; anterior margin truncate; lateral angles with a long hair and several short bristles; sides with numerous short hairs; a pitchy spot in anterior angles, from 159 38 M, A, Carriker & which run pitchy bands, parallel with the lateral margins of seg- ment and joining anterior angles of metathorax; a narrow cross band in the anterior portion; a median longitudinal ventral- patch, darker in anterior portion; postero-lateral margins nar- rowly edged with pitchy brown. Metathorax about as long as prothorax, with slightly concave anterior, and truncate posterior, margins; sides very slightly convex, diverging, bordered with numerous fine hairs; two pustulated hairs in the middle of pos- terior margin and one on each side in the middle portion of the segment, just with the lateral bands; whole anterior portion of segment pitchy brown; lateral pitchy bands, narrower anteriorly, join the abdominal bands; a median brownish patch divided longitudinally by a clear narrow line running the entire length of segment; regions between median blotch and lateral bands, clear pale brown. Legs long and stout, with anterior pair smaller, with femora and tibiae all margined with pitchy brown; interior portion clear, obscured in femora by a central patch of brown; femora with a row of pustulated hairs along anterior margin; a few on tibiae. Abdomen large, spindle-shaped, with lateral angles scarcely visible and sides fringed with numerous fine hairs of different lengths, longer posteriorly; posterior margins of segments al- most straight with a transverse series of pustulated hairs of dif- ferent lengths, the shorter hairs arising from smaller pustules ; heavy lateral pitchy bands from segment one to the middle of the ninth, unbroken at the suture, and having in segments’ two to eight three clear pustules, one in the anterior and two in the posterior portion of the segment, from which arise rather long hairs; clear pustules at the spiracles on segments three to eight; inner border of lateral bands rather uneven, fading into a wide, clear, submarginal longitudinal band, on which is a longitudinal row of very fine, short hairs, together with many scattering ones of the same size; median, transverse, deep brown bands on seg- ments one to eight, separated transversely by narrow clear su- tures, and longitudinally by a clear line in segments one and two, and interrupted by clear spots on remainder; sides of trans- verse bands uneven and separated from lateral bands by the clear 160 Mallophaga from Birds of Costa Rica, Central America 39 -submarginal area; tip of ninth segment encircled by ‘a brown band; a median longitudinal deep brown blotch, narrowing pos- teriorly; a row of about six short hairs in median portion of segment, Mate.—Body, length 8.47 mm., width 2.32 mm.; head, length 1.67 mm., width 1.57 mm.; darker than the female, with narrower submarginal clear bands on abdomen, and with the tip having two brown lateral blotches instead of one median blotch. Numerous males and females collected on specimens of Gypa- gus papa at Pozo Azul, Costa Rica, June, 1g02. This species also much resembles Piaget’s Lae. titan, more so than my oligothrix, but can easily be separated from it by the presence of median clear lines and spots on the abdomen, by a wider, submarginal clear area, by the absence of a narrow darker transverse band across the anterior margin of the segments, by the presence of clear pustules in the lateral bands, by the lighter ground color of the thorax, and different bands, and by the more slender tibiae. Laemobothrium oligothrix sp. noy., pl. IV, fig. 7 FEMALE.—Body, length 9.61 mm., width 2.71 mm.; deep smoky brown throughout, with pitchy markings on head and thorax and broad pitchy lateral bands on abdomen. Head, length 1.79 mm., width .76 mm.; sharply conical with the straight sides interrupted by the swellings of the antennal fossae, and the clypeus squarely truncate; palpi small, project- ing by the last two segments; temples produced posteriorly, bluntly pointed ; occiput transverse; clypeus, antennal fossae, and lateral portion of temples with numerous fine hairs; temples with three rather weak hairs, one of which is pustulated; whole head an even, clear brown, with an irregular pitchy band on each side, starting just behind the lateral angles of the clypeus, running backward along inner edge of antennal fossa, separating behind it, with one branch curving outward to margin at eye, thence backward to widest portion of temple, and the other branch pass- ing backward and inward to the occiput, with a break at the submarginal, pitchy, occipital band; mandibles large and heavy, 161 40 M. A: Carriker points deep brown; a crescent-shaped patch between the lateral bands of the head, at the point where they nearest approach each other. Prothorax somewhat shield-shaped, with antero-lateral portion emarginate with some short hairs, postero-lateral portion rounded, with a few short hairs; lateral wings clear, pale brown, whole median portion of segment deep brown; pitchy bands run from the lateral angles to the ends of the lateral bands of the meta- thorax, around on the margin at the lateral emargination and from the anterior portion of the emargination diagonally back- ward to the center of the segment, a longitudinal median ventral band of darker brown. Metathorax about as long as prothorax, with straight diverging sides and truncate posterior margin; sides with a few short hairs; heavy pitchy lateral bands, curving around on anterior portion of segment, continuous with lateral abdominal bands, and with branches running diagonally inward at the posterior portion of the segment and connected by a paler band; interior of segment uniformly deep smoky brown. Legs long and stout, concolorous with body, especially the two pos- terior pairs; anterior femora swollen and orbicular, posterior two lengthened; anterior margin and a portion of posterior margin- of femora and both margins of tibiae heavily edged with black, first joint of tarsi swollen, second long and slender; a series of short pustulated hairs along the inner margin of the black edging of the two posterior pairs of femora; other short hairs on mar- gin of femora and tibiae; femora deeper brown than tibiae and darker at base and in median portions. Abdomen large, elongate oval, with posterior, lateral angles scarcely visible; sides fringed with short slender hairs, heavier and more abundant at posterior portion; posterior margins of segments almost truncate, with a row of eight to ten short hairs on segments two to seven, lateral margins with heavy pitchy bands, narrowing posteriorly and ending at the middle of the last segment; short, pitchy bands run diagonally backward from lateral band in the first segment, corresponding to the bands in_ metathorax ; interior of abdomen an even, deep brown, separated from lateral bands by a narrow pale area; heavier, pitchy brown,_ 162 Mallophaga from Birds of Costa Rica, Ceutral America 41 narrow, transverse bands at the anterior margin of segments two to eight, ending at the pale lateral band; a darker median area in _ segment seven; segment eight translucent, with a deep brown blotch in anterior portion and a clear brown, crescent-shaped band around the tip. Mate.—Body, length 8.29 mm., width 2.17 mm.; head, length 1.65 mm., width 1.56 mm.; similar to female, except in size, and the tip of abdomen, which is slightly swollen iaterally and flatly indented at the tip, with a narrow band around anterior border, a dusky spot in each side of the median portion, and the same band around tip as in female. Numerous males and females collected from» specimens of Buteo borealis costaricensis, shot on the volcano Irazu, April, 1go2. This form is quite close to Piaget’s Lae. titan, but is dis- guished from that species by the clear temples, the absence of a dark band across the middle of the antennal swelling, by the heavier lateral abdominal bands, internal metathoracic bands, and absence of clear pustules on abdomen, and by the difference in the shape of the legs and fewer hairs on margins of abdomen. Physostomum jiminezi sp. noy., pl. V, fig. 1 FeMALE.—Body, length 3.06 mm., width .97 mm.; smoky ful- vus throughout, abdomen with darker lateral bands and median transverse bands; legs very dusky, femora margined with blackish. Head, length .71 mm., width .57 mm.; conical, with flatly rounded, bare front; palettes projecting by apical segment; sides of head slightly sinuated with one weak hair at anterior margin of antennal fossae; a slight ocular notch with one very fine bris- tle; temples produced backward, ending in a blunt spine, very slightly turned outward, with three weak hairs; occipital margin reentering, occiput flatly convex; a pale fulvous band across the forehead at the base of the palettes; a pale indistinct band along the lateral margins of the head, disappearing at the antennal fossae; antennal fossae of unusual size, with inner portion bor- dered with blackish but outer bounded only by a very faint line; a black ocular fleck; a narrow, fulvous, submarginal occipital 163 42 M.A. Carriker band; a slightly darker occipital patch; whole head smoky fulvous. Prothorax unusually large with clear lateral wings; anterior and posterior margins concave; lateral angles very obtuse; an- tero-lateral margins slightly concave, postero-lateral portions straight and diverging; median portion of segment dusky; pale internal bands running back from the lateral margins, near the anterior angles, to metathorax. Metathorax scarcely larger than prothorax, with anterior and posterior margins truncate; sides slightly sinuate, diverging; a faint clear line at the meso-meta- thoracic suture; anterior region of segment, and broad lateral bands of brownish fulvous, continuous with lateral bands of abdomen. Anterior legs short and weak, same color as head; two pos- terior pairs long and stout, with femora margined before and behind with blackish; the whole femora and tibia deep smoky brown. Abdomen narrowly oval, with scarcely protruding acute angles, furnished with one weak hair; posterior margin of first three segments flatly angulated; four and five transverse and six con- cave; broad, marginal, unbroken, lateral bands on segments one to seven, with inner border emarginate on segments four to seven ; eighth segment clear, vulva convex, fringed with fine hairs; in the middle of the broad lateral band is to be found the peculiar chain-shaped band of chitin common to the genus; broad, median, transverse bands of brownish fulvous, separated by narrow, clear sutures, on segments one to seven. Three females collected from Amizillis tzacatl, at Juan Vinas in March and three females from Selasphorus flammula on the volcano Irazu, Costa Rica, February, 1902. This species much resembles microcephalum Kell., but is eas- ily distinguished from that species by the large clear prothorax and the absence of numerous long hairs on the head and thorax. It seems to be a common parasite of several species of hum- mingbirds since it was taken on quite different species in very different localities. 164 bc eel SG Mallophaga from Birds of Costa Rica, Central America 43 Physostomum doratophorum sp. nov., pl. V, fix. 4 FEMALE.—Body, length 2.41 mm., width 1.10 mm.; short and broad, with sides of head deeply emarginate, abdomen broadly oval, whole body pale golden, darker on head, with clear brown markings on head, thorax, and abdomen. Head, length .58 mm., width .61 mm.; front broad, slightly concave, with rounded lateral angles; sides deeply emarginate slightly forward of the middle, and just at the base of the palettes, which are very large, filling the emargination and ex- tending around on the sides in front; palpi projecting by almost entire length, long and stout, apical segment largest, globular ; sides of head behind bases of palpi convex and diverging; a slight ocular notch; temples small, rounded, with two short weak hairs; occipital margin concave, but the qvide occiput strongly convex, with a narrow darker border; a narrow brown band across the forehead just in front of the palettes; anterior mar- gin of palettes bordered by a narrow dark brown band; antennal fossae of medium size, encircled by only narrow lines, and in- terior slightly darker than general color of the head; a black ocular fleck; a narrow, brown, marginal band, encircling the lateral emarginations at the base of the palettes and palpi. Prothorax of medium size, shield-shaped, with anterior mar- gin concave, anterior portion of sides slightly concave, and whole margin back of lateral angles nearly evenly rounded; five short, slender hairs on each side of the rounded lateral portion ; narrow, brown, marginal bands along the portion of the sides anterior to the lateral angles; narrow, pale, internal bands curving back- wards from the anterior angles to the anterior angles of the meta- thorax; two median, longitudinal bands enclosing a spear-head shaped area. Metathorax about as long as prothorax, sides broadly diverging, slightly sinuate, there being a slight lateral constriction at the suture of the meso- and metathorax and a nar- row clear dorsal line; posterior margin flatly rounded; posterior angles rounded and slightly protruding; narrow brown bands from the anterior angles along the lateral margin, broken at the suture, back nearly to the posterior angles, then curving inward and extending half way across the first segment of the abdomen, 165 44 M. A. Carriker where they. join the internal chitin bands; heavy, pale, internai bands run backward from the anterior angles into the abdomen ; the region of the mesothorax and posterior angles deep golden, about the color of the head. Legs long and stout, femora slightly swollen, tibia much swollen at the tips and second segment of tarsi unusually large, being almost globular; femora margined anteriorly and posteriorly, tibiae posteriorly, and tarsi anteriorly with brown. Abdomen short and broadly oval, almost clear, with the excep- tion of the transverse bands; posterior angles scarcely protrud- ing, without hairs; posterior markings transverse; eighth seg- ment flatly rounded behind, with a row of short fine hairs; deeply submarginal, chain-like, chitin bands, extending from the pro- thorax, run almost straight back as far as the middle of the sixth segment; between these bands are faint golden transverse bands, broken by clear sutures; outside, in each segment are spots of the same color as the transverse bands; a few short dorsal hairs on each segment inside and outside of the internal bands. Four females collected from Selasphorus flammula, on the vol- cano Irazu, April, 1902. I have placed this species in the genus Physostomum, but do not think that it rightly belongs there. However, since Mr. Kelloge*has placed a very closely related spe- cies (promineus) in this same genus, I will follow him in the mat- ter at present, though I believe that further collecting of Mallo- phaga from hummingbirds will produce additional species of this type, upon which a new genus can be safely established. Physostomum leptosomum sp. nov., pl. V, fig. 2 FeMALE.—Body, length 3.13 mm., width .82 mm.; head and thorax clear pale brown, with brown markings, abdomen brown- ish golden with deep smoky brown, lateral bands. Head, length .80 mm., width .61 mm.; slightly conical, with front broad and evenly rounded; sides straight with two short fine hairs in front of the antennal fossae; ocular notch small, with three short, fine hairs; temple produced posteriorly and bluntly pointed, with two rather long and one short hair; occip- ital margin deeply reentering, occiput very slightly concaye; the 166 Sree’ he dn or wis Mallophaga from Birds of Costa Rica, Central America 45 protruding palettes faintly tinged with brown; a brownish band across the forehead, joining the bases of the palettes and bear- ing a row of fine, dorsal hairs; antennal fossae clear, encircled by a narrow brown line, except on the inner side where the line expands into a brown band; a brownish, semicircular band curves from the lateral margin of the head at the base of the palettes upward to the transverse band, and between this and the antennal fossae are large irregular shaped brown blotches; a narrow, submarginal, occipital band, blackish in the median portion. Prothorax the same width as the head at the templar angles, quadrilateral, with anterior and posterior margins concave, the anterior half of the lateral margin straight and the posterior half evenly rounded; lateral angles very obtuse (scarcely noticeable), with a short spine and hair, two more hairs on lateral margin behind the lateral angles; broad lateral bands of clear brown and a median, longitudinal, ventral patch of paler brown. Meta- thorax scarcely larger than the prothorax, with anterior portion rounded and covered by prothorax; sides very slightly concave, posterior border truncate and same width as first segment of abdomen; one slender hair in the posterior angles and a spine in the anterior angles; heavy, brown, lateral bands, passing around on the anterior margin; brown bands pass diagonally forward from the posterior angles to the median anterior por- tion, broad and pale at base, narrowing and darkening ante- riorly ; internal portion of segment dusky brown. Legs of me- dium size, pale fulvous, lighter than body. Abdomen rather broad, sides parallel, with the tip abruptly rounded by a portion of the seventh and the eighth segments ; segments of nearly equal width throughout, with transverse pos- terior margins, except the sixth, which is slightly concave; one short hair in the region of the posterior angles of segments one to seven; eighth with a fringe of short hairs along the round margin; deep brown, lateral bands on segments one to seven and anterior portion of eighth, heavier along the inner portion and broken by sharply diagonal, clear lines at the sutures ; inte- rior of the abdomen an even golden brown, with a slightly darker 167 46 M. A. Carriker median patch covering a portion of segment four, all of seg- ments five and six, and a part of seven. Mare.—Body, length 2.43 mm., width .69 mm.; head, length .69 mm., width .55 mm.; differs from the femaie only in smalier size and in last segment of abdomen, which is without the dark lateral bands in anterior portion, is longer, more flatly rounded, and slightly sinuate on the sides; a dusky, median, transverse band in the anterior portion; genital hooks are slender, widely diverging in the median portion, with ends straight and con- verging, and extending the width of the sixth and seventh seg- ments. : Two females collected from Myiarchus lawrencei nigricapillus, and two males from Myiosetetes cayanensis, Juan Vinas, Costa Rica, March, 1902. This form has a superficiai resemblance to P. sucinaceum Kell., but differs in the shape of the head, mark- ings of thorax, presence of palettes, and size. Physostomum angulatum Kell. Kellogg, New Mallophaga I, p. 515, pl. LXX, fig. 5. One female of this well-marked species collected on Tanagra cana, Juan Vinas, Costa Rica, March, 1902. This is another strange instance of distribution, but there can be no doubt as to the identification of the specimen since it agrees perfectly with Mr. Kellogg’s plate and description of angulatum. Physostomum australe Kell. Kellogg, New Mallophaga UW, p. 516, pl. LXX, fig. 4. One female collected on Tanagra cana, at Juan Vinas, Costa Rica, March, 1902. The finding of this species on Tanagra cana is as unaccountable as that of P. angulatwmn, but the identification is absolutely certain. Physostomum subangulatum sp. nov., pl. V, fig. 3 FEMALE.—Body, length 4.37 mm., width 1.06 mm.; almost uncolored, with a faint tinge of golden on abdomen, with nar- 168 9 hh ae , Uy Ue Ae Mallophaga from Birds of Costa Rica, Central America 47 row, pitchy, submarginal, lateral bands on abdomen and thorax and pitchy and deep brown markings on the head. Head, length 1.05 mm., width .80 mm.; conical, with rather narrow, evenly rounded, bare front; large projecting palettes connected by a clear band across clypeal suture; sides of head slightly sinuate in region of antennal fossae; ocular notch shal- low with three short bristles; eyes small, nearly obscured by a black fleck; temples rounded, clear, with three weak hairs, pro- duced backward to a point; occipital margin deeply reentering and evenly concave, with a narrow black submarginal border; a brown occipital signature; antennal fossae small, inner border obscured by a pitchy band fading to brown; region between antennal fossae and lateral margins, and a short space in front, clear brown; a dark brown marginal blotch at the ocular notch; pitchy brown antennal bands run forward from the occipital band, passing along the inner margin of the aniennal fossae, thence forward to the palpi, with two round, darker spots on them, between the fossae and palpi; a short, curving band con- nects their ends with the margins at the base of the palettes; a narrow, curving, longitudinal, brown band runs along the mner borders of the anterior portion of the antennal bands. Prothorax hexagonal, lateral angles»rounded, with a bristle ‘and short hair; antero-lateral margin straight, postero-lateral margin curving, with one weak hair; anterior and posterior mar- gins deeply concave; pitchy lateral bands, marginal in front of lateral angles, submarginal behind, and joining anterior angles of metathorax; two narrow, median, longitudinal, pitchy lines, and two diagonal lines in each side of the posterior portion of segment. Metathorax larger than the prothorax, quadrilateral, with sinuated, diverging lateral margins and truncate posterior margin; posterior angles with one hair and a bristle; strong, -pitchy, submarginal bands running backward from the anterior angles to the posterior margin of segment; a second band start- ing at the anterior angles follows the margin backward to me- dian portion of segment, then curving inward cuts the submar- ginal band and passes forward to the posterior border of the prothorax ; the enclosed portion between the marginal and sub- 169 48 M. A. Carriker i marginal bands pale brown. Legs long, rather slender, and al- most colorless. Abdomen nearly parallel sided, with acute, scarcely projecting posterior angles, having one weak hair in segments one to four, ‘and two in segments five to seven; eighth segment evenly rounded, with two short hairs on each side; vulva convex, fringed with fine hairs; whole abdomen clear, with only faint tinge of golden, excepting the heavy, pitchy, submarginal, lat- eral bands, extending from the end of the metathoracic bands to the middle of the eighth segment, where they are marginal and narrow. Four females collected on Tanagra cana, at Juan Vinas, Costa Rica, March, 1902. This species is a curious combination of P. angulatum and australe Kell., having the head resembling australe and the thoracic and abdominal markings resembling angulatune. It seems quite remarkable that three closely related species of this genus should be found upon the same host, although they were collected on different individuals. As Mr. Kellogg says, this genus is a peculiar one and must be thoroughly revised as soon as sufficient material is accumulated. Physostomum picturatum Car. Carriker, New Mallophaga from Nebraska Birds. Jour. N.Y. Ent. Soee:, vol. XX, 120.4; One female collected on Helminthophila peregrina, Juan Vinas, Costa Rica, March, 1902. The finding of this species on this host in Costa Rica is not so surprising as might at first appear, since it was described from specimens collected on Helminthophila celata. Physostomum pallens Kell. ; Kellogg, New Mallophaga II, p. 49, pl. IV, fig. 7. This species described from specimens collected on Proteno- ‘aria citrea was taken on Compsothlypis pitiayuni, a closely re- lated genus, on the volcano Irazu, Costa Rica, April, 1902. But a single female was taken. 170 i << os eee ae ee ey _Mallophaga from Birds of Costa Rica, Central America 49 | Colpocephalum gypagi sp. nov., pl. VI, fig. 2 Fremace.—Body, length 1.96.mm., width .71 mm.; head smoky brown, with four large pitchy areas; body clear with lateral pitchy spots and paler transverse bands on segments four to seven of the abdomen; femora and tibiae heavily margined an- teriorly with deep brown. Head, length .35 mm., width .53 mm.; flatly rounded in front, with conspicuous ocular emarginations and with the palpi and antennae protruding by apical segment; front with six very fine hairs ; three more, slightly longer, on sides in front of the ocular emargination; ocular fringe rather sparse for the genus; tem- ples slightly angulated in the rear, with three rather long hairs, two marginal and one submarginal; occipital margin deeply concave, with six short hairs; very-large, pitchy ocular and occipital blotches, the ocular blotches connected with the occip- ital blotches by dark brown bands and the occipital blotches connected by a similar band. Prothorax small, oval, with lateral angles produced to a blunt spine, and furnished with a short spine and a long hair; pos- tero-lateral margins with two long hairs; posterior margin trans- verse, with six short hairs; deep brown chitin bands curving across lateral angles, which are brownish, fading inward; trans- verse band pale brown. Metathorax larger than the prothorax, quadrilateral, with widely diverging sides and rounded posterior margin; anterior and posterior angles pale golden, remainder of segment clear; region of posterior angles with numerous short, dorsal hairs. Abdomen large, widest at third segment, thence constricted and tapering sharply to the pointed tip; lateral angles project- ing but little, with several short, weak hairs, and lateral margins with several short hairs also; lateral regions of segments one and two with some short, dorsal hairs; posterior margins of segments one to eight with a fringe of fine hairs; first three segments almost clear, except a brownish spot in lateral portions of segment three; segments four to seven with pitchy blotches in the median portion of lateral margins, and pale brownish transverse bands, separated by clear sutures; segments eight and 171 50 M. A. Carriker nine uniformly brown except the tip of the-ninth, which is clear-} tip of ninth with two longish hairs and a fringe of fine hairs, also a few short hairs along the lateral margins; on the ventral surface is a transverse row of stout hairs across the posterior margin of the eighth segment, and a short row of stiff bristles in the lateral portions; also a fringe of stout marginal hairs along the anterior portion of the eighth, curving upward around the sides of the segment. Legs of medium length and stout, especially the anterior pair; smoky brown the same as the head, with heavy, deep brown anterior borders on the femora and tibiae. A single female collected from Gypagus papa, at Pozo Azul, Costa Rica, June, 1902. This form is of the same type as setosum Piag., from which it is distinguished by the absence of transverse bands on the first two segments of the abdomen, by the much narrower lateral bands on segments four to seven, by the presence of transverse bands on segments ‘four to eight, by the présence of a continuous fringe of hairs on the posterior . margin of all the abdominal segments, and by the much shorter hairs on the posterior angles of the abdomen. Colpocephalum osborni Kell. var. costaricense var nov. A large number of males and females collected from Buteo borealis costaricensis, on the volcano Irazu, Costa Rica, April, 1902. While these specimens resemble quite closely Kellogg’s osborni, there are sufficient important differences to give them a varietal rank, This form is larger, measuring: female, length 1.70 mm., width .56 mm.; head, length .31 mm., width .47 mm. ; male, length 1.57 mm., width .49 mm.; head, length .31 mm., width .46 mm.; the pitchy lateral spots are absent from the first and second segments of the abdomen; there is a dusky trans- verse band across the metathorax; the pitchy spots of the head are not so closely united, while the marginal bands of the legs are paler. With these exceptions it agrees with Mr. Kellogg’s descrip- tion of the species. 172 Mallophaga from Birds of Costa Rica, Central America 51 Colpocephalum extraneum sp. noy., pl. VI, fig. 3 FEMALE.—Body, length 2.15 mm., width .76 mm.; angulated before and behind, legs long and stout, meso- and metathorax divided, metathorax extraordinarily long and shield shaped, and abdomen constricted posteriorly. Head, length .45 mm., width .65 mm.; front very flatly rounded, sides sinuate, there being a depression at the point where the palpi project and at the ocular emargination; front with six short hairs, sides with four; palpi projecting by the ‘long apical segment; an antennae projecting by almost all of last two segments; ocular fringe very thick and long; temples expanded broadly, roundly angulated before and behind, with four long hairs on the lateral margin; occipital margin con- cavo-convex, with a narrow pitchy border; pitchy ocular blotches; mandibles rather small, points pitchy; brownish bands running from end of ocular blotch to frontal margin; a brownish band, curving backward-and broadening medially, connects the anterior portion of ocular blotches; whole temples clear brown. Prothorax hexagonal, with lateral angles produced to a blunt point, furnished with three spines; anterior and antero-lateral margins straight; postero-lateral slightly concave and posterior conyex; posterior angles with one hair, a median ventral blotch in anterior portion of segment; narrow brown edging to whole segment; dark brown coxal bands in the form of a flattened semi-circle across the postero-lateral portion of segments. Mesothorax distinctly divided from metathorax, with convex lateral and truncate posterior margins; a pitchy brown band around sides and anterior angles, broken medially, with short bands running backward from their ends. Metathorax very wide and long, at least three-fourths as long as abdomen; sides straight, diverging; posterior margin elliptical, the region pos- terior to the lateral angles being longer than that anterior to them; lateral-angles very obtuse, with three spines; sides of pos- terior margin with three long hairs; lateral margins with a nar- row pitchy band curving across lateral angles; dark brown bands running inward from posterior portion of lateral margins, al- most meeting medially; a lunate ventral patch in median ante- 173 52)" M. A. Carriker rior portion; two brownish patches in the posterior portion ; coxal outlines, pitchy, showing through (not shown in the plate). Legs long and stout; femora much swollen, especially posterior pair, and all with anterior margins brownish; tibiae swollen api- cally, with anterior and posterior edgings of brown; second joint of tarsi long; the whole concolorous with body, and having nu- merous short marginal hairs; a patch of short hairs on the dor- sal surface of the posterior femora. Abdomen short, scarcely wider than metathorax and con- stricted sharply in the posterior portion; segments subequal in length, with rounded lateral margins; posterior angles with three short spiny hairs; narrow pitchy lateral bands on seg- ments one to eight; deep umber brown transverse bands on seg- ments one to eight, extending inward about one-third the width of the abdomen and scarcely broken at sutures; ninth segment pale clear brown with fringe of fine hairs on the flatly rounded posterior margin; one long hair in posterior angles of the eighth segment and two on each side of ninth, a row of fine, pustulated hairs on posterior margin of transverse bands, except on first segment; median portion of abdomen pale, clear brown. A single female collected on Nyctidromus albicollis, at Pozo Azul, Costa Rica, June, 1902. Colpocephalum Iuroris sp. noy., pl. VI, fig. 4 FEMALE.—Body, length 2.03 mm., width .76 mm.; clear tawny brown, with the large abdomen completely obscured by con- tinuous transverse bands. 3 Head, length .43 mm., width .54 mm.; front very much flat- tened, sides sinuate; ocular emargination large but shallow and - ocular fringe strong; eye large, obscured, protruding from the emargination; temples expanded laterally, angulated behind, rounded before, with three long hairs; occipital margin concavo- convex, with two hairs; front with two hairs in median portion and three at the lateral angles; palpi projecting by half of apical segment; two hairs before the ocular emargination; two long pustulated hairs just within the ocular blotch; an irregular, brownish blotch along interior margin of ocular emargination, 174 Mallophaga from Birds of Costa Rica, Central America 53 and short curving bands from bases of mandibles to lateral an- gles of frontal margin, a pale brown occipital blotch; whole head _ elear tawny brown. Prothorax hexagonal, with expanded lateral angles, furnished with one bristle; coxal bands very plain; a narrow transverse chitin band; posterior margin convex with four hairs. Meta- thorax larger than prothorax, pentagonal, with posterior and anterior margins truncate and sides slightly concave, strongly diverging; a single spine in posterior angles; posterior margin with a row of short, stout hairs; anterior angles and lateral mar- gins edged with dark brown; brown bands (starting at lateral margins) cut across posterior angles into the abdomen. Legs long and stout, especially the front femora, which are much swollen ; concolorous with body. Abdomen large, oval, posterior angles projecting slightly, with two or three spines; a long hair in segments seven and - eight, and two long ones on each side of the ninth, with a fringe of fine hairs between; posterior margins of segments one to seven with a row of short hairs, heavier, tawny, lateral bands on segments one to eight, broken at the sutures; whoie interior of abdomen an even tawny brown. A single female collected on Zarhynchus wagleri, Juan Vinas, Costa Rica, March, 1902. Colpocephalum mirabile sp. nov., pl. VI, fig. 5 FemaLe.—Body, length 1.61 mm., width .66 mm.; clear with bands and markings of brown and pitchy; metathorax enor- mously developed, wider than head and abdomen and two-thirds the length of abdomen. Head, length .34 mm., width .56 mm.; clear, with front. almost evenly rounded except a slight depression at the projecting paipi; front with two long and six short hairs, one just before and three just behind palpi; a longer one pointing backward from anterior portion of ocular emargination ; ocular fringe long and thick; eye very large, clear, with a black fleck; temples ex- panded and nearly evenly rounded, with four long hairs, occip- ital margin deeply reentering, occiput straight; mandibles me- 175 54 . M. A. Carriker dium with chestnut tips; short, brown, curving bands from bases of mandibles to frontal margin; curving pitchy ocular blotches; short, longitudinal crescent-shaped, brown bands inside of ocular blotches. Prothorax large, hexagonal, with anterior margins straight, bluntly rounded lateral angles, concave postero-lateral margins, and rounded posterior margin; lateral angles with three spines; pitchy spots in the anterior angles ; narrow, broken, pitchy brown edgings to the lateral margins; conspicuous pitchy coxal bands, running from the middle of the antero-lateral margins to median portion of segment; a faint median ventral spot. Mesothorax distinctly set off from metathorax, with rounded lateral posterior margins ; mesocoxal bands of pitchy brown very distinct, running from anterior angles of metathorax around the lateral margins of mesothorax, into the posterior angles of pro- thorax; from the ends of lateral bands to middle of segment, then bending sharply back into mesothorax. Metathorax very large (.46 mm. X .62 mm.), clear, with straight diverging sides, very obtusely rounded lateral angles and truncate poste- rior margin; lateral angles with four spines and a very long hair; the posterior portion of lateral margin slightly angulated in the median portion, with one spine; deep brown, dorsal bands, forming a figure 8 across the middle of the segment, with short appendages at the ends on the anterior side and long narrow bands curving backward from the posterior portion of ends, into the abdomen as far as the fourth segment. Legs long and stout, clear, with brown spots at tips of femora and tibiae; numerous short, marginal hairs. Abdomen short, almost parallel sided, with segments five to seven much shorter than remainder, two segments, one, eight, and nine the longest, subequal; sides slightly convex, with a short hair in posterior angles of segments one to eight and a long one in segments one and two, and seven and eight; poste- rior margins of segments one to seven concave, especially five to seven, eight truncate and nine evenly rounded, with two long hairs on each side, a double fringe of very fine hairs at tip, and a submarginal row of short, stout hairs around the whole poste- 176 Bho ay Yibw » Mallophaga from Birds of Costa Rica, Central America | 55 rior margin; a number of short spines on the lateral portion of the posterior margin of segments one to five; short hairs along the posterior margins of segments three to seven, thicker in the portion just inside the lateral bands; irregular pitchy brown spots in the lateral portion of segments one to eight, not as wide as segment, a narrow longitudinal clear line separates these spots from the interior of abdomen; smoky brown transverse bands in segments four to seven, separated by narrow, clear sutures ; irregular, lateral, brown spots in eighth segment; inte- rior of segments one to three, and nine, a pale translucent brown- ish. The abdomen has the appearance of having a flattened lateral area, with the whole central portion convex. Mace.—Body, length 1.54 mm., width .51 mm.; head, length .34 mm., width .50 mm.; clearer than female, with much smaller metathorax ; abdomen a perfect oval, with lateral spots of regu- - lar size and shape in all the segments except the ninth; posterior margins of all the segments with a row of hairs; transverse bands on segments three to six separated by wider sutures; genital hooks very large and long, reaching from third to poste- rior margin of the eighth segment, with the anterior two-thirds a single heavy shaft, widening posteriorly and separating into a perfect trident. One female and three males collected on Zarhynchus wagleri, at Juan Vinas, Costa Rica, March, 1902. Nitzschia bruneri sp. nov. This form is very easily distinguished from pulicaris by the exceedingly short metathorax (length .34 mm., width .71 mm.), by the very slender hind femora and tibiae of the female, and by the paleness of the transverse abdominal bands. Measurements: Female, length 2.50 mm., width .95 mm.; head, length .50 mm., width .66 mm.; male, length 1.96 mm., width .75 mm.; head, length .45 mm., width .6r mm. While working over specimens of Nitzschia from Costa Rica, I again went over the material collected from Aeronautes mela- nolencus in Sioux county, Neb., and which I had named: puli- caris, var. tibialis. I now find that some errors were made at 177 56 M. A. Carriker that time and that several important points were overlooked, which clearly separate this form from pulicaris, and I accord- ingly give it full specific rank as Nitzschia bruneri. Nitzschia bruneri, var. meridionalis var. nov. FEMALE.—Length 2.73 mm., width .g9 mm.; head, length .52 mm., width .76 mm.; male, length 2.18 mm., width .74 mm.; head, length .51 mm., width .7o mm. — yn The variety is distinguished from the species by the darker color, being a translucent brown instead of golden, by the ab- sence of a marginal band on the lateral portion of the meso- thorax, by much darker thoracic and lateral abdominal mark- ings, by more and longer hairs at the posterior angles of the abdomen, by shorter and more spine-like hairs along posterior borders of segments, by much smaller posterior tibiae in the male, and, finally, by a difference in size. Numerous males and females collected from Chactura griset- ventris, at Pozo Azul, Costa Rica, June, 1902. While these specimens closely resemble bruneri, they can be scarcely called that, and have accordingly been given varietal rank. Menopon tridens costaricense var. nov. FEMALE:—Body, length 1.48 mm., width .61 mm.;_ head, length .32 mm., width .4g mm.; whole body uniformly translu- cent fulvous, with black tips to the mandibles, black ocular flecks, narrow blackish occipital margin, while the peculiar, characteris- tic, occipital process is deep brown; occipital margin with four hairs, while the posterior margin of the pro- and metathorax and the abdominal segments is furnished with a’ row of stout hairs; just inside the lateral bands is a longitudinal area covered with short, fine hairs. The rotundity of the abdomen is also a prominent character. A single female collected from Porzana cinereiceps, at Juan Vinas, Costa Rica, March, t902. Unlike the varieties described by Mr. Kellogg, this form has the lateral bands of the abdomen uncolored, as Piaget gives for tridens, but while it agrees with the species in this respect, it differs radically in others. 178 ie P it . aK ne ad - ie”! eA ae 7 i *, Mallophaga from Birds of Costa Rica, Central America 57 - Menopon ortalidis sp. nov., pl. VII, fig. 1 — Mare.—Body, length 1.81 mm., width .71 mm.; very pale throughout, with only a faint tawny tinge; markings of head and thorax and lateral spots of the abdomen light smoky brown. Head, length .37 mm., width .54 mm.; front, beyond ocular emargination, almost evenly rounded, with two short, fine hairs near the middle, two more just back of the projecting palpi, and two longer ones arising just in front of the ocular emargination ; the ocular emargination shallow with a strong fringe of hairs ; the temples slightly drawn out latero-posteriorly, with five long, stout hairs, four of which are pustulated; occipital margin con- cave, bare, with a slight, marginal, dusky band on each side of the occiput; the mandibles small, cinereous, and placed near the front of the head; palpi long and stout, projecting by nearly all of the last two segments ; just behind the base of each palpus is a dark cinereous, cone-shaped protuberance, between which are two short dorsal hairs; the eye is large, clear, with a short hair and partly obscured by a large black fleck on the anterior side; the antennae rather stout, apical joint much the largest, nearly globular, and projecting by a trifle more than half its length; a dark band along the inner border of the ocular depres- sion ; pale smoky, occipital bands, curving from each side of the occiput to the anterior margin of the ocular depression; three short, dorsal hairs along each occipital band. Prothorax very large, nearly as broad as the head and almost hexagonal in shape; lateral lobes expanded, dusky, and lateral angles bluntly rounded, with one short spine; the postero-lateral margins with six long hairs, and the posterior margin with four shorter ones; narrow, dusky bands start from the median por- tion of the antero-lateral margins, curve gently backward nearly to the posterior margin, then bend inward and unite, the whole enclosing a nearly quadrilateral space; fainter narrow bands run diagonally backward from the anterior corners of this quadri- lateral, nearly to the center.of the segment, joining a faint ventral spot; a still fainter band connects these diagonal bands trans- versely. Metathorax about the same size as the prothorax, with straight, diverging sides, and flatly rounded posterior mar- 179 58 M. A. Carriker gin; posterior angles with one long hair and posterior margin with a row of about ten slender hairs; the middle coxae show through as curving brown bands in the region of the anterior angles; pale brown bands cut across the posterior angles from the lateral margins and extend nearly across the first segment ot the abdomen. Legs long and stout, with swollen femora and slender tibiae, and well-developed tarsi; pale throughout, with a few short hairs. Abdomen clear with very pale smoky transverse bands (hardly noticeable) and wide, perfectly clear sutures; lateral angles pro- truding slightly, with one long and two shorter hairs; posterior margin of segments furnished with a row of fine hairs, while a second row runs across the middle of each segment except the first, eighth, and ninth; ninth segment with one long and two shorter hairs on each side of the rounded posterior margin, and some shorter ones between; in the lateral portion of segments. one to eight are smoky brown spots with a darker rounded pro- tuberance in the median portion; genital hooks short, of me- dium size, and typical of the genus. A single male collected on Ortalis cinereiceps, at Juan Vinas, Costa Rica, March, 1902. This species resembles pallescens N., but is much paler, has very slender tarsi, while the markings of the head and thorax differ considerably. Menopon fasciatum Rud., pl. VIII, fig. 4 Rudow, Zettsch. f. d. ges. Nat. XXIV, 403. Giebel, Insecta Epizoa, p. 270. Piaget, Les Pediculines, p. 418. FEeMALE.—Body, length 2.08 mm., width .93 mm.; head length .38 mm., width .69 mm. Mate.—Body, length 2.00 mm., width .84 mm.; head, length 35 mm., width .64 mm. Although the description is vague enough to fit any one of several closely related species, from the fact that my specimens were collected on the same host as Rudow’s, and that what de- scription he does give agrees very well with this specimen, there seems to be no reason for creating another species. 180 q . Mallophaga from Birds of Costa Rica, Central America 59 This form is very readily distinguished by the broad head and the clavate abdomen, with heavy, chestnut, transverse bands and narrow, pitchy, lateral bands. Several males and females collected on Gypagus papa, at Pozo Azul, Costa Rica, June, 1902. Menopon macrocybe sp. nov., pl. VII, fig. 2 FeMALE.—Body, length 1.34 mm., width .48 mm.; head wider than the abdomen; abdonien almost parallel sided, with heavy, transverse, smoky brown bands. Head, length .36 mm., width .57 mm.; very large, somewhat quadrangular, front almost truncate, sides convex and diverg- ing, interrupted by the shallow ocular emarginations; two long and one short hair in front of Ocular emargination; a short sparse ocular fringe; temples rounded, with four long hairs and several short ones; occipital margin deeply reentering, occiput transverse, with two long hairs; mandibles rather large, well toward the front of the head; dark brown antennal bands run diagonally backward from the clypeal angles to the bases of the mandibles, then straight back:to the large pitchy ocular blotch; a black ocular fleck; whole head evenly clear, pale, brown. Prothorax large, lateral angles bluntly pointed, anterior and latero-anterior sides straight; whole margin back of lateral an- gles evenly rounded, with one hair on each side; whole segment clear brown, darker in lateral portions, with a transverse band. Metathorax scarcely larger than prothorax, with straight, widely diverging sides and rounded posterior margin; six long hairs on each side of posterior margin, interior of segment clear brown, with lateral portion deep smoky brown. Legs large and stout, clear pale brown, almost the same color as head, with an- ‘terior margin of femora and both margins of tibiae edged with darker brown; second joint of tarsi very large in posterior pair of legs. Abdomen short, almost parallel sided, abruptly rounded by the large apical segment; segments subequal in length, with lateral angles scarcely visible, furnished with one long hair and one short one in segments one to eight; ninth with one long hair on ISI 60 M. A. Carriker each side of tip and a median fringe of slender hairs; a row of seven or eight short, slender hairs along the posterior margin of segments one to eight; segments one to nine with heavy con- tinuous transverse bands of deep smoky brown, darker in the lateral portion of segments one to three and separated trans- versely by clear sutures except between eighth and ninth. A single female taken on Buteo platypterus, at Juan Vinas, Costa Rica, March, 1902. Easily recognized by the broad head and parallel sided, banded abdomen. . Menopon praecursor meredionale var. nov. FrEMALE.—Body, length 1.73 mm., width .73 mm.; head, length .36 mm., width .54 mm.; differs from praccursor in brown- ish occipital bands curving outward from each side of the occi- put to base-of the antennae, and having a black spot at their base; eight hairs along the occipital margin instead of six; two long hairs in the posterior angles of the prothorax instead of one; posterior margin. very flatly rounded instead of slightly angulated; transverse bands of abdomen narrower and paler; in segments six to eight there are, in addition to the row along the posterior margin, two other rows of short hairs transversely across the segment. Mare.—Body, length 1.48 mm., width .59 mm.; head, length 31 mm., width .52 mm.; paler than the female as in praecursor, but has the transverse bands on the eighth segment of the same intensity as the others. Three males collected from Melanerpes aurifrons hoffmanni, at Juan Vinas, Costa Rica, in March, 1902, and three females and one male from Odontophorus leucolacmus, on the volcano Irazu, April, 1902. Menopon tityrus sp. nov., pl. VIJI., fig. 4 imMALE.—Body, length 1.04 mm., width .52 mm.; very short and broad, deep smoky brown, with wide short head having large pitchy ocular bands; broadly oval abdomen, with heavy transverse smoky brown bands, pitchy in lateral portions. : Head, length .24 mm., width .48 mm.; twice as long as broad, whole margin in front of the bluntly pointed templar angles a 182 INE CR ae oj oo eee Ny Mallophaga from Birds of Costa Rica, Central Ainerica 61 _ slightly flattened arc; whole occipital margin evenly concave with a narrow, pitchy submarginal band, and two median marginal hairs, two short hairs on the front, one at the slightly projecting labial palpi, two long ones and a short one in front of the ocular fleck ; three long marginal ones and several short ones on the pointed temples; a pitchy brown submarginal band around front, broadening at bases into brownish areas covering the whole sides of the head except the paler apical portions of the temples; a large black ocular fleck; a short dark longitudinal band runs forward from bases of mandibles almost to margin of head; two curving bands in occipital region, joining at anterior ends and then extending laterally to the dark portion of the head. Prothorax broad, with produced, bluntly pointed lateral angles, with anterior portion flatly rounded, and very obtuse posterior angles, making nearly straight postero-lateral margins and flatly rounded posterior margin; lateral angles with one long hair and a spine; posterior angles with a long hair, and posterior margin with six slender hairs; whole segment narrowly edged with chestnut ; narrow, lateral, deeply submarginal, and a narrow me- dian transverse band of chestnut; whole segment deep smoky brown. Metathorax scarcely larger than prothorax, with straight, widely diverging sides and flatly rounded posterior margin having row of about ten fine hairs and one in posterior angles ; a lateral emargination at the mesothoracic suture; ante- rior and lateral margins edged with deep chestnut ; segment deep smoky brown paler in mesothoracic region. Legs large and stout, posterior pair largest, with tibiae very large, longer than femora and edged with pitchy. Abdomen broadly oval, with broad, flatly rounded tip; pos- terior angles rather sharp, projecting, with one long and one short hair in segments one to eight; segments nine with several long, and a fringe of fine hairs on the flatly rounded posterior margin, posterior margins of segments one to eight with a row of about twelve to sixteen short hairs; broad continuous, trans- verse bands of deep smoky brown on all the segments, darken- ing to pitchy in the lateral portions of segments one to eight and separated transversely by clear sutures except between segments eight and _ nine. 183 . 62 M. A. Carriker MaAtr.—Body, length .91 mm., width .45 mm.; head, length -24 mm., width .45 mm.; slightly paler than the female, with narrower abdominal bands; apical segments of abdomen same shape as in female; genitalia long, very slender and widely sep- arated with tips curving inward slightly. One male and one female collected on Tityra personata, at: Juan Vinas, Costa Rica, March, 1902. This species approaches M. maestum Kell., but differs greatly in size, shape, and inten- sity of BSE Menopon distinctum Kell. Kellogg, New Mallophaga III, p. 126, pl. VIII, fig. 7. One male and one female of this well-marked species, de- scribed from Myiarchus cinerascens, were collected from M yiar-. chus lawrencei nigricapillus, at Juan Vinas, Costa Rica, March, 1902. Menopon stenodesmumi sp. nov., pl. VIII, fig. 2 FEMALE.—Body, length 1.54 mm., width .60 mm.; clear pale testaceus, with brown and pitchy markings on head and thorax, pitchy brown, lateral bands and clear brown, median transverse bands on abdomen. Head, length .33 mm., width .45 mm.; front rounded, with a depression at the ete ae palpi aaa another at the ocular. emargination, four hairs on front, between palpi; three short ones behind palpi and two longer ones just in front of the ocular emargination; ocular fringe heavy; temples expanding, clear, with four long hairs and a couple of short ones; occipital margin reentering, occiput very slightly convex, with a narrow, pitchy border ; eye large, clear, with a black fleck; an elongated, pitchy, ocular blotch; mandibles large, brown; antennal bands run back ° from frontal margin at palpi, past the bases of the mandibles and along the inner border of the ocular blotches for half their length, then bend abruptly inward and join, forming a backward ° curving band of deep brown across the middle of the head; a large brown occipital signature; antennae project slightly. Prothorax almost hexagonal, lateral angles produced, blunt, with two spines; posteric; angles very obtuse with one long © 184 xy « 4 Mallophaga from Birds of Costa Rica, Central America 63 hair, lateral margin flatly rounded with four short hairs; lateral ‘regions brownish; a median ventral brown spot, with pitchy lines running backward to its posterior angles from the antero- lateral margins. Metathorax much larger than prothorax, clear, bands of brown and pitchy; mesothoracic suture plainly marked; sides, back of suture, straight, diverging, posterior angles rounded, with one hair and three spines; posterior margin flatly rounded with numerous fine hairs; a pitchy band around ante- rior and lateral margin of mesothorax, broken in median portion, with narrow bands running slightly diagonally backward from the ends to the middle of segment; a ventral brown blotch at the _junction of the pro- and mesothorax and a larger wedge-shaped one in the median portion of metathorax; brownish bands rv” - straight backward from margin at mescthoracic suture across the lateral portion of segment and half way across first segment of abdomen, then bend abruptly inward from anterior portion of lateral bands, fading into the median blotch; some brown coxal lines visible, in addition to above. Legs of medium length, clear, with slightly swollen femora and tibiae except the poste- rior tibiae, which are slender and parallel sided; tibiae brownish at tips. Abdomen elliptical, clear, with rounded, projecting, posterior angles, furnished in segments one to seven with three spiny bristles, in eighth with five; ninth segment large, clear, rounded posteriorly, with three long hairs on each side and a fringe of shorter ones between; rather narrow, pitchy lateral bands in segments one to eight; posterior margins of segments with a row of fine hairs; median transverse bands as follows: a cres-: cent-shaped one open behind, extending across portions of first and second segments ; narrow, straight bands, separated by wide, clear sutures, on segments three to six; and a large, somewhat quadrilateral blotch extending from posterior portion of sixth into the anterior portion of the ninth. One female and one male collected on Empidonax atriceps, on the volcano Irazu, April, 1902, and one female on Tanagra pal- merum melanoptera, Juan Vinas, Costa Rica, March, 1902. — This form resembles Col. quadrimaculatum Car. more than 185 64 M. A. Carriker any other, but is distinguished from that species by the larger size, slenderer posterior femora and tibiae, and darker, narrower lateral bands of abdomen. Menopon thoracicum Gieb., pl. VII, fig, 3 FremaLte.—Body, length 1.4 mm., width .54 mm.:; pale ful- vous, with narrow, dusky, occipital margin; two blackish ocular flecks; fuscus markings on head and thorax and deep fuscus lateral bands on abdomen. Head, length .28 mm., width .40 mm.; front rounded, with four short hairs between the projecting palpi and two longer ones in front of the ocular emargination; the emargination dis- — tinct, rather shallow, and with ocular fringe; temples moder- ately expanded, rounded, with four long, slightly pustulated, hairs; occipital margin concave, transverse in center, with two short marginal hairs, and the whole narrowly margined with blackish ; a large, faint, occipital signature, with pale bands cury- ing from its anterior corners to the base of the mandibles, which are small, with dark points; ocular bands indistinct, filling the ocular depression, with a black fleck in the center and another on the large clear eye; a brown spot on the margin just in front of the palpi. . Prothorax with slightly produced, blunt anterior angles, bear- ing three spines, the posterior angles bear one long hair, and the flatly rounded posterior margin four hairs; the chitin bars quite distinct, in the form of slightly flattened semicircles in the region of the anterior angles, a pale transverse line; metathorax with quite a prominent suture setting off the mesothorax, which has an angulated posterior margin and heavy bands on the anterior angles; sides of metathorax straight, widely diverging, and with narrow marginal bands; posterior angles. obtuse, dusky, and with one long hair and three spines; posterior margin flatly rounded, with a complete row of hairs; pale brown coxal bands showing through; legs concolorous with body, bearing a few short hairs and with indistinct marginal markings on tibiae. Abdomen rather large, elliptical, lateral angles serrate, armed with one long hair and several short bristles; posterior margins 186 Mallophaga from Birds of Costa Rica, Central America 65 of segments with a row of longish, slender hairs ; ninth segment large, rounded posteriorly, with two long hairs on each side and a double fringe of very fine hairs on the tip; lateral bands broad, deep, smoky brownish, ending with the eighth segment; very dim, brownish, transverse bands on segments three to eight, separated from lateral bands by a clear place, and from each oiher by clear sutures; ninth segment dusky in lateral portion, tip clear. Matrr.—Body, length 1.00 mm., width .37 mm.; head, length .28 mm., width .37 mm., the head of the male being but slightiy different from the female, while the abdomen is much smaller and slightly darker in color; genital hooks large but simple, re- sembling more the common form of the Colpocephali. Numerous specimens of a Menopon were collected from Ca- tharus gracilirostris, Chlorophonia callophrys, and Piranga bi- denta sanguinolenta on the volcano Irazu, Costa Rica, April, 1902. These specimens can be referred, without doubt, to tho- racicum. Giebel’s description is, for the most part, quite com- prehensive, and every point which he mentions agrees with this form. 188 Mallophaga from Birds of Costa Rica, Central America 67 tion of segment, and extending into first abdominal segment ; whole segment deep smoky brown. Legs stout, with swollen femora and rather long tibiae and tarsi, concolorous with body, with some short hairs. Abdomen large, subelliptical, with one hair and several spines in segments one to eight; ninth segment rounded, with a fringe of hairs in the median portion; posterior margin of segments with a row of fine hairs; narrow, pitchy, lateral bands, broken at the angles; heavy continuous transverse bands of deep smoky brown, separated by clearer sutures on segments one to eight; ninth segment clear, with a brown posterior band and a brownish spot in lateral portions. Matr.—Body, length 1.52 mm., width .57 mm.; head, length .29 mm., width .49 mm.; very similar to female. Numerous males and females collected on Buarremon brun- _ meinuchus, on the volcano Irazu, Costa Rica, February, 1902. This species resembles, in a general way, Piaget’s M. extraneum, but differs greatly in markings of thorax, size of legs, and other - details. Menopon palloris sp. nov., pl. VIF, fig. 3 FEMALE.—Body, length 1.54 mm., width .53 mm.; pale, clear golden, with a slight smoky tinge;'no conspicuous markings, temples bluntly angulated anteriorly and very short, abdomen very hairy. Head, length .34 mm., width .5o mm.; front broad, flatly rounded, with four short hairs; sides sinuate, slightly diverging, with two long and two short hairs; anterior margin of temples almost transverse; a prominent ocular fringe, temples with an- terior and posterior angles (similar to Nitzschia) bluntly rounded, and four long pustulated hairs along the lateral mar- gins; occipital margin concave, occiput slightly convex, with two short hairs; a pale ocular blotch and a black fleck ; mandibles small, chestnut at tips; a pale band over palpi, which project slightly; a narrow, brown, occipital margin. Prothorax hexagonal, lateral angles slightly produced, bluntly rounded, with three spines; anterior, antero- and postero-lateral ia 189 68 ) M. A. Carriker margins concave; posterior margin flatly rounded, with six hairs; one faint coxal band visible on each side. Metathorax larger than prothorax, with almost straight diverging sides and very flatly convex posterior margin, set with a row of fine hairs; mesothoracic suture plainly visible; posterior angles with a long hair and a spine; a pale, lateral, marginal band, interrupted at the mesothoracic suture. Legs long and stout, with swollen femora and tibiae, but short tarsi; a few short hairs on margin. Abdomen subclavate, with lateral margins of segments con- vex, and posterior angles projecting, with one long hair and several stout bristles in segments one to seven; eighth with two long hairs and two bristles, and two long hairs on posterior margin; ninth evenly rounded at tip, with two long hairs on each side and a few short bristles; posterior margins of seg- ments with a series of fine short hairs; two other transverse rows of very fine short hairs across segments three to seven and a single row across two and eight; whole abdomen a uniform, translucent golden, with a slight tawny tinge. The male is slightly smaller, with abdomen somewhat con- stricted posteriorly. A single male and female collected on Steleidopterie ruficollis uropygialis, at Juan Vinas, Costa Rica, March, 1902. Of the type of MW. rusticum Piag. and dissimile Kell., but differs in the shape of the head, and the markings of thorax and abdomen. Menopon Iaticorpus sp. nov., pl. VII, fig. 5 FremMALE.—Body, length 1.40 mm., width .70 mm.; clear brown, with numerous markings and bands of deep smoky brown; abdo- men very large and broad, oval; head broad and very short. Head, length .27 mm., width .58 mm.; front flatly rounded, with rather prominent, though not deep, ocular emarginations ; two rather long hairs in front; labial palpi very long and stout, projecting by fourth and part of third segments; two short hairs and one long one just back of the palpi; two short hairs point- ing backward from front of ocular emargination; ocular fringe very sparse; temples short, produced laterally, and evenly rounded with two long pustulated hairs and several shorter ones, 190 : ead & St —— ifaliophaga from Birds of Costa Rica, Central America 69 e- oceipital margin concave, with four hairs, and a .narrow, mar- ginal, pitchy border; a pitchy brown submarginal band almost ‘entirely around the front of the head; mandible small, pointed ; two broad brown bands starting from the frontal band at the palpi, extend backward past the bases of the mandibles, around the inside of the pitchy ocular blotches, and backward to the occipital margin, their posterior portions spreading out laterally along the temple; lateral portion of temples clear; region out- side of the pitchy, curving, ocular blotches, deep brown; a large black ocular fleck, nearly obscuring the large clear eye; a large deep brown blotch with a clear circular center nearly fills the median occipital region. Prothorax large, with lateral angles produced and_ broadly rounded, with two spines; whole margin posterior to lateral an- gles evenly rounded, with five long hairs on each side; lateral wings deep clear brown; a pitchy band runs inward from lateral angles for a short distance along the antero-lateral margin, then curves backward nearly across the segment; narrow pitchy bands run diagonally backward from the anterior angles to the median portion of segment; a narrow, sinuate, transverse band; a triangular, median brown blotch. Metathorax short, broad, with nearly straight, widely diverging sides, and flatly rounded posterior margin, set with numerous short hairs; posterior an- gles with two longish hairs; pitchy bands around anterior angles, curving backward across segment; heavier pitchy bands running diagonally backward and inward, from anterior angles to middle of segment; narrower pitchy bands curving backward and in- ward from lateral margins, across the segment as far as the middle of segment two of abdomen; a deep smcky brown band across posterior portion of segment. Legs long and stout, with swollen femora, and tibiae slightly enlarged at tips, concolorous with body, and furnished with numerous short margina! hairs. Abdomen very large, a perfect oval, with lateral, posterior angles acute, but scarcely projecting, and furnished with one long hair and a bristle in, segments one to eight; ninth large, rounded behind, with three long hairs on each side and a fringe of fine hairs between; posterior margin of segments one to eight with IgI 70 M. A. Carriker a row of short hairs; segments one to eight with continuous transverse bands of deep clear brown across the pusterior portion of the segments; a large band in ninth, not reaching lateral margins. Mare.—Body, length .86 mm., width .41 mm.; head, length .25 mm., width .51 mm.; differs little from female except in — much smaller abdomen, and slightly smaller head, the legs being nearly as large; genitalia long, very slender, widely separated, with slightly curving tips. A male and female collected on Thamnophilus doliatus, at Juan Vinas, Costa Rica, March, 1902. This is of the same general type as M. maestum Kell. and tityrus sp. nov. LIST OF HOSTS WITH PARASITES Tinamus robustus Menopon ortalidis sp. nov. Docophorus sp.? (juv.) Lipeurus longipes tinami var. nov. - Lipeurus longisetaceus Piag. Ornicholax robustus sp. nov. Kelloggia brevipes sp. nov. Goniodes minutus sp. nov. Goniodes laticeps Piag. Goniodes aberrans sp. nov. Porzana_ cinereiceps Menopon tridens__costari- Guara alba Docophorus bisignatus N. Lipeurus sp.? (juv.) Gypagus papa Lipeurus assesor Gieb. -Laemobothrium —delogram- ma sp. nov. Colpocephalum gypagi sp. nov. Menopon fasciatum Rud. Ghee San HONE Accipiter bicolor Odontophorus guttatus Nirmus fuscus epustulatus Goniocotes eurysema_ sp. — vat RNs nov. Buteo borealis costaricensis Goniodes longipes Piag. Decophorus platystomus N. Odontophorus leucolaemus Nirmus curvilineatus Kell. Menopon praecursor mere- and Kuw. dionale var. nov. Laemobothrium _ oligothrix Ortalis cinereiceps sp. nov. Lipeurus postemarginatus Colpocephalum osborni aus- sp. nov. trale var. nov. 192 of hahha) : mI ey asi b\gh jaan Fs mn Pa oe ee ~ ° ae Mallophaga from Birds of Costa Rica, Central America 71 Buteo abbreviatus Docophorus platystomus N. Buteo platypterus Menopon macrocybe sp. nov. Leucopternus semiplumbea Docophorus platystomus um- _ brosus var. nov. Micraster guerilla Docophorus ultimus sp. noy. Docophorus _ transversifrons sp. nov. Piaya cayana mehleri Nirmus atopus Kell. Rhamphastos tocard Docophorus cancellosus sp. nov. Nirmus rhamphasti sp. nov. Chloronerpes yucatanensis Docophorus __ californiensis Kell. Melanerpes formicivorus Docophorus californiensis Kell. Melanerpes aurifrons hoff- mani Docophorus californiensis Kell. | Menopon praecursor mere- dionale var. nov. Dryobates villosus jardini Docophorus __ californiensis Kell. Trogon caligatus Nirmus hastiformis sp. nov. -Nyctidromus albicollis Colpocephalum extraneum sp. nov. Chaetura griseiventris Nitzschia bruneri meridion- alis var. nov. Amizillis tzacatl Physostomum jiminezi sp. nov. Selasphorus flammula Physostomum jiminezi sp. nov. Physostomum doratophorum sp. nov. Thamnophilus doliatus Menopon laticorpus sp. nov. Tityra personata Menopon tityrus sp. nov. Manacus candaei Docophorus bruneri sp. nov, Muscivora tyrannus Nirmus parabolocybe - sp. nov. Tyrannus melancholicus Nirmus parabolocybe sp. nov. Myiozetetes cayanensis Physostomum leptosomum sp. nov. Myiarchus lawrencei nigrica- pillus ' Nirmus atopus Kell. Physostomum _leptosomum sp. nov. Menopon distinctum Kell. Empidonax atriceps Menopon stenodesmum sp. nov. Momotus lessoni Nirmus marginellus N. 193 ge. M.A. Carriker Psilorhinus mexicanus Docophorus underwoodii sp. nov. Zarhynchus wagleri Nirmus francisi sp. nov. Colpocephalum nov. Colpocephalum mirabile sp. nov. Junco vulcani Docophorus communis N. Acanthadops bairdi Docophorus communis N. Chlorophonia callophrys Docophorus communis N. Menopon thoracicum Gieb. Calospiza guttata chrysophrys Docophorus communis N. Tanagra cana Physostomum Kell. Physostomum australe Kell. Physostomum subangulatum angulatum sp. nov. Menopon thoracicum majus var. nov. Tanagra palmarum melanop- tera Menopon stenodesmum sp. nov. Piranga bidentata sanguino- lenta Nirmus melanacocus _ sp. nov. Menopon thoracicum Gieb. ~ luroris — sp. Ramphocelus passerini_. Menopon thoracicum fuscum. var. nov. Pselliophorus tibialis Docophorus communis N. Pezopetes capitalis Docophorus communis N. Nirmus pseudophaeus — sp. nov. Buarremon brunneinuchus Menopon difficile sp. noy. Stelgidopteryx pygialis Nirmus atopus Kell. _ Menopon pallidoris sp. nov. | Ptiliogonys caudatus Docophorus communis N. Nirmus brachythorax ptilio- gonis var. nov. Helminthophila peregrina Physostomum Car. Compsothlypis pitiayumi Docophorus communis N. Physostomum pallens Kell. Wilsonia pusilla : Docophorus communis N. Catharus gracilirostris Menopon thoracicum Gieb. Merula grayi rubicollis uro- picturatum. Nirmus -caligineus sp. nov. Menopon thoracicum majus. var. nov. Merula nigrescens Docophorus communis N. 194 Mallophaga from Birds of Costa Rica, Central America COLPOCEPHALUM extraneum, 51 gypagi, 49 luroris, 52 niirabile, 53 osborni costaricense, 50 DOCOPHORUS bisignatus, 4 bruneri, 6 californiensis, 6 cancellosus, 10 communis, 9 platystomus, 4 platystomus umbrosus, 4 transversifrons, 5 underwoodi, 8 GONIOCOTES eurysema, 28 GONIODES aberrans, 35 laticeps, 35 longipes, 37 minutus, 33 KELLOGGIA brevipes, 32 LAEMOBOTHRIUM delogramma, 37 oligothryx, 39 LIPEURUS assesor, 27 longisetaceus, 25 longipes tinimi, 24 postemarginatus, 25 MENOPON difficile, 66 distinctum, 62 fasciatum, 58 laticorpus, 68 INDEX macrocybe, 59 ortalidis, 57 palloris, 67 praecursor meridionale, 60 stenodesmum, 62 thoracicum, 64 thoracicum majus, 65 thoracicum fuscum, 65 tityrus, 60 tridens costaricensis, 56 NIRMUS atopus, 12 brachythorax ptiliogonis, 21 caligineus, 22 curvilineatus, 12 francisi, 17 fuscus epustulatus, 11 hastaformis, 14 marginellus, 17 melanococus, 19 parabolocybe, 15 pseudophaeus, 20 rhamphasti, 13 NITZSCHIA bruneri, 55 bruneri meridionalis, 56 ORNICHOLAX robustus, 29 PHYSOSTOMUM angulatum, 46 australe, 46 doratophorum, 43 jiminezi, 41 leptosomum, 44 pallens, 48 picturatum, 48 subangulatum, 46 73 74 He CO bo Re e OOF & be aOonr FDS ee Am Ol me O bo mone ot ® GC bo M. A. Carriker EXPLANATION OF PLATES PLATE I . Docophorus transversifrons sp. nov. . Docophorus bruneri sp. nov. . Docophorus underwoodti sp. nov. . Docophorus cancellosus sp. nov. PLATE II Nirmus rhamphasli sp. nov., male. . Nirmus hastaformis sp. nov., female. . Nirmus parabolocybe sp. nov., female. . Nirmus marginellus sp. nov., female. . Nirmus francist sp. nov., female. . Nirmus melanococus sp. nov., female. PLATE III . Nirmus pseudophaeus sp. nov., female. . Nirmus caligeneus sp. nov., female. . Lipeurus longipes tinami var. nov., male. . Lipeurus postemarginatus sp. nov., female. . Lipeurus assesor Giebel, female. . Gontocotes eurysema sp. nov., female. PLATE IV . Goniodes minutus sp. nov., male. . Gontodes minutus, head of female. . Goniodes laticeps Piaget, female. . Goniodes aberrans sp. nov., male. . Gontodes aberrans, head of female. . Laemobothrium delogramma sp. nov., female, . Laemobothrium oligothrix sp. nov., female. PLATE V . Physostomum jiminezi sp. nov., female. . Physostomum leptosomum sp. nov., female. . Physostomum subangulatum sp. nov., female. . Physostomum doratophorum sp. noy., female, PLATE VI . Colpocephalum gypagi sp. nov., female. . Colpocephalum extraneum sp. nov., female. . Colpocephalum luroris sp. nov., female. . Colpocephalum mirabile sp. nov., female, 196 ’ aH, » Nin ban weg +, ckegicdepicaniel 4‘, gt Si Dill te et teh ee eel oe ‘ - Mallophaga from Birds of Costa Rica, Central America PLATE VII Menopon ortalidis sp. nov., male. Menopon macrocybe sp. nov., female. . Menopon thoracicum Giebel, female. . Menopon tityrus sp. nov., female. Menopon laticorpus sp. nov., female. PLATE VIII op toe . Menopon difficile sp. nov., female. . Menopon stenodesmum sp. nov., female, . Menopon pailoris sp. nov., female. . Menopon fasciatum Rud., female. Hm Oo bo PLATE IX 1-le. Ornicholax robustus sp. nov. 2-2¢. Kelloggia brevipes sp. nov. 197 75 PLATE I PLATE II Pipa eal PLATE IV PLATE. “VW Pie Wil re ) Po . An _ T - re inal yess . PEATE. Vil hp pt PEALE: Vill Pe SN C/ hoy N PLATE IX JF ithe pe 4 vi et Volumes Iand II of UNIVERSITY STUDIES are each complete in four numbers. Index and title-page for each volume is published separately. << ae A list of the papers printed in the first two volumes may be had on application. Single numbers (excepting vol. I, no. 1, and vol. II, no, 3) may be had ips $1.00 each. A few copies of eticeies I and II complete in numbers are still to be had. All communications regarding purchase or exchange should be addressed to. THE UNIVERSITY OF NEBRASKA LIBRARY LINCOLN, NEB., U. S. ye 4ACOB NORTH & CO., PRINTERS, LINCOLN, Vor. III. Juty, 1903 No. 3 UNIVERSITY STUDIES ZaNSURAN IASTITGF — NEA Published by the University of Neb 7 te JUN 29 1934 COMMITTEE OF PUBLICATION Ju 29 J3 | L. A. SHERMAN C. BE, BEN Now, yserd 7 Se ee H.B.WARD W.G.L. TAYLOR H. H. NICHORSON T. L. BOLTON R. E. MORITZ F. M. FLING, EpiTror CONTENTS I. GEORGE SAND AND HER FRENCH STYLE Prosser Hall Frye. 5 : : , 199 Il. Nores ON CERTAIN NEGATIVE VERB CONTRACTIONS IN THE PRESENT Louise Pound ., 2 5 : : : 223 III. ON THE VARIATION AND FUNCTIONAL RELATION OF CERTAIN SENTENCE-CONSTANTS IN STANDARD LITERATURE R.£E. Moritz . i : ‘ ‘ : 229 IV. ON THE ERRORS IN THE METHODS OF MEASURING THE ROTARY POLARIZATION OF ABSORBING SUBSTANCES Fred J. Bates. ; L é , : 255 V. THE MAGNETIC ROTARY DISPERSION OF SOLUTIONS oF ANOMALOUS DISPERSING SUBSTANCES Fred J. Bates. ; ; ; ; ‘ 265 LINCOLN, NEBRASKA Eutered at the post-office in Lincoln, Nebraska, as second-class matter, as University Bulletin, Series 8, No. 11 UNIVERSITY SPUDEES moL. TI] PULY F903 Nor3 I.—George Sand and Her French Style BY PROSSER HALL FRYE Though it is to be feared that the influence which Matthew Arnold? speaks of as exerted by George Sand upon his own youth is exceptional and that as a matter of fact she has never been particularly popular with English readers, yet she certainly ought in justice to be so, for more than any other great French novelist she wrote in the English way. The English judge writing by its spontaneity rather than by its finish. They have hardly been able to understand, at least until very recently, much less to sympathize with the feeling of those French writers, who in assuming the name of artists, have tried to indicate something of the slow, self-conscious elaboration of their processes. To the Englishman writing is a gift, not an art; and he has never been tempted to confound the two. This is the reason that style and construction have counted for so relatively little in the Eng- lish novel. Even so great a novelist as Thackeray has no com- position to speak of; while the fact that a person with George Meredith’s viciousness of expression should have won his repu- tation as an author, illustrates the native English indifference to grace of manner. And yet, to be just, Mr. Meredith does not 1George Sand: the novels, Histoire de ma vie, Correspondance, etc. See also, for original impressions, Caro: George Sand; Flaubert: Lettres a George Sand; des Goncourt: Journal; Sainte Beuve: Portraits contem- porains, note to George Sand; Heine: Lutezia; Matthew Arnold: George Sand; etc. 2George Sand. 199 2 Prosser Hall Frye in this particular suit much better with the English ideal than he does with the French; for the former in its regard ‘for spon- taneity does at least imply a respect for naturalness. The fact is, the English have formed their written upon the — model of their spoken style. They seem, as it were, to assume that their literature is written offhand, and must be judged, even a little indulged, it may be, with this circumstance in mind; as though it were to be expected of an author, not that he should necessarily give long time and thought to his expression, but that he should write quickly and fluently, above all naturally— in short, as though his best possession were the pen of the ready writer. What he has accomplished, then, is to be criticised in accordance with these conditions, not as aiming at perfection, at the expense of unlimited pains and patience, at any cost! On the contrary, the main requirement made of himself by the French writer is that he attain this perfection, which the former has left as unattainable or inconvenient or impertinent—a per- fection absolute and final, which he has always before his eyes as the goal of his aspirations and towards which he strives relentlessly. Time and labor are no object; only that when the work leaves his pen-cramped hand it shall be the best that can be made out of words, the very best without reserve or abate- ment. Ease, or at least the appearance of ease, may be desir- able; not, however, because it is the main purpose of writing to write easily, but because it is a property of elegance that what- ever is done, no matter with what difficulty, should be done too well to show the effort. But diffuseness, approximation, confu- sion, and the like unavoidable accompaniments of conversation- alism and improvisation are forever unpardonable equally with the appearance of stress and strain. While the English write prose with something of the carelessness of talk, the French write prose with the same care that we give to poetry. It is impossible to describe this state of mind better than Mau- passant has done in speaking of an author who stands in every respect in the most striking contrast with George Sand and who represents most characteristically the literary tendencies and ideals, if not the actual performance, of his countrymen—Gus- tave Flaubert. 200 “ea ’ 7 3 : : : ‘ a George Sand and Her French Style a “Haunted by this absolute belief that there exists but one way of expressing a thing, one word to name it, one adjective ‘to qualify it, one verb to animate it, he [Flaubert] would devote himself to superhuman efforts to discover for every phrase that -word, that epithet, that verb. In this way he believed in a mys- terious harmony of expressions, and when a word otherwise suitable seemed to him to lack euphony, he would go on search- ing for another with invincible patience, sure that he had not yet found the true, the unique word. “For him writing was a redoubtable undertaking, full of tor- ment, peril, and weariness. He would seat himself at his table in fear and love of that dear distracting business. ase “Then he would begin to write slowly, stopping again and again, beginning over and over, erasing, interlining, filling the margins, criss-crossing, spoiling twenty pages for one he fin- ished, and groaning with the effort of thought like a wood- sawyer. “Sometimes, tossing his pen into a great oriental pewter tray which he kept full of carefully cut goose quills, he would seize his sheet of paper, raise it to the level of his eyes, and leaning on his elbow, begin to declaim in a loud rasping voice, listening the while to the rhythm of his prose, pausing to catch a fugitive reverberation, combining the tones, separating the assonances, and disposing commas cunningly like resting places on a long road. “A thousand preoccupations would beset him at once, but this desperate certainty always remained fixed in his mind: “Among all these phrases, forms, and turns of expression there is but one phrase, one form, one turn of expression to represent what I want to say.’ “And, red in the face, with swollen cheeks and neck, his mus- cles tense like a straining athlete’s, he would struggle frantically with idea and expression, coupling them in spite of themselves, holding them indissolubly together by the force of his will, grasp- ing the thought and subjugating it little by little with super- 201 4 Prosser Hall Frye human effort and fatigue, and caging it up, like a captive beast, in a solid and exact form.’’? How excessive, but at the same time how indicative in its excess of the writer’s scrupulousness! And while the passion of perfection may not be so virulent with every one of his nation as it was with Flaubert, yet was there ever Englishman, how- ever exceptional, who wrote like this? It is necessary only to compare these remarks with our traditions of Scott’s unfaltering pen and Shakespeare’s unblotted page in order to recognize how different the spirit of French and English prose. This difference of style as between the two nations may be referred, at least in effect, to a variety of causes, the most influ- ential of which are probably these three. In the first place the Englishman has never made so wide a divorce between thinking and writing as has the Frenchman. ~ The former has temperamentally given thought such a decided preeminence over the presentation of thought that he has hardly considered the two as separate at all; but when he has had any- thing to write, has been content simply to think it out in words, and let it go at that. He has always managed to say what he wanted to say, if he has talked long enough; and writing is a sort of soliloquy in which no one can interrupt him. Consider how Browning conducts a poem, like a monologue upon which his readers are licensed for the nonce to eavesdrop, quite wel- come to whatever, if anything, they can manage to pick up. One can, to be sure, put down his book, or throw it away; but his attitude under such circumstances is one of haughty indif- ference—he writes no better. The Frenchman, on the contrary, while thinking, considers that he is in privacy and may be as informal as he likes. In expressing himself, however, he remem- bers that he is in the presence of others, whom he is eager to please and impress—he feels that he must strike and maintain his pose. It is now an affair of manners, and manners maketh the Frenchman. He looks upon his thought as one thing, the presentation of his thought as quite another. And so when he 1 Maupassant, Introduction to Lettses de Gustave Flaubert a George Sand. Compare also his introduction to Pierre et Jean. 202 George Sand and Her French Style 5 comes to write, it is the result rather than the process that he aims to give, and then crystallized in polished sentences which shall have something of the finality of a formula, and forestall i. posterity. When he has once said a thing it is said forever. From this peculiarity of his mind results the importance taken in his literature by epigram. Beside this intellectual difference between the two nations there exists also a difference of language which, though it may be sprung from the former, must be spoken of separately. French words, partly through the influence of the Academy, have comparatively little of that indistinctness or blur of out- line, that sort of emotional penumbra which is so noticeable with English words and to which English poetry owes in great - part its haunting suggestiveness. But they are defined and out- lined, stamped clean to the very edges, covering the ideas upon which they are set with a nicety and exactitude that make French, for all its narrow vocabulary, an ideal instrument of thought, particularly analytic thought. About most English words there is something vague, floating, elusive—something left over to be accounted for after they are applied to the ideas which they symbolize. And this fringe of meaning, which scatters such an iridescent halo about English poetry, makes it necessary in English prose, where such diffraction is an embarrassment, to qualify, limit, and extenuate in order to define the thought with accuracy. But these two conditions, far as they go, are not enough in themselves to explain all the phenomena we have been observ- ing and have still to observe. It is necessary to take account also of .a total difference as between the conceptions of genius held by the two peoples. Genius to the Frenchman means essen- tially an infinite capacity for taking pains—an intelligence cap- able of discerning the nature of the end proposed, of holding it steadily in view, and of applying cunningly and patiently every means at hand to its attainment. Characteristically, the ends of French genius are always rational, attainable by the emi- nently reasonable man—the man, it may almost be said, of com- mon ideas and uncommon energies. To every race genius is 203 6 Prosser Hall Frye the apotheosis it makes of its own best faculty; and intelligence is the Frenchman’s best faculty, as imagination is the English- man’s. “Our literature,’ declares Nisard* in his well-known characterization of the French spirit, “is, as it were, the living - image of this government of the faculties by reason. : This is the spectacle offered us by our masterpieces—they dis- play nothing but a higher reason, sufficiently reinforced by the love of truth to dominate the imagination and the senses and to draw admirable assistance whence ordinarily come the great- est dangers.” From this eminently practical point of view there is nothing absurd in Flaubert’s sitting down with the avowed intention of producing a classic—and succeeding‘ in doing so. While by the very fact his opinion concerning the spirit of the literature, which he knew weil enough to produce a masterpiece in it by malice prepense, takes on a representative character. “Talent,” he declares for his part, and to appreciate the force of the word the reader must remember that it is one maker of chefs-d' euvre coaching another,2 ‘Talent is only long patience. Everything which one desires to express must be looked at with sufficient attention and during a sufficientiy long time to dis- cover in it some aspect which no one has as yet seen or described. In everything there is still some spot unexplored. . . . The smallest object contains something unknown. Find it. To de- scribe a fire that flames, and a tree on a plain, look, keep look- ing, at that flame and that tree till in your eyes they have lost all resemblance to any other tree or any other fire. “This is the way to become original.” To the Englishman, on the contrary, genius signifies some- thing more, at least something other than the free play of intel- ligence. It implies inspiration, as he calls it—the revelation that seems to come down like a sudden light upon life, laying bare its very secrets, transmuting it with new meaning, and pos- sessing the writer, like one beside himself, with an enthusiasm, a power, an eloquence beyond his own. And this capricious, heady, lawless spirit, this emotional transport and exaltation 1fTistoire de la littérature frangatse. 2Introduction to Maupassant’s verre et Jean, translated by Hugh Craig. 204 ep ear ts George Sand and Her French Style z which visits the author without warning and relieves him of the labor of preparation, it is doubtful whether the Frenchman has ever yet quite succeeded in appreciating; whether Shakespeare does not still appear to him under the image of Voltaire’s drunken god adream; as the English have never learned, prob- ably never can learn thoroughly to admire the pale, refrigerated shimmer of Racine. “We are very much mistaken,” cries Zola,* “when we think that the characteristic of a good style is a sub- lime confusion with just a dash of madness in it; in reality the merit of a style depends upon its logic and clearness.” The Frenchman, in short, tends always to subjectivize his emotion and possess it, thereby making his literature objective, while the Englishman tends to objectivize his and to allow it to possess him, thereby making his literature largely subjective. And yet this difference, which is just the difference between art and genius, is the one critically differential of the two litera- tures. Language is at best an inadequate medium, no matter how well handled. And one in accordance with his tempera- ment will prefer the relatively imperfect embodiment of a lofty ideal; and another, the well-rounded embodiment of a relatively low ideal. The former produces a literature of aspiration, in which the whole structure of language is bent and strained by the stress of meaning forced upon it, a romantic literature, strong in poetry and weak in prose, like English. The latter produces a finished and finite literature, neat, elegant, and lim- ited, strong in prose and weak in poetry, a classic literature, like French. For the exuberance of life always tends to shatter and demolish form; and it is only by painful labor, by clipping and paring and pruning that a fresh and modern existence can be forced into vessels and moulds. This is probably something of what Flaubert meant by his celebrated and oft-quoted remark, “The idea springs from the form,”’? a saying so hard for the Englishman, and yet almost a shibboleth to his own disciples. At all events the remark has this much truth: in Goethe’s words, “die Kunst ist nur Gestaltung,” art is only form; and in deter- 1Le Roman expérimental. 2The Goncourts’ Journal. 205 8 Prosser Hall Frye mining his form, in finding what he can or can not put into language without splitting it, the artist does at least determine what his idea shall be. To this general effect George Sand writes to Flaubert:+ “It seems to me that your school doesn’t pay enough attention to the inwardness of things and is too much inclined to rest satisfied with their superficies. As a con- sequence of searching for form you neglect the profundities and address yourself only to the litterati.’” Ay; but he knew that he could not render the profundities without doing violence to the shape and figure of his work, and that he would not do. As Mr. Henry James says, “He had no faith in the power of the moral to offer a surface.’’? For these causes, principal among others, English literature is distinguished from French by its preference, at least in effect, for improvisation and inspiration. And it is for this reason, because these are so exactly the characteristics of her writing, that George Sand deserves the attention of the English reader. “No writer,’ asserts Mr. James, “has produced such great effects with an equal absence of premeditation.”* Her sponta- neity, ease, and fluency; her individuality, sensibility, and in- ventiveness are the positive virtues which most please the En- glish sense; while the vices of their reverse—her diffuseness, confusion, and haziness, her irregularity, extravagance, and wil- fulness, in fine, her lack of discipline—are all defects which the English least notice or most readily excuse. She had no art in the strict sense; but she had inspiration, its virtues and vices, its qualities and defects. The essential truth of this judgment of George Sand has never been disputed by her countrymen or indeed by herself. “She knows,” writes Balzac,* “and said of herself just what I think, without saying it to her, namely, that she has neither force of conception, nor gift of constructing plots, nor faculty of reaching the true, nor the art of pathos, but—without know- ing the French language—she has style; and that is true.’ But \Correspondance, Lettre CVXUIX. 2Essays in London and Elsewhere: Gustave Flaubert. 8French Poets and Novelists: George Sand. ‘Correspondance, March 2, 1838. 206 George Sand and Her French Style 9 in spite of the charm of her writing, almost irresistible in the wooing of the soft slow sentences,’ the inevitable weaknesses of the facility which stood her in place of literary method have been observed over and over, particularly where they are most noticeable, in her construction. Her lack of fundamental plan, of architectural design, has impaired a work that otherwise would have in perfection, as it now has in bulk, few peers. Sen- tences she could write, and chapters, exquisite in touch and feel- ing,—few better; but alas! for all their delicacy, fragments. When it comes to building up piece by piece a single whole, an entire fabric with the subdual of many parts to the perfect har- mony of one great purpose,—there her weakness, the weakness of facility, is manifest. “Le génic,’ she says herself, “vient du ceur et ne réside pas dans la forme;’? and it was her misfor- tune to take her own statement too literally—so literally, indeed, that in Flaubert’s sense she had no form at all. For to Flaubert form meant something more comprehensive than style. “While attaching great importance to observation and analysis, he attached an even greater importance to composition and style. In his opinion it was these two qualities in especial which made a book imperishable. By composition he understood that obsti- nate labor which consists in expressing only the essence of the successive acts of a life, in choosing only the characteristic traits, and in grouping and combining them so that they shall concur perfectly to the effect intended.” * It was not merely his language, then, for which Flaubert was so anxiously concerned in his obstinate wrestlings with expres- sion—it was as well the figure, the shape, the whole concrete plastic embodiment—the Gestaltung—under which he should ex- hibit his conception, at once the emanation and the incorporation of the idea as surely as the pose .of a statue is decisive of the final impression produced, to which the style was to add its par- ticular evocation of sentiment like the music of an opera. This 1Compare Taine’s essay, George Sand, for an appreciation of her style. IHistoire de ma vie. 8 Maupassant’s Z/ude, prefixed to Lettres de Gustave Flaubert & George Sand. 207 IO Prosser Hall Frye was his conception of form, a complete organic whole, a creation in all its parts fatally answerable to the thought of its creator. In the words of Stevenson, who sufféred under much the same infliction of literary conscience as Flaubert :* “For the welter of impression, all forcible but all discreet, which life presents, it [art] substitutes a certain artificial series of impressions, all indeed most feebly represented, but all aiming at the same effect, all eloquent of the same idea, all chiming to- gether like consonant notes in music or like the graduated tints in a good picture. From all its chapters, from all its pages, from all its sentences, the well-written novel echoes and reechoes its one creative and controlling thought; to this must every inci- dent and character contribute; the style must have been pitched. in unison with this; and if there is anywhere a word that looks another way, the book would be stronger, clearer, and (I had almost said) fuller without it. Life is monstrous, infinite, illogi- cal, abrupt, and poignant; a work of art in comparison is neat, finite, self-contained, rational, owing, and emasculate.” It is hardly surprising that of form in this consummate inter- pretation, as the deliberate artist understands it, George Sand should show small sense. With her quick, sensitive, and rather shallow nature she was by no means so likely to distinguish her- self through the manifestation of intellect and will in literature as through the manifestation of sentiment and emotion—not so much in composition as in style. For these, as nearly as they can be discriminated, would seem to be the particular powers of the two.” A Greek tragedy imposed, not by its emotional and sentimental surface-play, but by its deep purposefulness, its se- vere determinism; and so to a lesser degree the drama of Racine, and to some extent all genuinely characteristic French work as compared with English; while a poem of Shelley’s or Tenny- son’s, on the contrary, pleases by the prismatic shimmer of senti- ment with which it is overlaid. The one is typically the affair of composition, the other of style. And toward the latter ex- treme George Sand’s writing naturally gravitates in spite of the 14 Humble Remonstrance. 4Compare Pater’s Hssay on Style. 208 George Sand and Her French Style LE general tradition to which it belongs. It is full of color and feeling, it is splendidly romantic; but when one comes to con- sider it as a whole, to look toward its end and reflect upon its tendency, one is struck by its ineptitude to its purpose. In this respect her work corresponds very closely with the account she herself gives of her own intellectual condition :* “Wisdom,” she remarks very justly, “consists perhaps in classifying one’s impressions, in keeping them from encroach- ing upon one another, and in isolating, if necessary, the particu- lar impression one wishes to receive. In this way arise the great works of genius.’ And of herself: “In order to put an end: to my lack of mental discipline I have prescribed myself a regular life and a daily task—and then two-thirds of the time I lose myself in dreaming or reading or writing something very different from that in which I ought to be absorbed. Had it not been for this intellectual dissipation I should have acquired some sort of an education, for I comprehend readily enough— indeed, if anything, I get to the bottom of things a little too readily; I should have forced my memory to classify its ideas. To understand and to know has been my constant aspiration ; but of what I have wished to realize I have realized nothing. My will has never governed my thought. . . . The external has always acted upon me more than I have acted upon it. 1 have become a mirror from which my own image is obliterated, so completely is it filled with a confused reflection of figures and objects.” These characteristic mental traits of hers show themselves in her writing in several ways. For the careful and consistent reader, one of the most painful experiences is prepared by the frequency with which she falls away in the latter part of her novels from the high standard of her beginnings,—and that not merely in her early work, when she was learning her trade, but in the work of after periods as well, when she had served a long apprenticeship to her art. It is sad to notice, for instance, that M. Faguet speaks of the first volume of a story like the Beaux Messieurs de Bots Doré as a chef-d’euvre and then drops the remainder of it into the oblivion of silence as though in mercy \mpressions et souvenirs. 209 E2 Prosser Hall Frye of its defects. And it is sadder still to find for oneself a book of such fair promise, which might have been completed fault- lessly within the limit of three hundred pages, running on into a wreck of diminishing climaxes and crises and feeble after- thoughts, until it expires tardily of sheer exhaustion, without the needed apology for being so long a-dying, at more than ~ twice its natural age,-—spoiled for no other apparent reason than that the writer wrote too easily to stop when she had finished. Of her might be said what Dryden says of Fletcher: “He is a true Englishman—he knows not when to give over.” It is hardly exaggeration to advise one wishing to read George Ba best work to read only the first halves of her novels. And yet the difficulty were not to be so escaped. This fault of saying too much, this plethora of words occurs again and again over smaller areas than an entire book. With the inveter- acy of disease it infects the whole system. The author is not willing to make the reader a suggestion, to drop him a hint, to risk herself to his perspicacity. She must needs explain—often more for her own sake than for his, it would appear—until there is left over event and motive hardly a single shadow for him to penetrate, but everything lies exposed in an even glare of reve- lation, like the monotonous landscape of our great western prai- rie, without concealment or mystery. There are no skeletons in George Sand’s closet; she has got them all out into the middle of the floor. And her dialogue is as prolix as her analyses. Her characters seem possessed with her own fondness for explica- tion, and invariably talk matters out to a finish, however trivial, so that the reader is constantly outrunning the writer with a sense at the end of disillusion and disappointment. This cir- cumstance is partly accountable for the feeling of commonplace- ness which frequently torments one in his George Sand, even in what he is conscious on reflection are the rarest afercus. The development of her thought is so slow, so gradual, so far fore- seen that her utterances are stamped with none of that surprise which we have come to consider as the hallmark of a profound saying. One is so long prepared that, when the announcement \F tude sur le XIX® siecle. 210 George Sand and Her French Style 13 finally comes, it falls flat on his tired ear like an assertion of the obvious. Perhaps this faultiness, behind which lies always her too ready fluency, may be explained, or at least illustrated, by her manner of work. It is well known nowadays, when the personal habits of authors are more studied than their books, that she wrote at night for certain fixed hours with the regularity of a day-laborer. “She works every night from one to four, and then sets to work again during the day for a couple of hours—and .. . it makes no difference if she’s disturbed. . . . Imagine that you have a faucet open in the house; some one comes in, you close it. . . . That’s the way with Mme. Sand.”* The story goes of her that if she happened to finish the novel on which she was employed an. hour or even less before her time was up for the night, she would calmly set the manuscript away, the ink still damp on the page, and placidly begin another, compos- ing rapidly as she went until the clock released her.2 Whether rightly or wrongly one misses something here—the fond linger- ing over the old work, the patient review and minute revision, the reluctance to part with the child of the brain which makes every finis to the author a lover’s parting and which is so char- actersitic of the French writers of the century. It is another story that is told of Flaubert :* “When he read to his friends the tale entitled, (2 Ceur sim- ple, several remarks and criticisms were passed on a passage of ten lines, in which the old maid ends by confounding her parrot with the Holy Ghost. The idea seemed too subtle for the mind of a peasant. Flaubert listened, reflected, recognized the justice of the observation—but was seized with agony. ‘You're right,’ he said, ‘only—I should have to alter my phrase.’ “That very evening, however, he set to work. He spent the night in changing ten words; he blackened and canceled twenty sheets of paper, and finally left things as they were, unable to construct another phrase whose harmony would satisfy him. “In the beginning of the same tale the final word of a para- Vournal des Goncourt, March 30, 1862. 2/bid., Sept. 14, 1863. , 8’ Maupassant’s E/ude, 14 Prosser Hall Frye graph serving as the subject of the following, might give rise to an amphibology. This distraction was pointed ‘out to him; he recognized it and attempted to change the sense, but could not recover the sonority which he wished for, and, discouraged, ex- claimed: ‘So much the worse for the sense; rhythm before every- thing!” Can there be a more significant contrast than that between these two pictures: Flaubert, the great, rough, positive Norman hesitating irresolutely over a novel for seven years, unable either to perfect or relinquish it; and George Sand, the woman, feeble and timorous, one might suppose, resolutely laying aside one piece of work and taking up another in order to fill out half an hour of scheduled time? By comparison there is something very like gradeur—the grandeur of renunciation, perhaps—in this ability of hers to put away the past when she was done with it, to leave her work to its deserts without just one more backward look, just one more correction, and to pass on confidently to the next duty without worrying over what was gone. “Consuelo,” she writes in reply to a letter of Flaubert’s, “la Comtesse de Rudolstadt, what in the world is that? Can it be something of mine? I have forgotten every last treacherous word of it. Do you read it? Does it really amuse you? In that case I will re- read it one of these days, and if you like me I shall like myself.” It shows at least a self-detachment, a sobriety and moderation not always evident in French literary workmanship of a modern school with its long brooding of the thought—often serving little better purpose than to addle the eggs—and its slow coagulation of the phrase, such as we have come to associate even with Balzac, who would never let his copy go, as Gautier tells us, till it was wrung from him by his implacable taskmaster, the pub- lisher. ? But for all this excess of care we might well wish that George Sand had, without going too far, shown a little more concern for what she had done, a little more for what she was about to do, were it reasonable to suppose that all her errors were due to her habits of work and could have been retrieved by revision. 1Coriespondance. 2Portraits contemporains. 212 - e 7 4 George Sand and Her French Style Is . Much, however, of her defective construction must be charged to another cause. A certain indefiniteness of conception, a failure to decide the end from the beginning and write up to it—in _ short, a powerlessness to fix and realize the idea of a book, is equally a condition of her structural frailty. ‘Descriptions and paintings are no proof that one knows how to write; they prove only that one has strong sensations. What is expected of the writer is the expression of general ideas, and by that he is judged.”* For after all our talk about concreteness and what not, does not every great novel rest finally upon an idea, which the story serves as a specific instance to illustrate? It is difficult perhaps to determine but it is surely legitimate to ask whether the masters have not invariably seen in their fables something wider than the single incident recorded, something standing to that incident in the relation of a general principle to a particular case. “It is not enough to have seen, to have observed; it is essential besides that something general in the case of science, something universally human in the case of art, should be, as it were, engaged in our very observation.”* Certain it is, at all events, that we can not think of a novel in any sense great which does not result for us in some conclusion, much more compre- hensive than the case in point, in regard to human life and con- duct. It may not be expressible in other terms than those par- ticular ones in which the author has rendered it; it may not lend itself to intellectual formularization at all; but there it is in the reader’s mind as the residuum of his reading—the book simple, concrete, and special; the idea complex, abstract, and universal. And it is hardly reasonable to suppose that the writer could have got it thither unless he wrote with it constantly before his eyes. So true is this that the idea a book leaves with us becomes its criterion. “When such a philosophical theme,” so Taine in- sists, “meets a person capable of carrying it to the end and ex- pressing it completely, the novel is of the first order.”* While Lessing makes a similar distinction from the complementary 1Brunetiére: Le Roman expérimental. ?Brunetiére: La Raportage dans le roman, 8 Essay on George Sand. 213 16 Prosser Hall Frye point of view: “To create for a purpose, to imitate for a pur- pose is what distinguishes the genius from the little artist who creates only to create and imitates. only to imitate, who is quite content himself with the minor satisfactions of technique, makes this technique his sole aim, and requires that we also shall be content with just that same sort of minor satisfaction which arises from his artistic but purposeless exercise of his tech- nique.”’* And to the same effect, were it not otiose to do so, it would be possible to cite the criticism of every age which has had a great literature ;? while a lack of sense for this “sorte de lieu commun moral” is an almost infallible sign of critical and literary decadence. For life is a moral affair; and if literature succeeds in its purpose of representing life, its perusal, like ex- perience, will result in the attachment of correct values to human action, not because it is the business of literature to inculcate morals, but because it is the business of literature to represent life, and life is a moral affair. The mere stylist like Gautier is felt to be less than first rate, in spite of the seduction of his man- ner, simply because he has no great ideas of human life to com- memorate. But this is very different from expecting a novel to be written for the promotion of social or religious doctrine or for the ex- ploitation of theories or hypotheses of any kind. To attempt to use literature for such a purpose or to require of it the solution of philosophical problems is evidence of a strange perversion on the part of writer or critic. Philosophies are at best fluctuating and transitory ; they change from generation to generation. The consequences of human action are alone of eternal interest to the human kind. And he who builds beyond the moment must build not upon the former but upon the latter. Nor do such ideas as a rule or as an exception afford a just measure for the evaluation of human life. On‘the contrary, they tend to force life and its expression into narrow, ready-made equations, true enough for the day but by so much the falser for the mor- 1\Hamburgische Dramaturgie, Stick 34. *For instance, Johnson: Rambler, No. 4; Addison: Spectator, No. 70; Dryden: Grounds of Criticism in Tragedy; etc. Indeed the idea has a clear literary pedigree back to Aristotle. 214 ~ George Sand and Her French Style 17 row; in other words, to reduce it temporarily to order by the _ summary process of straight-jacketing it. One attempting the _ representation, or better the interpretation of life, ought to bring _ to its study no preconceived ideas. All such ideas should, where _ they enter literature at all, be strictly distinguished as foreign to its purpose; that is, as extra-literary. They may not always be impertinent or uninteresting, but they are subordinate and ines- sential; and where they rise into prominence and importance _ above the life of the book, they are so,—both impertinent and uninteresting from the point of view of literature. And yet one finds persons enough to read a novel for nothing more than its historical background, or its treatment of a political issue or some other vexed question. In spite of the modern populariza- tion of literature,—perhaps its vulgarization,—one has not ceased to recommend Scott for the historical information to be got out _ of him; or George Eliot for her curious cases of moral casuistry ; or Mrs. Ward for her religious disputations :—clearly literary impertinences in any case and not the vitality that gives these writers their strength. The best training for a novelist is not a system but an expe- - rience—a first-hand knowledge of men and their ways acquired from the give and take of existence, where the hard facts, by dint of battering the consciousness, finally gain recognition. If there is one thing, though but one, for which we are indebted to naturalism, it is the conviction that literature and science are in thus far alike—that both proceed not from speculation but from _ observation. This is the open school in which the novelist learns his lesson, not in the cloisters of a creed. Here he learns of ‘human responsibility, of the consequences of human action, of the fatality of the human will; here he learns “what life and death is; and here finally he gets his ideas of the world direct from the world itself, not in set formulae or generalized pre- scriptions, but embedded in the tissue of individual examples by which he conveys them to others. Literature can never be stud- ied from any mirrored image, not even from literature itself, without distortion or conventionality. Some arrangements of facts he must make, no doubt; but these are not the classifica- 215 - 18 Prosser Hall Frye tions of a rigid system, they are the peculiarities of flesh and blood. Such was George Sand’s training. It is well understood now that she belonged to no sect, accepted no creeds, held no tenets or dogmas, literary or otherwise, which might have controlled her at the outset though at the risk of cramping her early genius. But unfortunately, while she began writing solely from her ex- perience and observation, she began at a moment of violent reac- tion and revolt, when her feelings were still running riot with her reason. And this circumstance imparted to her first work, together with a spirit of reality and naturalness hitherto want- ing to French fiction, a wildness and incoherence that marred the product. The naturalness and reality, for which she had her observation and experience to thank, gave her instant popu- larity, her writer’s capital at the start; while her revolt produced the mental and moral confusion of her first period. . Free of creeds and dogmas as she naturally was, she could have met with nothing more unlucky for the development of her genius than that, almost immediately, and before her liter- ary character was formed, she should have fallen under the in- fluence of those who were essentially theorizers and doctrinaires. An admirer of Rousseau from the first, with an obscure bias in her nature toward a hazy humanitarianism, she devoted the pro- duction of her second period, inspired by her masculine friend- ships and attachments, to the ill-advised attempt to make the novel an instrument of social reform.t No one can doubt that her enthusiasm: over Lamennais’ Christian communism was sin- cere for the moment, but equally so for another moment was her admiration for Pierre Leroux’s socialism, and for still another her interest in free‘masonry. The fact is that these notions for which she was momentarily inspired were never hers by orig- ination and that she never made them so by adoption. The per- sonal weight of those who professed them imposed them upon her feminine susceptibility; and with the artist’s impulsiveness she worked them off upon her novels. Naturally her presenta- 1In his essay on George Sand in French Poets and Novelists Mr. James traces after Taine a very suggestive connection between her ‘‘psychology”’ and her descent. 216 George Sand and Her French Style 19 tion of them was confused and uncertain. And the result is » aes alive much the same with other novels of hers of this and other periods, which are not strictly Tendenz perhaps but may be fairly classed together with the preceding as extra-literary, since they were “not written under purely literary inspiration nor with purely literary motives, and since—the most important test—who reads them reads them primarily for something over and above their litérary interest—for the side-light generally which they throw upon the life, character, or thought of their author or of her time. “I have found,” says Coleridge, and the remark is as true of the novel as it is of poetry, “that where the subject is taken immediately from the author’s personal sensations and experi- ences, the excellence of a particular poem is but an equivocal mark, and often a fallacious pledge, of genuine poetic power.’’* Woman as she was, her feelings when aroused were ever of a vehemence to overbalance her critical judgment; and in writing for the gratification of these feelings rather than from the in- stinct of letters she was likely, no matter at what time of life, to reproduce the emotional confusion of her earliest period. A remark that she herself makes in her memoirs concerning le Piccinino is significant in this connection, and justifies in closing as well as illustrates my use of the word extra-literary as a gen- eral designation for all this kind of work. “Ce que je pense de la noblesse de race, je Vat écrit dans le Piccinino,’’ she says, ‘‘et je nat peut-étre fait ce roman que pour faire les trots chapitres ou jaz développé mon sentiment sur la noblesse.’’* It is often so, too often, in fact, that the purpose of her novels early and late, as she confesses here, is to be sought and found outside of character, situation, and plot. De Musset himself, whatever else he may or may not have stood for, was one of the few exclusively literary ascendencies to which she ever submitted. He it was who awoke her to the existence of such a thing as form and taught her all she ever learned except of herself about style. It is impossible to esti- mate how great was the detriment to her genius that she should have been so_long under influences that, while intellectual, were 1\Biographia Literaria. 2Histoire de ma vie. 217 20 Prosser Hall Frye in no sense literary, and should have been obliged to work her way alone out of much that was harmful to her spirit. Had her flow been less full and copious, it may well be questioned whether the stream would not have choked in the sands of so- ciological and metaphysical discussion with which she was sur- rounded; and she have ended where George Eliot began, as a mere controversialist. It is not a little singular that these two women, the greatest littératrices of their respective countries, should both have been for a time under the dominance of in- spirations other than literary, and should have been more or less diverted from their proper paths and more or less hindered. in their proper activities by philosophical speculation. Of the two George Eliot was more inclined to such thought, and never, indeed, got quite clear of the clutter of erudition, while George Sand was in reality of no great philosophical bent and never assimilated such ideas thoroughly enough to handle them with firmness. As a result of her feeble grasp of.such subjects and of the ‘vivacity of her feelings, she was at her best when she centered her novels neither in a doctrinal motif nor a merely personal emotion, but in some simple episode of common life, which she had noticed and been touched by. Her masterpieces are few in number—as any one’s must be—but they are perfect in their kind :—la Mare au diable; la Petite Fadette, Francois le champi. Les Maitres sonneurs, of the same attempt as the others, errs by excessive development; it overreaches and outruns itself and in spite of much good grows wearisome by its length; while Jeanne and the Meunier d@ Angibault, which are sometimes classed with these, show traces of confusion due partly to the intro- duction of extra-literary ideas and partly to the mixture of idyllic and social elements; so that none of these latter three can be ranked as masterpieces beside the former. Her own dis- trict of Berri, which she always loved and to which she returned more and more in later life, furnished her with the setting for these flawless gems. After the welter of passions and ideas, into which she had been cast young and in which she was long whirled, had subsided, and she could attend to the voice of her 218 ee George Sand and Her French Style 21 own desires; when her love of the unaffected and the natural asserted itself and she had leisure for quiet contemplation in the face of nature ;—then she was quick to recognize and respond to the charm of just such characters and incidents as she met in her Vallé noir of the romantic name, and as she has rendered with exquisite sensibility. The simple, unpretentious life of the peasant amid his fields with his robust loves and hates, hopes and fears was a discovery in comparative humanity to French letters. The healthfulness and freshness of these idyls, full of the air of wood and lawn, the breath of morning and evening, is a revelation after the stale intrigue skulking away in the close and tainted atmosphere of city rooms. They justify to the En- glish reader the existence of French fiction. It may be, as M. Brunetiére declares, that George Sand made the French novel capable of sustaining thought;* it is of infinitely greater credit to her to have shown that it was possible for the French novel to carry good, clean, wholesome sentiment. No reader of mod- ern French fiction can return to these stories without feeling that there life has been triumphantly vindicated against natur- alism, and without feeling, too, that his heart has been purified and gladdened by contact with a great art. Of George Sand’s latest work it is hardly necessary to speak here. Its merits and demerits are essentially and, as far as we are concerned at present, those of the earlier periods. There is noticeable a constantly growing disinclination to air her per- sonal convictions and feelings, together with a marked tendency to rationalize the action, which is quite new, and a very per- ceptible loss of reality, the result, perhaps, of waning enthusi- asm, perhaps of overstudy of the plot, for she could do nothing well that she could not do naturally. It is better to leave her at the moment when her gift for improvisation, the heritage of the born story-teller, which I have tried to show in its strength and weakness, was at its fullest. This impression is certainly the pleasantest of her to carry away, and, what is more impor- tant, it is also the truest; for it is in respect of this quality that she holds among French novelists of the century—eminently a century of novelists—a unique position. Others may have writ- \Manuel de l’ histoire de la littérature frangatse. 219 22 Prosser Hall Frye 6 tae cs ee hth adalah sai ten with equal facility; but no one has written so easily and at the same time so well. Dumas and Hugo in his prose are the only ones who have approached her in point of spontaneity and excellence. That the former is still greatly her inferior is gener- ally recognized, so that it needs here no discussion to prove that his style is of a lower order and that his matter is such that it can impart no imperishable value to his work. With Hugo the case is rather different. He had transcendently the trick of : the phrase ; but catch-words do not make literature. And I think — that as time goes on, whatever his fate as a poet, his fame as a novelist will lower to George Sand’s, if it has ngt actually done so already; because he lacks, at least in prose, the sincerity that alone gives the writer’s utterance weight and authority, while the trace of charlatanry in his novels, as must be the case where the thought waits upon the word, will, when they are farther removed from a fashion to which they have catered, be felt more and more to vitiate his work in this kind. I say nothing of the relative volume of these two authors’ productions, just as it would be to consider such a matter in a question of their comparative spontaneity. And I say nothing of their relative importance to the historical development of fiction, nor urge that his contri- bution to the growth of the novel was inferior to hers, notwith- ’ standing his place poetically in romanticism. I am trying merely to estimate their value for present readers and not their places in the evolution of fiction. And I am even willing to let Hugo ~ pass as an exception to my remark; for whatever may be the weight of the two in technical literary performance, George Sand has, I think, a message for the present day to which Hugo can not pretend. To none of George Sand’s serenity can Victor ~— Hugo, or any other modern French writer that I know of, lay claim. In spite of the spiritual turmoil of her period, in, spite of her personal difficulties, her work at its best is eminently . serene. It possesses in a high degree the twin characters of all work that is great and sane,—simplicity and serenity. And this is assuredly the wisest lesson that can be read to our own two vices of extreme at this moment,—to our impateince and our intemperance. To the vague trouble, the haunting disquietude 2 220 - George Sand and Her French Style 23 that disfgures almost all our work to-day, not only our literary ‘and artistic work but our work of every kind particularly here in America, her writing offers the best contrast and correction. If there are two qualities that we lack just now, they are the qualities of patience on the one hand and of moderation on the other ;—the patience to await results and to labor honestly for them, and the moderation to be satisfied with a fair day’s labor and a fair wage for it. For either we neglect our task, shuffling it hastily aside to turn to something new, or else we are mas- tered by it and become its slaves. And the reason is, as we should find if we took the trouble to analyze ourselves, that we are not serious about high things. About low and small things we are deeply, passionately serious; we are serious about our material rewards, about the price of our work, about our popu- larity that people know our names and faces and the figures of the fortunes that we have made; and. we allow ourselves to be diverted from the things that are really high and serious,—from the aim and purpose of our work itself, from its issue and in- fluence. About these matters we are no more serious than was Flaubert, when he spent his leisure picking over words and shuf- _ fling the cadences of his artificial phrases. He had his reward: he founded a cult and provoked much technical discussion among the curious, and he introduced into French literature a trouble of which it is not yet rid. But he lost the hearts that George Sand won and holds at home and abroad. There is trouble in her books, to be sure, but it is the trouble of life bravely faced and nobly overcome. She does not allow her personal “anxiety about her work to enter and disturb the ultimate peace of art. One feels that she, like Shakespeare, was greater than her task. It is work done without the haste of impatience or the waste of fret; and in consequence it is good and great,—done, I venture to say, in spite of our momentary aberrations, in the true spirit of English work. In its patience and moderation, in its magnifi- ° cent spontaneity and naturalness,.and, above all, in its serenity it is an especially opportune example to the vices of the time, to which we, no matter what our occupations, can return again. and again with a sense of relief and renewal. 221 > Bear a ae I1—Notes on Certain Negative Verb Contractions in the Present BY LOUISE POUND For some time lexicographers have recorded in standard dic- tionaries the colloquial and vulgar forms ain't and hain’t, and their predecessors an’t and han’t. Ain’t especially is now in so widespread usage as to deserve notation in the completer gram- mars also, though Dr. Sweet seems to be the only one yet to give it place. He notes, with his usual tolerance for the idiomatic or colloquial (New Eng. Gr., 1900, § 1491) : “The negative forms [of to be] in the pres. are generally sup- plied by (eint) in familiar speech, which is, however, felt to be a vulgarism, and is avoided by many educated speakers, who say (aim not) instead of (ai eint), (aa ju not) instead of (eint ju).” It was natural (1) that there should be confusion between ain't and hain’t, and (2) that these forms should occur in all persons; the first because of the light quality and frequent insta- bility of English initial h, the second because distinction in per- son and number is not usually observed in negative contractions. So the familiar don’t for does not, dost not, as well as do not. So occasionally with arent, especially in the speech of children. “Aren’t I a good brother to you?” George Eliot, Mill on the Floss, v; “I’m no reader, I aren’t,”’,[b., iv; “I’m a pretty con- siderable favorite with the ladies—arn’t 1?” Captain Borough- cliffe, dramatization of. Cooper’s Pilot (1825), II, ti; “I’m a sort of a kind of a nonentity—arn’t I?” Jb. The vowel sound (é, ei) of ain't and hain’t seems, however, less expected. Obviously the normal contractions of the nega- tive present of to be should be, Ist sg. @z’t (am not), 3rd sg. in’t (is not), plur. dv’t (are not). Similarly Aen’t (Eng. 223 2 Louise Pound © hawt) for has not and have not. Not (é, ei) in any person of either verb. In the eighteenth century, outside of contractions purely dia- lectal (for the numerous dialect contractions of the negative, cf. the Eng. Dialect Dict., Ed. Wright), three forms are found. These are in’t, ant (for pronunciation, see below), and ain't, aivt being apparently the latest form of the three, while im’¢ is the first to disappear. Cf. Richardson, Pamela (1740), IL, Ixxii, “Oh, dear. heart, thought ‘I, in’t it so!’; F. Bumney, Cecilia (1782), I, viii, “However I assure you it in’t true.” Dr. Murray, New English Dictionary, quotes as his earliest example _ of an’t, the plural, 1706, E. Ward, Hud. Rev. (1711) I, I. 24, “But if your Eves an’t quick of Motion, They'll play the rogue.” His first example of an’t am not is 1737, Hist. Reg., - I, i, “No more I an’t, sir’; and for an’t =is not, 1812, H. and J. Smith, Rejected Addresses, 69, “No, that an’t it, says he.” The earliest occurrence of ain’t noted is 1778, F. Burney, Eve- lina, I, xxi, “Those you are engaged to ain’t half so near re- lated to you as we are.” Some other eighteenth century exam- ples of an’t (ait is rarer) are: Sheridan, Rivals (1775), IV, i, “I suppose there an’t been so merciless a beast in the world” ; F. Burney, Cecilia, I, ix, “Why, sure, madam, an’t you his hon- our’s lady?”; /b., I, iti, “It won’t do; an’t so soon put upon.” In the nineteenth century, amt and ain’t are found side by side, with ain’t monopolizing popularity only in the last half: Dickens, Pickwick Papers (1836-37), vi, “There an’t a better spot”; Jb., xix, “Very easy, ain’t it’; Nicholas Nickleby (1838-39), lvi, “It an’t time’; Jb., ii, “Her name ain’t Nick- leby”; Oliver Txist (1837-38), xxvii, “An’t yer fond of oysters”; Ib., v, “Yer the new boy, ain’t yer”; Edwin Drood (1870), v, “It ain’t a spot for novelty”; Thackeray, Vanity Fair (1847-48), xiv, “The sneak ain’t worthy of her.” In America also, an’t was widely written in the first half of last century. Sylvester-Judd, Margaret (1855), has only an’t. “Them an’t yarbs. They won't doctor’ I, v; “She an’t a flower,” Jb. Cf. 224 Negative Verb Contractions in the Present 3 also Cooper, The Pioneers (1823), vii, “Ain’t Marmaduke a judge?”; Jb., xx, “The Squire ain’t far out of the way”; W. G. Simms, Guy Rivers (1835), vi, “I ain’t the man to deny the truth” ; /b., “I ain't slow to say that”; /b., xxx, “I an’t afraid.” In present American dictionaries (cf. the Century, Standard, etc.) ant is not entered as obsolete or obsolescent; although a fairly close survey of contemporary American colloquial and dia- lect literature reveals no examples, but rather, in contrast with results for the first half of the nineteenth century, the complete ascendancy of ain’t. It would seem time to enter an’t as dying, or dead, in America, and restricted to dialect speech in England. In the London Illustrated News, April 4, 1903, occurs, in a “story by “Q”, the form amn’t, “Am I captain here, or amn’t [?” The story is Cornish, and the contraction probably local, hence belongs rather to dialect than to general English. In the eighteenth century, lasting into the nineteenth, the expected han’t was the familiar contraction of have not, has not. Cf. Congreve, Way of the World, 1700, III, iii, “Why then, be- like, my aunt han’t dined”; De Foe, Colonel Jack (1722), iu, “No, it is well if you han’t”; Sheridan, Rivals (1775), I, i, “I doubt, Mr. Fag, you han’t changed for the better’; Dickens, Our Mutual Friend (1864-65), xii, “Why han’t you gone to Lawyer Lightwood?”; George Eliot, Mill on the Floss (1860), xi, “We han’t got no treacle.” So in America, Simms, Guy Rivers (1835), vi, “Han’t I told you”; /b., “Here’s none of ‘us that han’t something to say agin that pedler”; Judd, Margaret (1855), II, xii, “Marm han’t said.” For hain’t, cf. Simms, Guy Rivers, vi, ““Hain’t he lied and cheated’; Jb., “Il, who hain’t the courage”; Ib., “We hain’t got much law and justice in these pairts”; M. C. Graham, Stories of the Foothills (1895), i, “He hain’t no notion o’ doing that,’ “If I’d really had any idee but I hain’t”; J. Fox, Jr., Little Shepherd of Kingdom Come (1903), “I hain’t got no daddy, an I hain’t never had none.” There was confusion between an’t and han’t, as now between ain't and hain’t. For the use of ant for han’t cf. Sheridan, 225 4 Louise Pound Rivals (1775), IV, i, “I suppose there an’t been so merciless a beast in the world” ; Dickens, Our Mutual Friend (1864-65), xii, “Have you finished?” ‘No, I an’t”; Bleak House (1852-53), vii, “No, I an’t read the little book wot you left,” etc. Han’t for an’t is found in Pickwick Papers (1836-37), xi, “Where’s that villain Joe!” “Here I am; but I han’t a willin.” Examples of ain’t for hain’t are: Dickens, Bleak House, “We ain’t got no watches to tell the time by”; S. O. Jewett, A Dunnet Shepherd- ess (1899), vii, “Ain’t William been gone?”’; Jb., Where’s vora? “Ain't you got the Queen’s. luck?” The still more illit- erate hain’t for ain’t is heard often, but is less frequently writ- ten, “I hain’t nothing but a boy,” J. Fox, Jr., Little Shepherd of Kingdom Come, 1903; “Oh, you’re a reglar tin peddler, hain’t yer?”; “You ain’t green,” “You bet I hain’t”; Julian Ralph, Trip with a Tin Peddler, Harper’s Mag., 1903; “Dat rifle hain’t neber gwine kick,” Her Freedom, negro story by V. F. Boyle, Century, Feb., 1903. In the latter story was noted an example of ain’t for don’t, “I sho ain’ want ter mairy Rias.” Negro dia- lect has cain’t, beside ain’t and hain’t, as “I cain’t go in dar—no eain’t. 710: As suggested in opening, an’t and han’t some time ago found their way into the dictionaries, ain’t and hain’t more recently. Bailey enters none of the four forms; Dr. Johnson han’t only. An edition of Webster’s Dictionary, as late as 1855, when ain’t had become about as widespread as an’t (cf. examples supra) enters an’t, han’t only. Both Webster’s Dictionary and the Im- perial Dictionary (1856), based on Webster’s, give the following curious etymology : 4 “An’t, in our vulgar dialect, as in the phrases, I an’t, you an’t, he awt, we amt, etc., is undoubtedly a contraction of the Danish er, ere, the substantive verb, in the present tense of the indicative mode, and not; I er-not, we ere-not, he er-not; or of the Swedish @7, the same verb; infinitive vara, to be. These phrases are doubtless legitimate remains of the Gothic dialect.” With regard to the vowel sounds in the contracted forms an’t, 226 Negative Verb Contractions in the Present 5 han’t, the lexicographical evidence within reach shows not a little diversity. Was the sound the expected Eng. (a), Am. (e, a), or was it (é€), and the difference from ain’t, hain’t, for the most part one of spelling only? Dr. Johnson, as so well known, does not mark pronunciation, nor does Richardson. Knowles, A Pronouncing and Explanatory Dictionary of the English Lan- guage, London, 1848, gives (a, €) for am’t, entering no ain't, and for hawt (a), recognizing no (hént). Webster’s Diction- ary, ed. 1855, gives (e) only for an’t, han’t. The Imperial Dic- tionary (English), 1856, gives (€) only, like Webster’s. Dr. Murray, N. £. D., 1885, gives (a) only for an’t, reserving (€) for ain't. Of present American dictionaries, the 1900 Webster gives (€) only for both an’t and han’t, mentioning, however, an English han’t; the Standard gives (a) only for an’t, but (ce) only for han’t; and the Century gives (a, é) for an’t, with the note, “In the second pronunciation also written ain’t”, and (é) only for han’t. Perhaps before affirming too much, there should be closer ex- amination of lexicographical and other evidence that was pos- sible from the material at hand; nevertheless it seems probable, despite the testimony, English and American, given above, that the entry of the sound (é) for amt and han’t is nothing more than a legacy from the period when ain’t and /hain’t were not yet recorded, the older forms having to do double duty. (v. _ Knowles.) When has a followed by nt had the value of (€)? It would be difficult to believe that an’t and ain’t, forms which occur side by side in so many early and middle nineteenth cen- tury texts, when orthography within individual authors is fixed, were for the most part nothing more than variant spellings for the same (ént). If the entries are to be exact, our American dictionaries should probably adopt the New English Dictionary (1885) differentiation of an’t and ain’t, and furthermore clearly distinguish han’t (2, a) and hain’t. The spelling of colloquial and vulgar forms is generally pretty phonetic. An’t is not now written for (ént), and in the days when it was commonly found was most probably pronounced as spelled. So with its com- panion form. ry 6 Louise Pound For the vowel sound of ain’t, so far as I am aware, no ex- planation has been offered, at least no accepted explanation. The (€, ei) could hardly be due, as I have once or twice heard sug- gested, to the analogy of mayn’t, or of some such form perhaps as the dialectal bain’t. If following analogy the contraction would more likely have fallen in with shan’t and can’t. Nor could it have assumed the vowel of hain’t. The latter arose, probably, slightly later, and has itself to be explained. Rather may (é, ei) be the result of some sort of unconscious compro- mise between the vowels of am, are, and is. The vowel of are would be moved forward and higher, and that of am be made. higher and closer, through the influence of 1s. A compromise between (4), (#), and (i) could hardly result otherwise. The vowel of han’t then went the same way. 228 I1I.—On the Variation and Functional Relation of Certain Sen- tence-Constants in Standard Literature BY R. E. MORITZ ‘*Surely the claim of mathematics to take a place among the liberal arts must now be amitted as fully made good. . . . It seems to me that the whole of zesthetic (so far as at present revealed) may be regarded as a scheme having four centers, viz., Epic, Music, Plastic, Mathematic. There will be a common plane to every three of these, outside of which lies the fourth; and through every two may be drawn a common axis opposite to the axis passing through the remaining two,” J. J. Sylvester, Philosophic Magazine, 1878 (1), p. 184. “Tt therefore seems clear that mathematics can be shown to sustain a certain relation to rhetoric and may aid in determining its laws.” L. A. Sherman, University Studies, vol. I, p. 130. Pythagoras taught that the essence of things is numerical re- lation, Aristotle that number is the means to true knowledge. In more modern times Lagrange conceived of the possibility of representing the complete history of the universe by one huge differentia! equation, the actual state of progress at any moment by a single time integral between the limits, minus infinity and zero. According to Herbert Spencer’s definition.of a law of nature, every such law can ultimately be embodied in an alge- braic equation. Solvay, a Belgian scientist, recently established an equation governing the energy set free by an organism in vital phenomena for a given food supply. Even Rhetoric bows to Number. During the closing decade of the last century it was demonstrated that all good writers lisp in numbers, that there is cadence in the essays of Bacon and Emerson and in the histories of Hume and Macaulay just as truly as in the “winged words” of a Homer, a Milton, or a Goethe. This cadence is as inaudible as the music of the spheres to ordinary ears, but it reveals its sweet measures to the patient 229 2 : R. E. Moritz investigator. There is a sentence-rhythm, as it has been called by its discoverer, which is the hidden mark, the cipher, the cryp- togram, with which each author unconsciously endows the prog- eny of his pen. Measured, it yields a number, an author-constant, corresponding to the wave-length of this rhythm. And just as the wave-length of standard musical tones has been shortened in the course of time,* so the author-constant has gradually di- minished from the beginning of English literature to the present day. a Such is a brief paraphrase of a theory set forth at some length by Professor L. A. Sherman? in two articles published in 1892 and 1894 respectively. In these, as in a later article by G. W. Gerwig,* results are announced and tentative statements are made which lead one to infer that English prose writers conform to sentence-instinct, which has all the force of a positive law in controlling their written utterances. After tabulating the sen- tence-lengths* measured in words of from 500 to 800 consecu- tive sentences from each of the authors De Quincey, Macaulay, Channing, Emerson, and Bartol, Professor Sherman remarks: “Now that the number of words in consecutive sentences was definitely exhibited, strange facts and features of style were in- dicated or suggested. The length of one sentence, it was shown, might be echoed unconsciously into the next, as notably in Ma- caulay’s groups of seventeens. . . . But the really remark- able thing was the apparently constant sentence-average in the respective authors. Could it be possible that stylists, as eminent and practiced as these, are subject to a rigid rhythmic law, from 1 Ellis, an English physicist, has shown that the concert C4 normal tone has reduced its wave-length (in air) from 2,33 to 1.99 feet in the course of 130 years. 2, Some Observations upon the Sentence-Length in English Prose, Univer- sity Studies, Lincoln, Neb., vol. I, no. 2, p. 119. On Certain Facts and Principles in the Development of Form in Litera- ture, Ibid., vol. I, no. 4, p. 337. 8On the Decrease of Predication and of Sentence Weight in English Prose, 101d sy VOL: nOLek pre liee *The study of sentence-lengths was suggested by T. C. Mendenhall in his articleon Zhe Characteristic Curves of Composition, in Science, March 11, 1887, who in turn credits August DeMorgan, the well-known mathema- tician, with the priority of the thought of applying numerical analysis to the study of literary style. 230 - Lee Sp aur 4 nil te EN, Sea Des Variation of Sentence-Constants in Litera:ure 3 which even by the widest range and variety of sentence-lengths and forms they may not escape? At once pushing the suspicion to a proof, I made, first, an extended test in Macaulay’s essays: result, 23+, the number obtained before; then in Channing: average again, 25.”* After several other tests with similar re- sults he continues, “No evidence appearing to the contrary, if seemed likely enough that sentence-rhythm was a_ universal law.”* (The italics in the preceding quotations as well as in those which follow are mine.) Mr. Gerwig occupied himself particularly with the average number of predications per sentence and the per cent of simple sentences in various authors, to the number of one hundred. His conclusions are summed up in the following words: “A very little investigation served to convince me that the same re- markable uniformity which had been found in the average num- ber of words used by any given author per sentence would also hold in regard to the number of finite verbs, or predications, found in each sentence. The results obtained convinced me also that there was a uniformity in the number of simple sentences per hundred of a given author.”*® Mr. Gerwig then examined Chaucer’s Tale of Melibeus and 2500 sentences from Macaulay, and, finding the expected uniformity, he says: “Other authors _ were taken in the same way until it was demonstrated that the average of 500 periods of any author who has achieved a style was approximately the average of his whole work.”’* In par- ticular he discovered that “while Chaucer and Spenser habitu- ally put over five main verbs in each sentence they wrote, and less than ten simple sentences in each hundred, Macaulay and Emerson used only a little over two verbs per sentence, and left over thirty-five sentences in each hundred simple.”® In neither of these quotations is there any explicit statement to the effect that the principle suggested or announced is inde- pendent of the nature of the composition, but the implication \Wnive: sity Studies, vol. I, no, 4, p. 348. *Jbid., p. 349. 8University Studies, vol. II, no. 1, p. 17. 4Jbid., p. 18. *Jbid., p. 19. 231 3 4 R. E. Moritz certainly is that there is a single set of constants for each author. Professor Sherman speaks of “a rhythmic law from which an author may not escape,” and asserts that “the determining factor (of sentence-length) in each case is the relative capacity of the author to respond to what may be called the sentence-sense in his own mind,’? “which, if it could have its will, would reduce all sentences to procrustean regularity.”* Here we notice the excepting clause “if it could have its own will,” indicating that the writer was conscious:of some kind of limitation of the prin- ciple in question; and elsewhere we read, “to avoid complica- tion, no consistent attempt has yet been made to determine the sentence-average in works of fiction. Here of course the matter is mainly narrative or descriptive, thus reaching the imagination of the reader more directly, also much of the language is quota- tion and dialogue.”* Yet dialogue was not ruled out in com- parisons. In the test case of the 40,000 sentences in Macaulay’s History of England, it was found that “After the dialogue pas- sages and consequent reduced averages, seemingly by a sort of reaction, full rounded periods and high averages take their place.”* Later on, this restriction seems to have been entirely ignored, as, for instance, when Miss C. Whiting,® in studying the relative sentence-lengths during different periods of English literature, by an examination of 500 sentences from each of 60 authors, admits De Foe’s Robinson Crusoe, Fielding’s Tom Jones, Goldsmith’s Vicar of Wakefield,-Scott’s Kenilworth, Eliot’s Mid- dlemarch, Howell’s Rise of Silas Lapham, etc., alongside of Mil- ton’s Areopagitica, Dryden’s Discourse on Satire, Gibbon’s Rome, Emerson’s American Scholar, Channing’s Self-Culture, Spencer’s Data of Ethics, ete. Gerwig remarks that “the question incidentally arose whether a writer had the same sentence-structure in poetry as in prose,” ® 1University Studies, vol. l, no. 2, p. 119. 2/bid., vol. I, no. 4, p. 353. 8Jbid., vol. I, no. 2, pp. 129, 130. 47bid., vol. I, no. 4, p. 352. 5 The Descent of Sentence-Length in English Prose, master’s thesis, Univ. of Neb. (unpublished). 8University Studies, vol. II, no. 1, p. 4. 232 Variation of Sentence-Constants in Literature 5 and realizes that “it would be manifestly unfair to compare trun- cated, dramatic dialogue (such as Shakespeare’s), abounding in exclamations and broken sentences, with the even flow of the ‘Hind and the Panther.”' To. equalize matters, he proposes to examine only passages three or more lines in length. But whatever limitations to the Sherman principle? may have been recognized by its author and Mr. Gerwig, they were all | swept aside at one stroke in an article which appeared in 1897.* In this article, the principle of constancy of sentence-length, predication-averages, and simple-sentence-percentages is set forth as follows: “Ten years or more ago Professor Sherman, while investigating the course of stylistic evolution in English prose, made the discovery that authors indicate their individuality by constant sentence-proportions, personal and peculiar to them- selves. This was demonstrated especially with the number of words used per sentence in large averagings. It was found that De Quincey, Channing, and Macaulay, if five hundred periods or more were taken, evinced this average invariably, and in the ear- liest as well as in the latest period of their authorship. This discovery led to the suspicion that good writers would be found constant in predication averages, in per cent of simple sentences, and other stylistic details. Acting upon a suggestion to this effect, Mr. G. W. Gerwig, then a pupil of Professor Sherman, undertook an investigation that established the constancy of pred- ication, as well as simple-sentence frequency, in given authors. . . . Professor Sherman and Mr. Gerwig have thus estab- lished by the examination of a great many authors, that writers are structurally consistent with themselves; that they possess a ‘certain sentence-sense peculiarly their own. These investigators have established that by this instinct authors use a constant aver- age sentence-length, and a certain number of predications per sentence, and that a given per cent of their sentences will be simple sentences. . . . The work of these investigators cov- Vbid., p. 3. 2In honor of its discoverer, I shall call the principle of sentence-instinct, into whatever form it may ultimately crystallize, the Sherman principle. SUniversity Studies, vol. Il, no. 2, p. 131. 233 6 R. E. Morite ers a large amount of material and a wide field of literature. They have examined and compared the works of ancient and re- cent authors, early and late writings of the same author, and writings of the same author of different character, such as his- tory and dialogue, poetry and prose. The results thus far ob- tained are sufficient to show that it is not possible for a writer to escape from his stylistic peculiarities.” * There is no uncertainty, no indefiniteness, no ambiguity in these statements. The principle as here set forth is as thor- oughgoing as it is simple. A good writer can not escape from his stylistic peculiarities any more than he can change his pace or alter his voice. Whether he write history or drama, poetry or prose, these stylistic characteristics are ever present with him. They permeate all his writings. They manifest themselves in certain numbers, such as sentence-length, predication-frequency, simple-sentence-percentage, etc. Once discover these numbers — and you have marks which will serve for detective purposes quite as well as a man’s chirography. They are the earmarks by which to trace anonymous and disputed writings to their sources, the touchstones to disclose the spurious and the false. That this is the thought is clear from the sequel of the article referred to. If Shakespeare’s plays were written by Bacon, they will reveal the same constants as the other writings of Bacon. In the author’s own words, the end sought is evidence touching the authorship of Shakespeare’s plays; whether Bacon wrote Shakespeare’s works, or at least whether the Baconian and Shakespearean writings were the work of one and the same per- son, or of different persons. Consequently, an examination is made of the prose in fifteen of the Pays, of Bacon’s Essays, and the New Atlantis. To eliminate possible errors, arising from careless and inconsistent punctuation, all the material is re- punctuated according to modern principles. Fairness is added to consistency by omitting on the Shakespearean side of the in- quiry all inorganic, broken, and suspended diction. Then follow twelve pages of figures representing totals and specimen results, and finally comes the 1Wyniversity Studies, vol. 11, no. 2, pp. 147-148 234 A | Variation of Sentence-Constants in Literature 7 GRAND SUMMARY neue z No. of |Sentence-| Predica- | Simple examined|seutences| length tions |sentences Shakespeare ............ 61956 5002 12.39 1.70 39 MSACOUR sats ero a's Sc cee 66524 2041 32.59 3.45 14 Instead of adding evidence to the Bacon-Shakespeare contro- versy, this argument seemed to me merely a reductio ad absur- dum of the Sherman principle itself as stated by the writer. It served to convince me of the existence of certain limitations to the principle in question, which had not been recognized, certain restrictions which had been violated. For who is not aware that dramatic prose generally, if not invariably, contains shorter sen- tences, consequently, other things being equal, fewer predica- tions per sentence, and a larger per cent of simple sentences than do other forms of prose composition? The argument, based upon a difference in sentence-constants merely, could be used equally well to prove that Dryden, the dramatist, did not write the famous Essay on-Satire. The only real information conveyed by the above summary is not the fact of. variation, which is a common notion of all who are at all familiar with the writers in question, but the widely divergent results indicated by the numbers 12.39 and 32.59. Can this difference in the sentence-constants be en- tirely accounted for by the difference in character of the material examined? To satisfy my own mind in the matter, I was im- pelled to make a test. Goethe’s works were at hand. [ selected one of his prose dramas, Goetz von Berlichingen, and his Bild- hauerkunst, a collection of essays on art. The results were: GOETHE GOETHE (Goetz von ee (Bildhauerkunst) First five hundred periods. . ASF First hundred periods erty oe 27.1 Second ‘‘ es AER Gy BECOME ie ae Nar ee sec 32.9 Third —-“* = aie one A Dhird- < Ae hee hen 33.7 Fourth ‘ ne ab a PL eo Fourth ‘ Ata adlnss A 27.0 Fifth ‘ 2 ss 8.1 Fifth ‘ $e Tes tactiet ans £2 36.9 Average for 2,500 periods...... 8.5 Average for 500 periods...... 31.5 235 8 k. E. Moritz These results showed even a greater divergence than those ob- tained from a comparison of Shakespeare’s and Bacon’s prose. Possibly Goethe occupied a unique position in this respect. I continued the test with an examination of Schiller. This time I selected Die Rauber, a prose tragedy, and his History of the Thirty Years War. SCHILLER SCHILLER (Die Rauber) (History Thirty Years War) First five hundred periods... 12.2 First two hundred periods... 29.9 Second ‘ eS MOL see EL Second Yo.9t¢ phd EAC 8 Third “ ri pane eee bey @hird j.* e (6 1 ieee Fourth ‘ 7 ests Pe ORS Fourth ‘‘ ‘f CO Sle oes Fifth ‘ ee y gorge ese UA Fifth “ es eee rach it Average for 2,500 periods.... 11,5 Average for 1,000 periods..... a ee This would not do. Whatever sentence-rhythm these German writers manifested was certainly greatly dependent upon the par- ticular style of composition employed. Perhaps I had made a mistake in making tests from books nearest at hand, and should have limited myself to English authors. So I took up Swift and Dryden, who, I thought, would be unobjectionable from any point of view. I examined a prose drama and an essay from each, and decided to limit my examination in each case to five hundred periods.? z SWIFT SWIFT ; ‘(Essay on the Four Last Years of (Polite Conversation) Queen Anne) First hundred periods....... 10.8 First hundred periods....... 51.6 Second ‘‘ a av be 12.0 Second ‘ GS are 2 49.3 Third Ea eas ge 12.1 et hiirdess es RTRs asenes 6: 58.2 Fourth ‘“ SOP aa eed: IIS AT} Fourth ‘‘ SE dea on 62.1 Bitth ck Decl 54 ES ADLO Fifth aL (top 2 Sen eaaee 53.0 Average for 500 periods...... 12.4 Average for 500 periods...... 54.8 1Professor Sherman expresses the opinion that three hundred periods will generally reveal the sentence-rhythm of any author who has achieved astyle. University Studies, vol. 1, no. 2, p. 130. 236 Variation of Sentence-Constants in Literature 9 DRYDEN DRYDEN (Zhe Mock Astrologer) (Essay on ieee First hundred periods Satps bete 15.0 First hundred periods. . . 45.0 RPOLLUE eptrini es ss ats Say Ss 16.5 ECON ee se 48.1 a eh § yo a TNO ag apie hy 19.6 hire” << ae wey 40.1 Fourth ‘ Soices rates stants Tie Fourth ‘ Sete lash onese 44.8 Fifth oh Seer eS ak conte 16.3 Fifth rs SBa Dike sh 33.3 Average for 500 periods...... 16.9 Average for 500 periods...... 49.3 In these authors the divergence of constants was even greater than in the case of Goethe and Schiller. Shakespeare’s 12.39 and Bacon’s 32.59 seemed no longer remarkable, the ratio of these two constants being 2.6, while the corresponding ratios for Goethe and Swift are 3.7 and 4.4 respectively. These results urged me to continue the investigation. I soon found that the sentence-constants varied not only when a com- parison was made between drama and history, or essays, but in other forms of composition as well. I append a few of my results :* GOLDSMITH (Present State of Polite Learning in Europe) GOLDSMITH (She Stoops to Conquer) First hundred periods Bate 13.9 First hundred periods gee c yin 30.4 PRIN se cesen NeS aiahaes 13.1 CONE On be oo ho 4 Io eee 24.6 Bird. °° SPR EEN ara: 13.2 ray: 5° Ae oat 25.7 Fourth ‘ Meera ace ». 14.9 Fourth ‘ SASS Brae 23.5 Fifth o EEA ESR. 12.4 Fifth “ SS oP rag ieee Os 20.4 Average for 500 periods. . Neo 13.5 Average for 500 periods...... 24.9 GOLDSMITH SCOTT ( The Vicar of iA eclerad: (Ivanhoe) First hundred periods. . . 31.2 First hundred periods ware ena 46.2 Se NSS SES emer aa a 30.8 ps Eliot (Aa aS ee ore 39.3 Third‘ SoG eae 27.5 Ehixd: 5 pabedrt. ay ieee 33.7 Fourth ‘“ SL Ra ee We 25.8 Fourth ‘“‘ py aimee es tae 29.8 Fifth es oats iene 26.5 Fifth TEEN: ophats cieeyeaeoes Average for 500 periods...... 28.4 Average for 500 periods...... 35 4 1In each case the count was made from the beginning of the work cited. Introductions, headings, footnotes, sentences containing long quotations, verses, and, in the case of dramas, stage directions were consistently omitted. 237 se) SCcoTT ( Auchindrane) First hundred periods Pere aiere 19.1 BICCORG eo Sf No nrel es ccc tio 21.9 puhind te bee ws ictate, a8 23.9 Fourth ‘ ST cdottarece ere 19.2 Bitte a." a 21.8 Average for 500 periods...... 91.2 CARLYLE (Signs of the Times, etc.) First hundred periods rettete 27.4 Second ey sc ve ae. arene netavers 31.8 ‘Liurd 5. "GD taisniGt ranch 26.9 Fourth ‘‘ a aN Pi 34.5 Fifth rs SES Ene aihts rs 42.3 Average for 500 periods...... 32.6 BAYARD TAYLOR ( Zhe Prophet) First hundred periods Rae 15.8 Second ye 2 Sis Ute ae seas ats 1138 Third: <~A° an aa 11.3 Fourth ‘“ SR erecta aunts 14.8 Fitth < LP Peters a 5. 16.4 Average for 500 periods. epee 14.0 LOWELL (Letters, vol. IL) First hundred periods Meare 18.8 SECO Goins eres ce ee nha natane 18.2 Third: +5 gee eae 22.5 Fourh ‘ Peas teats 21.8 Fifth e BAO oe Seca 17.2 Average for 500 periods...... 19.7 HOLMES (Guardian Angel) First hundred periods Sb skits 26.4 Second > 6 ih ie ane 29.0 hind? a: OSS esses acts reaate en Fourth=* SEE warrant: 31.1 Fifth oS es 25.6 Average for 500 periods...... 98.9 238 R. E. Montz SCOTT (Life of Napoleon) First hundred Sere ee ete 46.0 Second ‘‘ PP ee inl Third se 08 Reena 50.0 Fourth ‘ 16s eee 49.8 Fifth re ‘h. 7 oe 46.5 Average for 500 periods...... 47.9 CARLYLE ( The Hero as Divinity) First hundred periods i sexeerd 25.2 Second 9-0) sic SS ee ace 21.5 hind aa SS. 5 oink eae 24.3 Fourth ‘ HO is Sie. 23.5 Fifth 7 iS: Saeed 23.4 Average for 500 periods...... 93.6 BAYARD TAYLOR (Balearic Days) First hundred pe: sane 30.3 Second ‘ age pee Third oe See ROR eee Fourth ‘ Re Rumer oo 26.6 Fifth ca AS ee 27.6 Average for 500 periods...... 98.3 LOWELL | (Fable for Critics) First fifty periods Ra eo 106.5 Seconds) ic Fae aero 88.7 fb shbalt RRM GPP baie edn ho 87.6 Remaining 30 periods....... 108.3 Average 180 periods ........ 96.6 HOLMES (The Autocrat at the Breakfast Table) First hundred periods. . 1958 Second’ “ooo Se sce 18.8 Phir rte ee Se eee 18.0 Fourth ‘ SS oe coe Lees 23.8 Fifth He em tsangrs Scary 5 20.1 Average for 500 periods...... 20.1 1s do aah Variation of Sentence-Constants in Literature Il LONGFELLOW LONGFELLOW ( The Spanish Student) : ( Ayperion) First hundred periods....... 12.0 First hundred periods....... 20.5 Second ‘ nie Sekar 8.8 Second ‘ SESH saeeach. si & 25.4 hard .‘** Beam antes! 11.1 Third ~.,‘ Se rire dats . 27.5 Fourth ‘ ete pte 8.8 Fourth ‘ Da Seen aC 24.3 Fifth a oe ar a ae 10.1 Bitth< 5 Sas wie me 21.3 Average for 500 periods...... 10.2 Average for 500 periods...... 93.8 The list could be indefinitely prolonged, but it was not neces- sary. The evidence seemed to me to show that sentence-lengths, and presumably also predication-averages and simple-sentence- percentages, are quite as much dependent upon the nature of the composition employed as upon the author’s sentence instinct. In fact, | surmised that the variety of sentence-lengths which an author employs is limited only by his versatility as a writer. Acting on this surmise, I decided to make a test. Goethe was the author selected for this purpose, for he seemed to meet most nearly ideal conditions. His style is unquestioned, his punctu- ation is consistent and scientific. Most important of all, he is the most universal writer of the last century. He ranks high as a writer in the fields of poetry, drama, fiction, biography, art, literary criticism, travel, and science. The results* fully cor- roborated my conjecture. 1JIn securing these as well as the preceding results, the sentences were actually counted, but where the nature of the text permitted it, the number of words has been obtained by counting the lines and multiplying by the average number of words per line. In this way the drudgery of the work may be greatly shortened without impairing the accuracy of the results, since counting, like every other arithmetical operation which is not carefully checked, involves unavoidable accidental errors. To test the accuracy of my method, I determined by means of it the number of words in Chaucer’s 7a/e of Melibeus and obtained 16,633 words as against 16,659 which was obtained by an actual count. In less than two hours I determined the number of words in Macaulay’s //story of England to be 979,668 as against 974,195 obtained by actual count. Assuming that this last number is absolutely correct, my estimate involves an error of 14 per cent, which may be safely neglected since it will be shown that the average sentence lengths based upon 500 sentences involve an average error of over 2 per cent. 239 12 GOETHE (Dichtung und Wahrheit) First hundred periods Nasties. 31.5 SECOMG Kee) eshte iho Sareehiny 36.1 Rhird =. yoo eS eects 30.9 Jeroyelfeelet) ele SEER Saree ee 28.9 Fifth * tee ete AR 31.1 Average for 500 periods...... 31.7 GOETHE (Faust: Second Fart) First hundred periods skate Re 16.3 SS[ececon 6 Unmatd Gain noe (AUS Met oer eter E 17.2 Phirds he ee eae NN 15.5 Fourth ‘“‘ nie Naru breil W503 Fifth He Re ce ah ae 16.5 Average for 500 periods...... 15.6 GOETHE (Die Leiden des Jungen Werthers: Book 2) First hundred periods vis hier Wis 20.7 Second sh ork Sn mee spe icy 20.5 Third. 7.“ SOR. Bathe hea 1931 Fourth ‘‘ Ore emcee lites 22.7 Fifth =~ ete RE Een 18.2 Average for 500 periods...... 90.3 GOETHE (Der Birgergeneral) First hundred periods 5 he 6.7 Secoudwg Miptxee eee oocyte 5.1 ol Ploy hs bo pike Sree iaio.c 4.5 Fourth ‘ ic te Shoah 3.9 Fifth x En, Maton Sse 4.7 Average for 500 periods...... 5.0 GOETHE (Die Wahlverwandtschaften ) First hundred periods ISR Ea cit 21.9 SECOHGEY Hy oe tiopire oe ves teke 23.6 imstbilk Wat Dy Dar | Fourth ‘ pn Cine: St ae 25.1 Fifth y us 24.5 Average for 500 periods...... 93.4 240 R. E. Moritz GOETHE (Faust: First Part) First hundred periods Hel #oaeAd 14.6 Second: <7, ¢ ee eee 15.2 Thainds esr oe eee 13.6 Fourth ‘‘ io cera! saa 13.0 Fifth a Ari 2 12.0 Average for 500 periods...... 13.7 GOETHE (Reinecke Fuchs) First hundred periods eaten 18.5 Second). ee 16.5 Sica. © 20° OA only Rest aa 16.9 Fourth ‘ “s 14.8 Fifth Je KS 16.6 Average for 500 periods...... 16.7 GORKTHE (Briese aus der Schweiz: Part 2) First hundred periods. . 23.2 Second 2° (2 yt ee eee 26.5 Third’ 4-" Sf OV yerertaees 25.6 Fourth ‘ OE Ua 23.0 Fifth = sf 28.4 Average for 500 periods...... 95.3 GOETHE (Entwurf einer Farbenlehre: Di- dactic Part) First hundred periods a eee 28.4 econd sin Sa sate eee 25.7 ‘Third =. 59 iC" raat eee Fourth ‘“ ‘S22 Saas 22.3 Fifth os fo so ee 26.1 Average for 500 periods...... 96.4 GOETHE (Literatur: Recensionen) First hundred periods Senet 37.3 Secotrd!s ore eae eee 32.9 Third 2.24 Stay Sane 36.9 Fourth “ Be ars) ae Fifth be fo. DY Sapa 34.9 Average for 500 periods...... 34.7 Variation of Sentence-Constants in Literature 13 GOETHE GOETHE (Ltalidrische Reise: Rom.) (Letters to Frau von Stein) First hundred periods emt ge 23.1 First hundred periods nets) ote 13.5 MIMO ee Ny eS ae 23.7 PIEOMNEL «ew eit se sed o0 8 12.5 Thisd , .‘' sey sam Ree 22.7 Dhird-. 34" Bape ewn thre 11.3 Fourth ‘ Sper Pana caee 21.5 Fourth ‘“ PG Metin ke 11.2 Fifth 3 Wess ep SOR? | Fifth + Se Mrsie oe evoistons 12.4 Average for 500 periods...... 99.7 Average for 500 periods...... 12.2 GOETHE ( Die Metamorphose der ss zen) First hundred periods. . . 33.8 WRCODLy oe ss eee roses, 5 34.1 Remaining 88 EES Soe re 33.3 Average for 288 periods...... 33.7 If we arrange these results, including the numbers previously found for the Goetz von Berlichingen and Bildhauerkunst, in order of the*sentence-length we have the following instructive table: TABLE I VARIATIONS IN SENTENCE-LENGTH OF GOETHE'S WRITINGS Rm CURCCOCTICT alli asia) coiaal< coin "s/h op) 8 ciate a cietelo lohel ois ssh, © Sigiblae eon 5.0 PI WASTE ROUTER ION 80 a. Says ein’ si pee wigs e eee W ole. e's At ee wan ole ome 8.5 err errant eat VeNtetnl ys 6 Foss creche Whee cise te boys Ged ee eel rat 12.2 TE vga SHV at SCT EI ey Bet S72 de GE Pe ee a neces ge A ee RPE PT 13.7 BASE ROE COMM Hatt Sila otal ec wl tdclok eS Dale ens ti tig aheny alee eins «Bote 15.6 etek OAPARC L SE AU aint ores ies aalccse eae ws whem eiat> NewteteiaLioe ave mre 16.7 Me tena se 1h WW ETLD CES Co 2 es Soc ad esis ae brass Sees Seely eed ace 20.3 MPA ATMOCHE IR GISE, ROTI co ccie ho ace ee oahu eee po sb oe sale lag ae stes 22.4 Die Wahlverwandtschaften..... 23.4 ere TES LETS IE WEIS 1S cin ar eace tants ie oite maine ne pres earintoh ew n) ohalnade rs gies 25:3 ea trattel vette SAR DOMICILE > '-55 < vocints, fellas A fotaes «ee alecbincac eles G10 (riences 26.4 PA CATR EMT SIBLE east pate Ae cis isu ae std A clei averted ale sioh se vie ee 31.5 Dichtung uw. Wahrheit ......... 6. eee eee eee ee eee eee ee Maiogs 3).7 Metamorphose d. PAlanzen..........---e eee e ee eee eee eee eee 33.7 Paterattir. RECeNSIONEM 9.6. os cere ees biele cee eh ewe cee ce ce enes 34.7 241 14 R. E. Moritz The above list includes romance, drama, allegory, criticism, biography, description, science, correspondence, buc with tne ex- ception of Faust and Reinecke Fuchs the works are all in prose, so that the fact of variation in sentence-length appears, even if we consider prose literature alone. There can be but little doubt that an examination of all of Goethe’s writings would furnish a chain of sentence-lengths varying by almost insensible grada- tions from five to thirty-five or forty words per sentence. Other authors may not yield such wide extremes, but it is the fact rather than the extent of variation which is essential in this discussion. It is hardly necessary to comment further on the above results. They demonstrate absolutely thé unreasonableness of applying the Sherman principle indiscriminately to various types of com- position. They demand that, before we compare the sentence- constants of different authors, or the normal sentence-constants of the writers of a given period, there shall be some agreement or understanding regarding a standard type or standard types of composition.'. Any conclusions concerning the stylistic evolu- tion of literature, which are based upon the principle of con- stancy and a disregard of the principle of variability of sentence- length must be considered worthless, even if they lead to desired or plausible results. As an illustration I will only cite the inves- tigation by Miss C. Whiting’ on The Descent of Sentence-length in English Prose. Assuming the Sherman principle, it appears, from an examination of single works by each of sixty authors, ranging from Chaucer to Henry James, that there has been a decided diminution in the sentence-length. Averaging her re- 1Some of the specimens which I have examined, such as Schiller’s Rauber and Goethe’s Goetz von Berlichingen, may be objected to because of the rather abnormal nature of the compositions. They abound in exclama- tory and interrogatory dialogue. I admit that they are not normal types, but it is the necessity of agreeing upon what shall constitute normal types which I desire to point out. Shall we disregard all interrogatory and ex- clamatory passages? If not, what proportion shall be counted out? Or shali we omit all sentences less than four, five, or six words in length? Or adopt Mr. Gerwig s rule and examine only passages three or more lines in length? The principle of variability could not be denied even though it were claimed that all of Goethe’s sentence-lengths, but one, are unnatural. 2Master’s Thesis, Univ. of Neb. (unpublished). 242 ‘ Variation of Sentence-Constants in Literature 15 sults, I find that the normal sentence-lengths of writers before the seventeenth century, during the seventeenth century, the eighteenth century, and the nineteenth century, are represented respectively by the numbers 48, 42, 36, and 27, if the original punctuation is used, and by 42, 38, 34, and 26, respectively if the works examined are repunctuated., But in the selection of the works examined, no thought seems to have been given to uniformity in the form of composition; at any rate we find works so widely divergent in structure as Chaucer’s Tale of Melibeus, Bacon’s Essays, Milton’s Areopagitica, Bunyan’s Pilgrim’s Prog- ress, Locke’s Essay on the Human Understanding, Taylor’s Sermons, Hume’s History of England, Goldsmith’s Vicar of Wakefield, Irving’s Life of Washington, Spencer’s Data of Ethics, and Howell’s Rise of Silas Lapham enjoying equal suffrage in the comparison. Professor Sherman brought to light the remarkable fact that an author’s sentence-length remains practically constant through- out a given work, an extensive test having been made on’ the 40,000 sentences of Macaulay’s History of England, but here we have the incontestable fact that the sentence-length of one and the same author may vary by almost insensible degrees be- tween limits so widely divergent as five and thirty-five. The conclusion from which there seems to be no escape is that the sentence-length of a work depends both upon the writer's sen- tence-instinct and upon the particular form of composition into which his thought is cast. That is to say, sentence-rhythm, inas- much as it manifests itself in constant sentence-length, is a func- tion of at least two variables + and y, where .x signifies the author’s sentence-sense and y the form into which he moulds his thought.? 1T trust that this statement will not be interpreted as contradicting or ~ de-util zing the Sherman principle. All that it insists upon is the necessity of somod fying the principle as to recognize the facts of variability in the sentence-constants. The results ob ained from Macaulay’s History are in perfect harmony with these conclusions. Professor Shermin found the average sentence- length of the Astory (5 volumes containing 41,500 periods) to be 23.43. This is practically the same as the sentence- ength 23.65 1or Mach:avelit, 24 for P.t/, and 23.C0 for the Zssay on History by the same author. Now the 243 16 Rk. E. Moritz I shall now endeavor to show that the principle just stated applies also to the other sentence-constants, predication-averages, and simple-sentence-percentages. It will not be necessary to produce a chain of different predi- cation-averages or simple-sentence-percentages corresponding to the chain of sentence-lengths which we found in Goethe. We need only show that there exists a functional relation between the various sets of constants, such that a variation in one set produces a variation in each of the other sets. Mathematically expressed, we need only show that P=f (L) S=$ (P) } where L=sentence-length, P=predication-average, S==simple-sentence-per cent, from which it immediately follows that S itself is a function of L. A priori we should expect no less than that the shorter sen- tence contains fewer predications, and that as the sentence grows shorter the percentage of simple sentences increases, the limits being respectively single predications in the one case and none but simple sentences in the other. However, inasmuch as no attempt appears to have been. made to verify this prediction, the only statement that I can find bearing on it being exactly op- History of England, particularly the second volume, contains much dia- logue, which might cause us to expect a lower average than is actually the case. The explanation is, that taking the A/7s/ory as a whole Macaulay’s normal style predominates to such an extent as to practically obliterate the ‘‘bearish’’ tendency of the dialogue passages. This can be easily demon- strated. There are in all 45 hundreds of periods whose average is less than 20 words per sentence. These we may take to represent approximately the dialogue portions of the A/zstory. The exact average of these 4,500 periods is 18 62, that is, 4.81 words less per sentence than the average for the entire Fiistory. If we replace these sentences by others of the normal length, we swell the total aggregate of words by 4,500 4.81 or 21,645 words. That is, if the portions of the A/istory which contain an excessive amount of dia- logue were replaced by an equal number of sentences of normal length, the five volumes of Macaulay’s History would contain 41,500 23.43-+-21 645 or 993,990 words. Dividing this number by 41,500, we obtain 23.95 words per sentence, a result not essentially different from the actual average. For the data employed see Univ. Studies, vol. I, no. 2, p. 180; Zb:d., vol. I, no. 4, pp. 351, 352. 244 Variation of Sentence-Constants in Literature 17 posed to it,’ it may be worth the while to examine such data as are available, for the purpose of detecting some relation. For- tunately there is some material at hand with which to work. Among the hundreds of works examined by Mr. Gerwig? for predication averages and simple-sentence-percentages there are some twenty titles which also occur in the list of works subse- quently examined by Miss Whiting* with reference to sentence- length. Both sets of data are presumably based upon the orig- inal punctuation,* thus making a comparison possible. While the number of works thus furnished is not as large as one would desire, it has the advantage of precluding any results which could be attributed to an unconscious bias in case the works had been selected by myself. TABLE II AUTHOR AND TITLE N | PLS L Ay) iChaueers Zale Of: Meltbeus. oes. othe 400 | 5.19 |} 4.5 | 48.0 BU Moxe, zie ofieichard ThE a. o.oo, Pee. 500 | 3.65 | 15.0 | 36.5 C |Spenser, View of Present State of Ireland| 500 | 4.68 | 10.6 | 49.7 D |Hakluyt, Voyagesofthe Eng.NationtoA.| 300 | 4.44 | 8.7 | 56.8 E |Hooker, Aaclesiastical Policy......0.... 500 | 4.12 | 12.0 } 40.9 Bi Siduey, Lefemse- pF Pestana. 3. ek te es ee 479 | 3.98 | 10.0 | 39.3 Gees el er ARIES ree. Gace: ocin siorece +g 'efeh dine 500 | 3.50 | 17.0 | 37.1 Ee EAC OME SSA Y SP acelc as aie sivers att cites cates 1558 | 3.58 | 12.0 | 32.9 er AMON, AA KCOPOR1ELEM oo oa bse ewes os ot 500 | 4.87 | 6.0 | 43.7 MASP Enya LILY WAGE Oo ova wa dhsin 5, 2 so 500 | 3.91 | 10.0 | 37.5 EUS WILE© 2 ACCT, tHE: LUO. weiss 8 ob recs bie 300 | 3.32 | 16.3 | 43.0 L |Hume, Aistory of England ............ 500. | 3.29 | 12.0 | 38.2 VE CI SLIM PI CLLCH SIN) nc torts. can tastes afew aleeces 500 | 2.54 | 26.0 | 28.7 iE (tel as vat hs Keep Sa Oe 716177 a ee ee 500 | 2.57 | 31.0 | 25.9 O |DeQuincey,Confess'ns ofan Opium Eater| 500 | 3.69 | 14.0 | 32.6 P |Macaulay, History of England®........ 41500 | 2.30 | 34.0 | 23.3 Q |Newman, /dea of a University.......... 100 | 4.65 | 8.0 | 41.7 R_ |Emerson, D.vinity School Address...... 100 | 2.14 | 41.0 | 18.0 Seal CI 2CAL- /OOLEMS ca oe snc ciainee 300 | 2.33 ! 40.3 | 15.9 1Gerwig states that ‘‘the proportion between the average number of predications and the percentage of simple sentences is approximately con- stant’? TZynesis, Univ. of Neb (unpublished). 2University Studies, vol. II, no. 1 , pp. 31-44. 8Master’s Thesis, Univ. of Neb., 1898 (unpublished). 4Three works, Latimer’s Sermons, Ascham’s Schoolmaster, and Bacon's £ says, presented such anomalous results that I questioned their correct- ness. A rough count revealed a large discrepancy between Gerwig’s and Miss Whiting’s figures which I attribute to a difference in the punctuation of the texts examined. 5 Bacon’s constants are taken from Hildreth’s paper, and Macaulay’s and Bunyan’s sentence-lengths from Professor Sherman’s, 245 18 R. E. Moritz In the preceding table the figures in the second and third columns are reduced from Mr. Gerwig’s tables, the fourth column is made up from Miss Whiting’s results: N==the number of periods from which the averages are taken. P=the average number of predications per sentence. S=the per cent of simple sentences. [=the average number of words per sentence. If we compare the figures under P with those under L, we see that Macaulay in his history uses almost exactly ten words to each finite verb employed, and so do More, Hooker, Sidney, and Channing, while Chaucer, Spenser, Lyly, Bacon, Milton, Bun- yan, De Quincey, and Newman come approximately within ten per cent of this amount. But there is a marked tendency to depart from the ratio ten when the sentence-length approaches extremes as in the case of Hakluyt, Emerson, or Bartol. Ob- viously, when L is less than ten, the ratio L to P must be less than ten. The data are, however, too limited to reveal much - more than the mere fact of interdependence. The fact of interrelation is perhaps more apparent from a graphical representation as shown in fig. 1. The values of L have been used as abscissas, ten times the values of P as ordinates, and the resulting points have been marked by letters correspond- ing to those to the left of the names of the authors in the table above. The graph shows that four points are approximately in range with the line whose equation is L-10-P. ‘T) while all the points but four lie between the dotted lines, mak- ing an angle of only six degrees with the former. A more striking, though less obvious, relation exists between predication-averages and simple-sentence-percentages. An in- spection of the table convinces one that there is some sort of reciprocal relation, the larger P going with the smaller § and vice versa, but much more than this is not warranted by the figures in our table. Nor should we expect the table to reveal any very definite relation, when we consider the uncertainty of the numbers involved. There is a large probable error in all 246 NU BS Na ER Poot 5 2: EE : : OW) | | a & Be a | ne STA ae ee ns Oa a Cte Oe Pe PO a rie Fads rhea. 8 Fe eae seypeetend Variation of Sentence-Constants in Literature 19 the figures employed, but the magnitude of the probable error in the S’s can only be appreciated by a glance at the figures from which some of the averages are gotten. Thus we find Lyly’s 17 to be the mean of the numbers 26, 14, 20, 15, 8,1 these numbers representing the simple-sentence-percentages for consecutive hun- dreds. Similarly, De Quincey’s 14 results from the numbers IO, 19, 15, 7, 21.” If larger averagings had been made, essen- tially different results would have been obtained in many cases as the following example shows: VARIATION IN SIMPLE-SENTENCE-PERCENTAGES FOR CONSECUTIVE HUNDREDS IN Spenser’s View of the Present State of Ireland :8 Average of 300 sentences yields 12.0 per cent simple sentences ae ce ee ce ec iad “ce 11.5 « ce 500 “ ce 10.6 “ec <¢ “cc “c ce 600 “ec sé 10.5 ec “<é sé a ce 700 “cc “se 9.7 6é “ce “ce “cc “cc 800 “i 73 9.0 “e “cc “eo “cc a 900 ce “< 8.9 “cc “a «é «< se 1000 “cc “cc 8.0 “c sé “ac Even so great a stylist as Macaulay gives averages for S so widely divergent as 41 and 27 for different parts of his History of England, though each average is based upon 500 consecutive periods. * . Although disastrous for immediate progress, these facts did not lessen my confidence in a more definite law than that which was thus far apparent. The first problem was how to reduce the probable errors to manageable proportions. Of course, the obvious solution consists in making larger averagings, say, of 10,000 or more sentences from each work, but this involves an enormous amount of unattractive labor. To avoid this, I de- vised a seemingly crude experiment on Mr. Gerwig’s figures for English prose works.® I struck averages of both P and S$ for 1\Wniversity Studies, vol. 11, no. 1, p. 38, 2Jbid., p. 34. 8 Jbid., p. 42. 47bid., pp. 25-26. The numbers represent the averages for the 27th to the 33d, and for the 284th to the 289th hundreds respectively. ‘Jbid., pp. 31-44. These figures comprise averages of about 60,000 periods, taken from 71 different authors, 247 20 R.E. Moritg each five hundred periods from the same work, and then grouped together all the averages for which the predication-average is between 1.50 and 2.00, similarly those for which the predica- tion-average is from 2.00 to 2.25, from 2.25 to 2.50, etc. We thus get the following table: TABLE III 1 P=1.50-2.00 =F D022 1D P=3.25-3.50 [2a h eS Pa |S) Peres — Ascham, ......| 3.49 19 Macaulay..... 1.88 48 ; 2 Coleridge ..... 3.33 19 Symone: !.25..0 2.88 (a8 WOR a ee oat Gaels, tal elaaeraeies amt 3.43 | 16 a ES kin tet ay ees 2.51 31 Hume:.t2- 5; 3.29 12 On aie ar ae 2.69 30) Huxley 2.3 sialleorse 16 P=2.00-2.25 || Darwin .......] 2.64] 24 || Moore.........] 3.38 | 11 | a Deva ee eel eda 27 Ruskin ........ 3.50 18 PySoy | Rises, eee: 2.69 | 20 || Scott.........-. 3.36 | 16 Bartal, 210 44 Janiuss.5. ee 2.54 26 P=3.50-4.00 RO One ote 210) 44 Lowell’...25::. Boone ae Pils ee EGHae Combe 2.10 CS ee es J Blaine......... 2.23) 39 Sao ee gait ot tl Rddizenaye 3.67| 12 Channing..... 2.1 0 sige eau : s Barrow? «sc. cee 3.74 20 Cope tee 2 204 44 Shaftesbury...| 2.61 28 Bolingbroke ..| 3.65 14 Emerson ...... 2.14 38 Ee Cor ou eT cl SL MSE ee a Ee 3.65 14 Macaulay...... 2.22 32 Bunyan .......} 3.91 10 COST Fe 2.22 32 De Quincey ...| 3.69 14 Bon Gece aies ore - a hbge a3 Sheen Bs a poWat o siatotelors ¢ z2 Ves ee 3.6% i) “ 2.18 | 40 P=2.75-3.00 || Tyty rT 3.50 | 17 “ 2.17 36 Pa aatenee MGT eerioatne 5.65 15 i 217 36 ee ore, seeeee | 3.98 10 Ceae pir | a6 || Amold........ 2.1 | -20. |levery 1200 | ale apd Say 2.17 | 36 || Browning. ...-) 2.91 | 25 || wordsworth ..| 3.87] 17 Phelps ........ A as a | Pei 2.88) 30 | ———— pope eS ES |e el Geotge ya 2.92 | 23 ‘Ss Se ee Te ta hes pemldemith = yl ode“ ls ogy =4.00-4.50 Hamerton..... 2.85 20 Pyves P=2,25-2.50 Dippiasen tees] 2.855) 21 SSS SS —<—_— WSOP 6 ties wel 2 OS 26 4.22 12 P | Ss || Newman ...... 2:97| 16 Halas pare 412| 4 Shakespeare ../ 2.76 | 31 || steele ...11.1/|} 4102 | 10 TEAM ree rinie’s 2.86 Dl peta ieee Emerson ie | eens ea oad go See Chaucer ....... Aa 8 % 5 a | PRE aOR co lero e core ome eel Doty Re hh me of : 8 32 ————— } Ba P=4.50-5.00 32 P=3.00-3.25 ie James <1001..0.] 2545 | 34 ps || Dryden...) 4.89] 6 Macaulay ..... 2.31 | 32 Latimer.......} 4.75 | 18 hes eaten oe 2 31 39 Muilton's<. J2cee- 4.87 6 Bee mete seat ae 32 || Bacon......... 3.12} 19 a | Sabet 26 32 Carlylessoacactie Ole 18 —5 O0-F SRD aaa 2.26 32 Franklin....., 3.04 19 fa nits 2.26 |- 82 || Holland>......| -3.03| 21 Pass: SE ESAS ese 2.26 | 32 || Johnson....... Sees a6 pees mr Pace rete 2.26 32 Mandeville....| 3.08 22 Chaticers =. | oven. 4 Phillips. ......] 2.47 53 Stevenson..... 3.01 24 Spenser.... ..| 5.44 8 BeLLey con wenn 2.48 26 White pont 3.15 15 eC MAR GS 5.44 8 ETA Variation of Sentence-Constants in Literature 21 Let us now average the P’s and S’s of each separate group and we obtain: TABLE IV PREDICATIONS AVERAGES cal ra PER SENTENCE Z BETWEEN ———— | —$——___ | 1 1.50 and 2.C0 i > a age es 0 Sg 7 5 le eats hee DS oho, wae Wi] Dye Beg tO ee POS Pe B82 ef =8 00's |: 2" Tall bites Jol (8 Vesa TIM es Es aaa 3 fs Relea Pod eet) Se) ome Sule. OO” wlan 4 Oreo: Or) A OO, Ae ()P ch Ae TOs 40s 500) 4S a: 141395 GOs +£5.01 Oe COR AWE Owe AENAODOONKK GO bt we) Lo PYsS 13.54 13.38 13.41 13.33 13.87 13.59 13.52 13.55 13.19 13.94 13.92 The average values of P and S which we have thus arrived at, aside from the general reciprocal relation already referred to, manifest a uniformity which can not possibly be attributed to chance. Take the square root of 53.0, the first number under S, and multiply it by 1.86, the eae number under P, result: ee ee Next take the square root of 39.1, the second number under S, and multiply it by 2.14, the corresponding value for P, and the result is again 13.+- Proceeding in like manner with the follow- ing pairs of numbers we obtain N Nw N .861/53.0=13.+ -14//39.1==13.-++ 341 32.9=13.+ .6274/25.9=13.-+ 4847/8 .3=13.-- 5.38//6.7= The exact values are found in the last column of our table. short, we have quite uniformly 249 meal te In 22 R.E. Montz Py SSC (2) where C=-13.57 the mean of the slightly varying values 13.+- How nearly this equation fits our data may best be seen from the graphical representation in fig. 2. The curve pe JESCBT See : Yas ) as well as the P’s and S’s from our table have been plotted, by using the values of S as abscissas and ten times the correspond- ing values of P as ordinates. The resulting points have been numbered 1, 2, 3, etc., to correspond with the index numbers in the table. The relation symbolized by equation (2) may be easily stated in words. For, if P,, S,, and P,, S, are any two pairs of corre- sponding values of P’s and S’s, we have by virtue of (2) Ly Sie s 57=Pry|S2 nearly, hence Py P2=/ S21: VS, that is to say: The predication-averages of various works are inversely pro- portional to the square roots of their simple-sentence-percent- ages.* ! 1This law, owing to the meagerness and uncertainty of the data upon which it is based, must be considered no more than a rough approximation to the truth. There is no reason to suppose that it will give with equal accuracy the true relations between the values Pand 5S for every other work. Of the other common sources of error in inductive reasoning, biased data, and reasoning from accidental common marks of a limited number of samples to a property of the class, the first I think has been eliminated by my mode of selecting the data. As to the other I can only say that the mathematical probability is practically zero that each of eleven pairs of numbers chosen at random should manifest a definite relation like the one in question. I have computed the limits between which each value of S could vary and yet satisfy the relation P\/S=13-+ for the given values of P, The numbers in bold face are the .S’s from our table; immediately above 250 we r EA AEST Sea i v ? ay Variation of Sentence-Constants in Literature 23 It will be interesting to test this law on some particular work not included in our table of averages. Macaulay’s History of England is the only work available for this purpose, for it is the only work whose constants P and S have been determined with sufficient accuracy, without tampering with the composition or punctuation of the author. Using S==34.2,1 that one of the two constants P and S which is most readily determined by count, our formula gives ee 184. 2,32, ay 34.2 and below are the limits within which they could vary and still satisfy our law. Upper limit....... 56.7 | 42.8 | 35.8 | 28.6 | 23.7 | 20.4] 17.1] 14.4] 11.3] 8.4] 6.8 Ss 53.0) 39.1) 32.9) 25.9) 23.2) 19.2) 15.9) 13.4) 10.0) 8.3 | 6.7 Lower limit.,..... 48.0 | 36.8 | 30.8 | 24.6 | 20.3 | 17.5 | 14.8] 12.3] 9.7] 7.1] 5.8 Now not only are the limits comparatively narrow, but in most cases S occupies a mean position between them. I am aware, howeyer, that strange numerical relations may occur where there is no law. In evidence of this J take liberty to quote an example from C. S. Peirce’s essay, 4A Theory of Probable Inference, published in the Johns Hopkins University Studies of Logic, an essay which every one who ventures upon the field of scientific induction would do well to peruse. The first five names of poets and their ages at death, taken from a certain biographical dictionary are, RABAT rh odasees 48 - A DEIM ER, Me trace ce es 76 Abulolasy wecesase cee 84 Abunowas.............] 48 PICCOTGS soicic ce witeereeas lle” Now, although no sane person would expect a law connecting the digits of the numbers representing the ages at deaths of poets, it is nevertheless true that for the given five numbers, 1. The difference of the two digits, divided by ¢hree, leaves a remainder of one. 2. The first digit powered by the second, the result dividel by ¢hvee, leaves a remainder ove. 3. The sum of the prime factors (including one) of each number is divis- ible by three. 1Gerwig gives 34. But Gerwig gives only the integral parts of the values for S, while the values for Pare given to two decimal places. To be con- sistent, he should have either retained the first decimal places in S, or have dropped also the second decimal places in P, 251 24 Rk. E. Moritz which coincides nearly with the value 2.30, determined by the laborious process of actually counting the finite verbs in over 40,000 periods. I now offer the following as a tentative statement of the Sher- man principle: I. Good writers manifest their individuality by the uncon- scious use of certain average sentence-proportions which remain nearly constant throughout a given form of composition. 2. These sentence-proportions vary widely, not only in various authors, but also for different forms of composition employed by the same author. 3. These proportions are interrelated, and in some instances at least thew laws of dependence may be mathematically ex- pressed. The principle as now stated opens an almost unexplored field for investigation. The present inquiry has been limited to the constants L, P, and S, but there are many other sentence-pro- portions, such as the percentages of various kinds of ,clauses and phrases, the relative abundance of adjectives, conjunctions, and other parts of speech, the ratio of Anglo-Saxon derivatives to those derived from other languages, the average length of the words employed by an author, etc. Are these proportions more or less variable than those already considered? Which of these are independent, and which are interrelated, and by what laws? Which of these are most persistent and hence most character- istic of a given author? These and many other questions press for answers. Concerning the constants L, P, and S many questions remain open. We have seen that L depends at least upon two factors, the author’s sentence-sense and the form of composition em- ployed; but are these the only factors upon which it depends? If so, can the element of variation in L due to the form of com- position be determined and the residue, due to the author’s sentence-sense alone, be set free?’ Can the relation between sen- tence-length, sentence-instinct, and sentence-form be as defi- nitely expressed as we expressed the relation between P and S? 252 Variation of Sentence-Constants in Literature 25 These are problems which it seems to me need not remain unsolved. The greatest immediate need is for more abundant data. In the data thus far obtained it has been assumed that 300 periods will furnish with comparative accuracy the constants sought, and in but few instances have more than 500 periods from single works been examined. Neither the probable nor the actual error arising from this hypothesis has been determined. In order to get some light on this question I have computed the following table from averages of 400 periods and 500 periods each of 50 authors :* TABLE V AVERAGE DE- LIMITS OF Re aEACRS eeOM FITS VIATION IN THE} DEVIATION PROBA- AVERAGHS 4 ORs b eo waren aca one BLE Dod Positive | Negative] ERROR PERIODS Per cent Percent | Percent| Percent Sentence-lenoth 5.0.5 5.26: 2.35 9.5 6.3 1.6 Predication-averages ......... 2.06 4.4 9.7 1.9 Simple-sentence-percentages. . 5.98 28.8 t.2 5.3 This table shows that the constants L, P, and S obtained from an average of 400 sentences, will differ by 2.35 per cent, 2.06 per cent, and 5.98 per cent respectively from the values of these constants based upon averages of 500 sentences, and in- some cases variations amounting to 9.5 per cent, 9.7 per cent, and 28.8 per cent respectively may occur. We should, therefore, even by examining 400 sentences from each of 50 authors, obtain less than 2 per cent accuracy in the constants L and P, and about 6 per cent accuracy in S. The numbers in the last column show the probable deviation between 400 and 500 sentence-averages for any other work whose constants have not yet been deter- mined. It appears, therefore, that so long as averages are based upon 500 or less periods the fractional parts in L and S may be entirely neglected, and in P everything after the first decimal place may be disregarded. 1 The table is based upon Miss Whiting’s results for Z and Mr. Gerwig’s tables for Pand 5S. 253 IV.—On the Errors in the Methods of Measuring the Rotary Polarization of Absorbing Substances BY FRED J. BATES In the half-shade polariscope the settings are made with the two halves of the field of equal intensity. Let 47O and MO, fig. 1, represent the directions of the vibrations of the wave-lengths in the three similar beams as they emerge from a transparent substance in a magnetic field. Let 4 be the amplitude and A; to X,, the wave- lengths, A, being the shorter. Let AZ be the direction of vibra- tion after passing into the analyzing nicol, anda, to a the angles which the vibrations in each half of the field make with the normal OH. We then have as the condition for equal illumination of the two halves of the field: (Ay sin a,)?-+ (Ao sin az)?+------+ (4,4 sin an—1)°>+(An sin Oy, a= (Aisina, )?+-(Aesin a,,)?+----+ (A,_1sin a2)?+ (4, sin a1)”, (1) where each term is the intensity of the wave-length as it reaches the eye. Consider any two wave-lengths, A, and X,, fig. 3, and let them have amplitudes 4, and 4,. Then (A, sin as)?+ (A, sin a, )?=(A, sin a, )*-|-(A4, sin as)?, (II) Let 4,°,=K A,’ (IIT) then ae ie ae reas -L sin? 0, = + sin? a,, (IV ) since, by transposing, (IV) gives 1/x(sin’a, — sin? a,)==(sin? as — sin?a,), (V) 255 2: Fred J. Bates which can be true only when AK=1. Consequently (V) can only be made to hold, and still maintain the condition (III), by chang- ing the angles between the directions of vibration and the normal, that is, OH takes the position O77’. : Hence (I) for light of any amplitude becomes (Aisin [a1--8a])’-+ (4esin [og+8a] )?+----+ (4,_; sin [a,-,+8a] )” +(A, sin [a, 60] )?= (Aisin [a,, 48a] )?+(Aesin [a,_4+ 8a] )?+----+- (4,4 sin [oo ba] )? + (A, [sin a= 6a] )? (VI) AG He Be Fig. I. YOM -E" E A FS E Ve EY’ Let us now consider a specific example and solve for da. As the source of light take the wave-lengths from 566 mp to 580mm cut from the solar spectrum. (aj-++a,) (fig. 1), which is the angle be- 256 Errors in Methods of Measuring Rotary Polarization 3 tween the direction of vibration in the nicols of the polarizing system, is obtained by direct measurement when there is no rota- tion of the light.!_ We thus find (a; --a, )==20.4'; Also for Ain pp Rotation 566 ATT 580 4.53° That is (a,—a, )=14.4' 7. @==17.4! ay =a Considering now the luminosity curve, S, of the sun? and the transmission curve of a fuchsin solution,’ 7, in fig. 2, and taking m=5 we have TABLE I PER CENT TRANS- N - NIN pp a IN MINUTES pinreeas ake ae MISSION IN FUCH- SIN SOLUTION. 580 17.4 86 21.7 576 13.3 94 10.5 573 10.2 99 5 570 ies 103 2.3 560 3.0 104.5 1 where A==573 pp for example is taken as the mean intensity of the interval from 571.5 pp to 574.5 pp. By successive approximation (VI) is solved with sufficient accu- racy, and we obtain by considering only the luminosity curve: 1The analyzer was set for blackness of one-half of the field and then rotated to a similar position for the other half. The angle of rotation is (aa, ). i ‘ 3Marscart, Op/ique, I, p. 104, pl. I. 8This curve was kindly determined for the writer by Professor B. E: Moore. The concentration, parts by weight, was 0.000024. 257 4 Fred J. Bates 86 sin? (17.4’-+.4')-+-94 sin? (13.3'-+.4’)-+-99 sin? (10. 2’-+.4’) + 103 sin? (7.1'-++.4’)-+ 104.5 sin? (3’+.4°)= 86 sin’ (3’'—. 4’) --94 sin? (7.1’—.4’)-++-99 sin? (10. 2’—.4‘) + 103 sin? (13.3—.4)-+101.5 sin? (17.4—.4) (VII) Hence, with a transparent substance, giving a rotation of about 4.77° for 566 wy, the normal to the principal axis of the analyzing nicol, for the light source considered, will be rotated from the median position OH through an angle, 8a—. 4’ (—=.006°) to a posi- tion O/F7’. In a similar manner we now change the coefficients of equation (VII) to correspond to the luminosity curve resulting from passing sunlight through a fuchsin solution (see Table I) and we get approximately 86 X 21.7 sin? (17.8’—4. 1’) +94 X 10.5 sin? (13.7’—4.1/) + 99 5 sin? (10.6'—4. 1’) +103 X 2.3 sin? (7.5 —-4.1')-+ 104.5 X1 sin? (3.4'—4.1')= 86 X 21.7 sin? (2.6'+-4.1')+94 X 10.5 sin? (6.7’°-+4.1/)+ 99 <5 sin? (9.8'-+-4.1') +103 X2.3 sin’ (12.9'+4.1') + 104.5 1 sin? (17'+-4.1’). Hence the normal moves through an angle of ‘4.1’ from OH" to OH”. Thus it becomes evident that the change in the luminosity curve of the light reaching the eye, due to the absorption of the fuchsin solution, gives with the half-shade system an apparent rotation of the plane of polarization of approximately 4.1’ (0.065°) under the conditions assumed above. Under the same conditions the magnitude of this effect observed experimentally was 0.055°. Errors due to the change in the luminosity curve similar to the one just discussed also enter in a proportionately greater degree into Wiedemann’s* and allied methods of measuring the rotation of plane polarized light in and near an absorption band. 1G, Wiedemann, Fogg. Ann, 82, p. 215. 1851. 258 it oe x —* t Errors in Methods of Measuring Rotary Polarization 5 In these methods there is used as a measure of the rotation the displacement of either a black space or a clear space in a spec- trum. The former is produced by a quartz plate cut at right angles to the optic axis and an analyzing nicol; the latter, which was first used by A. Schmauss,* by an additional plate of selenite, “ which gives a channelled spectra with the interference bands absent for those wave-lengths whose planes of vibration coincide with one of the principal axes. From space to space in these methods is a rotation of 180°, and the maximum sensibility seems to be obtained with about 1.5 mm. of quartz between the _ nicols. The working width of the space thus obtained depends upon the observer, but is assumed to be about 40 pp. Both methods are essentially photometrical since the position of the space is determined by the condition that the light at equal dis- tances on each side of the central region be of the same inten- sity. Even if the clear space is located by having an equal num- ber of the interference bands on each side of the central region, the process is nevertheless photometrical because the bands do not entirely disappear at any portion of the space, and the bands at which the observer ceases to count must be of equal intensity. When either of these methods is used to measure the rotation in the edge of an absorption band, there is a shift of the space referred to, and consequently an apparent rotation of the light due to the absorbing substance. Consider, for example, the dark space method. Let the space be produced by 4.5 mm. of quartz and the direction of vibration of wave-length 589.6 uu be exactly at right angles to its direction in the analyzing nicol. Under this condition S”, fig. 3, is the luminosity curve of the sun as it reaches the eye; and F’ the same curve modified by the insertion into the path of the light of a 1 cm. length of the alcoholic fuchsin solution 0.000024. The ordinates of S’ are obtained by resolving the direction of vibration, giving it its relative inten- sity from curve S, fig. 2, along the direction of vibration in the analyzer. F’ is obtained by giving each ordinate of 5S” its relative transmission from curve F, fig. 2. It will be observed from the fact that the luminosity curve of the sun is not a horizontal 1A. Schmauss, Ann. d. Phys. 2, p. 280. 1900, 259 6 Fred J. Bates straight line that, even on the curve S, the center of the space, no matter what its width, is not on wave-length 589.6 wp, which is the point of zero intensity, but at some distance to the left. Really this is the only wave-length that is completely absent, but for several pp on each side of it the eye can detect no change in the intensity of the blackness and so is compelled to judge by means of points of equal intensity on each side of the central region. The space is thus located with an apparent center to the left of 580.6 pu. Assume that the boundary of the black space is defined by the wave-lengths 582 wp» and 594 pm, which are of equal intensity, thus giving it a space with a working width of I2 pp. ; With the transparent solvent in the field it is evident that the position of the black space will be determined as though its center is at C (fig. 3). When the fuchsin solution replaces the solvent, the center, passing to curve F, is found to be at C’, or a change of position of 0.7 wy. This gives a rotation of 0.6° when there has been no change in the position of the planes of polari- zation. The conditions assumed above being very similar to those used by various investigators were easily realized experimentally. An auxiliary cell with the fuchsin solution was thrown into the path of the light and the analyzing nicol rotated until the black space was in its original position. The following data were thus obtained for this part of the spectrum: FUCHSIN SOLUTION our IN DIFFERENCE 16.00° 16.60° 0.60° 15.95° 16.43° 0.53° CALCULATED ROTATION OBSERVED ROTATION 0.62° Ne 0.56° : The above is also just as applicable to the clear space method. 260 oe Noa <: Errors in Methods of Measuring Rotary Polarization 7 Such close agreement is of course partly accidental, since con- secutive measurements can not be made by Wiedemann’s method closer than 0.10°. The magnitude of the error will depend upon what width of space the observer uses. As the shift of the space was toward the red, thus giving an apparent positive rotation, it is evident that as the space moves farther into the absorption band of the fuchsin this positive rota- tion would increase until it attained a maximum in the edge of the band. From there on it would gradually decrease and be- come zero at some point in the interior where the absorption is the same for both edges of the space. As the space approaches the other edge of the band from the interior a gradually increas- ing rotation would be observed, attaining a maximum in the edge of the band, and then decreasing to zero as the space leaves the band. This maximum is, however, in the opposite direction from that previously mentioned. Since the former is a positive, this is a negative rotation; that is, for solutions it would drop below the rotary dispersion curve of the solvent. When the half-shade is used the apparent rotation due to the error increases very perceptibly with, but is not directly propor- tional to, an increase in the strength of the magnetic field, be- cause a doubling of the field gives a corresponding increase in the angle (ai—a»), all other conditions remaining the same. In the black and clear space methods, however, the maximum value of the error varies only very slightly, for small rotations, with the intensity of the field, because the rotation of the sub- stance studied constitutes only a small fraction of the total rota- tion. The thickness of quartz necessary to give the desired space in the spectrum is at least 1.5 mm. This gives an initial rotation of 32.5° at the D line, to which is added the rotation, for example, of 1 cm. length of water for a field of 8000 lines, making a total 34.2°. If the field be doubled the total becomes only 36°, which is sufficient change to affect the maximum error only slightly. In the three methods mentioned the maximum value of the spurious rotation will be nearly directly proportional to the con- centration for dilute solutions, and the sharper the absorbing band the more marked the anomaly. 261 8 Fred J. Bates These apparent rotations, however, are not entirely due to the simple cutting down of the amplitudes of certain wave-lengths by the absorbing substance. Superimposed upon this is the Purkinjie effect, or change in the luminosity curve for different intensity. This effect is ordinarily negligible, but under certain conditions it may become considerable. This, we are led to sus- pect by a study of Ko6nig’s* luminosity curves for gas lights at different intensities. Similar curves for sunlight are apparently lacking, so that no calculation bearing upon this point could be made. With the wave-lengths from 573 py to 587 wy as the source of light and a rotation of 4°, the setting of the analyzer differed by 0.02° when the light intensity was reduced 44 times by the rotating sector. The angle between the principal axes of the nicols in the polarizing system was several minutes of arc. In fig. 2, R is the luminosity curve resulting from passing sun- light through a I cm. thickness of alcoholic fuchsin solution, concentration 0.000024. It is obtained by modifying the ordi- nates of the luminosity curve S by the transmission curve F, Whenever the difference in the rotation of two substances is be- ing measured, or either the optical center of gravity of the light source is being determined for absolute measurement with the half-shade system, or the position of the observing telescope is being calibrated in wave-lengths for either the clear or the black space methods, the errors due to the use of two luminosity curves of which FR and S are a specific example always enter the mea- surements unless proper compensation be made. These curves may be almost identical for transparent and vary greatly for ab- sorbing substances. In the former case the errors are ordinarily negligible; in the latter they may become very large. From the theory it is evident that the angle will be a maximum at those points in the spectrum where the comparative slope of the two curves is such as to give the greatest difference in the intensity of the longest and shortest wave-lengths reaching the eye. Con- sider curves S and R with the half-shade system. At 600 pp the error would be approximately the same as that calculated in equation (VI), (0.006°), since the conditions are about the same. 1 Beitrage zur Psychologie und Physiologie der Sinnesorgane. 262 Errors in Methods of Measuring Rotary Polarization 9 At about 567 wp» it would seem to be a maximum, because here the variation in the intensity of the extreme wave-length used will be greatest. At some point in the interior of the band it ‘will become zero. In passing from the absorption band the error will not be nearly so marked as it was upon entering. Curves L and H, fig. 2, obtained experimentally, are given to show how the error changes in this particular instance. Their ordinates are differences in the rotations, one small division representing 0.005°. Curve L is for the alcoholic fuchsin solution 0.000024. H is the same with a concentration of 0.000012. A cell of water ‘giving 4.5° rotation at the D line was between the poles of the magnet. The angle between the elements in the half-shade sys- tem was 1.10°. The intensity of the light source was 14 pp. A 1 cm. length of the solution placed outside the magnetic field was thrown in and out of the path of the light and the differ- ences in the rotation of the water noted. The effects of bleaching must be carefully guarded against. After-sunlight has been passed for five minutes through solutions of such substances as fuchsin and cyanin the absorption is greatly _diminished. Much careful work has been done in determining the optical center of gravity of different light filters* for various parts of the spectrum, and the calibration is done with substances with no particular selective absorption. These filters let through light differing in wave-length by not less than 34 pw or 40 pu. Hence -a slight difference in the luminosity curves will give, with the half-shade system, a comparatively large error in the rotation. When the non-absorbing substance was in the field, a rotating sector was experimented with to equalize the intensity of the light ; but, since this diminishes the amplitude of all waves equally and changes the luminosity curve only in proportion to the Purkinjie effect, the spurious rotations were not eliminated. The insertion into the path of the light, outside the magnetic field, of an optical thickness of the substance equal to that in the field in such a manner as to compensate for the unabsorbed beam, 1H. Landolt, Das optische Drehungsvermogen, p. 387. Chr. Winther, Zeilschr. Phys. Chem., July, 1902, p. 161. 263 IO Fred J. Bates seems to be the simplest method of keeping the luminosity curve unchanged. The elimination of these effects should in many cases bring about a more perfect agreement, especially for solu- tion of the measurements of different investigators on the rota- tion of plane polarized light, both natural and magnetic. The opportunity is here taken to thank Professor D. B. Brace for his helpful suggestions. 264 V.—The Magnetic Rotary Dispersion of Solutions of Anomalous Dispersing Substances* BY FRED J. BATES Becquerel? in 1880 observed the anomalous rotary dispersion in oxygen. Others have failed to verify his results. In 1898 Macaluso and Corbino® observed it in sodium vapor. They found that the rotation increased with great rapidity as the absorption band was approached, attaining a maximum in the edge of, and dropping to a small positive rotation in the interior of, the band. Zeeman? has also studied this phenomenon and finds for sodium vapor, not too dense, a negative rotation in the interior of the band. Biot,® Arndsten,® Landolt,’ Nasini® and Gennari, Cotton® and others found anomalous rotary dispersion in optically active liquids, and in 1896*° Cotton observed a tendency toward what he interpreted as an anomaly in ferric chloride, copper acetate, and several other liquids, when they were placed in a magnetic field. He found as an absorption band of a certain solution was approached, going from red to violet, an increase in the rota- tion above what would have been obtained had the solution been transparent. In the solution of some other substance in which he could approach a band from violet to red he found the rota- 1Read before the American Association for the Advancement of Science at its Washington meeting, 1902-3. 2H. Becquerel, Compt. rend. 90, p. 1407. 1880. 3 Macaluso and Corbino, Compt. rend. 127, p. 548. 1898. 4P. Zeeman, Proceedings Royal Academy of Amsterdam. May 31, 1902. 5M. Biot, dan. de chém. et de phys. (3) 10, p. 5. 1844. 6M. A. Arndsten, Ann. de chém. et de phys. (3) 54, p. 403. 1858° 7H. Landolt, 2e7b/. 5, p. 298. 1881. 8R. Nasini and G. Gennari, Zeitschr./. physik. Chemie 19, p. 113. 1896. ®Cotton, Ann. de chém. et de phys. (7) 8, p. 347. 1896. Cotton, L’£clairage electrique 8, p. 162 and 199. 1896. 265 2 Fred J. Bates tion dropped below the normal. This he interpreted to indicate two points of inflection in the rotary dispersion curve for each band. He never succeeded, however, in establishing a curve for both sides of the same band, nor in the interior of the band. In 1900 Schmauss* observed an apparent anomaly in fuchsin, cyanin, naphthalene-red, eosin; and didym glass. He studied the first four in very dilute alcoholic solution, the greatest concentration by weight for each substance being 0.000024, 0.000039, 0.000005, 0.000004 (parts by weight) respectively. His rotary dispersion curves for these liquids all show four points of inflection, and a very marked rise above the curve of the solvent as the violet region is approached. The fuchsin solution 0.000024 gave a rotation of 2.81° for wave-lengths 544 mu and 2.34° for 545 mp. The former is 0.44° greater than that of the solvent, and the latter 0.13° less. For wave-length 450 pp the solution gave a rotation of 0.58° greater than that of the solvent. In subse- quent papers* he studied didym glass, liquid oxygen, and solu- tion of litmus, anilin blue, and three rare earths. Schmauss’s results indicate that the anomalous effects increase with increasing concentration ; that the maximum rotation is inde- pendent of the field strength, and that the negative rotation in the interior of a band decreases with increasing field strength. The observations of Zeeman on moderately dense sodium vapor as well as all of Schmauss’s results seem to confirm the theoreti- cal deductions of Voigt. The writer has been studying solution of anomalous dispers- ing substances and thus far has not succeeded in observing an anomaly. The substances were fuchsin, cyanin, anilin blue, and litmus. The first two were in alcoholic, the latter two in water solutions. The method employed consisted in placing a tube of the solu- tion between the pierced poles of an electromagnet. Light par- allel to the lines of force was successively passed through a half- shade polarizing system, the tube, and an analyzing nicol. After measuring the rotation, the tube was removed and the rotation 1A. Schmauss, Aun. d. Phys. 2, p. 280. 1900. 2A. Schmauss, /d7d., 8, p. 842, 1902, and 10, p. 853. 1903, 266 Magnetic Rotary Dispersion of Solutions 3 of a similar tube filled with the solvent noted. Since the tubes are the same length the difference between the two rotations is the rotation, for the wave-length of light used, due to the pres- ence of the dissolved substance. The source of light was the sun. The desired wave-lengths of sufficient homogeneity were obtained by means of a spectral system similar to that described by Doubt.* _ The instrument used in making the measurements was a sensi- tive-strip spectropolariscope.t_ The polarizing system consists of two thin pieces of Iceland spar mounted in a cell of a-mono- bromonaphthalene and transmitting the ordinary ray. The smaller of the two pieces covers but half of the field of the larger one, and the angle between their planes of polarization can be varied at will. Both the color and polarizing systems will be treated in detail by the writer in a paper soon to be published. The conical pole-pieces of the electromagnet used tapered from 20 cm. to 5 cm. in diameter. With 23 amperes at 500 volts and the pole-pieces separated 17 mm., a field of over 18,000 lines per sq. cm. was obtained. This intensity was used with the litmus and anilin blue solution, but was allowed to drop to 15,000 lines in studying the fuchsin and cyanin. This was sufficiently near saturation to prevent the slight fluctuations of the current inter- fering with the setting of the polariscope. A star-shaped brass frame, fig. I, rotating on pivot bearings, was rigidly supported between the poles of the magnet. The tubes rested in the V-shaped slots and were held in place by spring-back clips. A slight rotation removed one tube from the path of the light and substituted another. The tubes were I cm. in length and closed with cover-glasses each 0.2 mm. in thickness. Since we compare one tube with another it is necessary that their optical paths be the same. Factors tending to make their paths unequal may be introduced, first, by the tubes not being of the same length; second, the brass frame failing to keep their axes parallel; third, the cover-glasses not being of the same thick- ness. These difficulties were all eliminated by surrounding the 1T. E. Doubt, Pril. Mag. Aug., 1898. 2D. B. Brace, Zbid. Jan., 1903. 267 4 Fred J. Bates frame with a stationary glass cell (100 mm. * 50 mm. * 12 mm.), fig. I, filled with the solvent. If, now, the tube containing the solution be rotated into the path of the light, it displaces solvent equal to its own length. The tube containing the solvent serves merely as a carrier for the cover-glasses, which compensate for those closing the tube containing the solution. If the cover- glasses are not of the same thickness the only change that can occur in the optical path will be caused by replacing of the extra thickness of glass by the solvent when the next tube is thrown in. A difference of 0.02 mm. in the thickness of the cover-glasses would give an error of 0.004° in the rotation. Thus it is evident no mechanical errors can appreciably affect the measurement. Fig. l In the usual methods for measuring the rotation of plane- polarized light there appear in the observing telescope, when the magnet is excited, several partially overlapping images of the field. This is due to the successive internal reflections of the beam of light within the tube. When the analyzing nicol is set for the extinction of the ray that has passed through the liquid 268 Magnetic Rotary Dispersion of Solutions 5 but once, the images produced by the rays that have passed through it, three, five, etc., times, become relatively bright, since the rotation is proportional to the length of the liquid traversed. An indistinctness results. These reflected rays were eliminated from the field of the observing telescope by making the ends of the stationary cell slightly prismatic. With such a sensitive polarizing system all glass, even when quite thin, shows more or less double refraction, and the elimina- tion of this becomes a serious problem. By examining a great many microscope cover-glasses a sufficient number of optically good ones was obtained. The thinner the glass the freer it is likely to be from strains; but if it is less than 0.2 mm. in thick- ness the pressure of liquid against it will produce not only a lens effect and destroy the adjustment of the optical system, but also make the glass doubly refractive. In constructing the tubes a cover-glass with a diameter 2 mm. greater than that of the tube was laid on a level surface and the tube placed on it. Without touching tube or cover-glass, either beeswax or a mixture of fish glue and glycerine, depending on the solvent, was flowed around them. In this way tubes were finally obtained whose end-plates showed practically no double refraction. The glass on the sides of the cell containing the tubes being 1.5 mm. thick, holes were bored and windows of cover-glass were mounted in the manner just described. The analyzing nicol could be rapidly rotated when necessary. The setting, however, was made by means of an accurate screw carrying a drum graduated to 0.005°. A tenth of the scale divi- sions representing this quantity could be readily estimated by the unaided eye, With the apparatus as described and the magnet excited, suc- cessive settings for different tubes could be made in a few sec- onds, and the field was neither distorted nor contained any ex- traneous light. Successive settings on a non-absorbing solution could be made whose extreme values differed from each other by less than 0.007° ; and with the sensibility diminished to give sufficient light to read through the absorption band itself of the liquids studied, the error was less than 0.01°. 269 6 Fred J. Bates . In comparing a non-absorbing with a highly absorbing solu- tion the difference in the intensity of the light reaching the eye is very great. The eye is thus unfitted for rapid setting from the non-absorbing to the absorbing substance, and it is not de- sirable to allow a considerable interval of time to elapse between such observations. In order to overcome this and also to elimi- nate spurious rotations which are discussed in a succeeding paper, an auxiliary cell of the solution with the same absorbing power as the solution tube was placed before the collimating slit and outside the magnetic field. By means of the lever the observer threw this cell into the path of the light simultaneously with the tube of solvent. The luminosity curve, as well as the intensity of the light reaching the eye, was therefore the same for both tubes. The tubes, one filled with solvent, the other with solution, were placed in the star-shaped frame, the lower portion of which was emersed in the cell filled with the solvent, fig. 1. A setting was now made with the solution, say, in the magnetic field; this tube was then thrown out and the tube containing solvent thrown in, and a new setting made. The next settings were from solvent to solution. The time required to exchange the tubes was a fraction of a second. The measurements are given in Table I. The numbers oppo- site a wave-length are the differences in thousandths of a degree in the rotations for centimeter lengths of solvent and solution. The minus sign before a number indicates that the rotation of the solution was less than that of the solvent. The concentra- tions are parts by weight. Magnetic Rotary Dispersion of Solutions 7 * Saas era ANILIN . CYANIN FUCHSIN FUCHSIN LITMUS oan ; : Alcoholig Water Water ; ae Alcoholic solu- | Alcoholic solu- | solution wheats solution | solution ; length : length tion concentra-| tion concentra-} concen- concen- | concen- ; tion 0.L00019 | tion 0.000025 | tration tration | tration : 0.000012 0.0013 | 0.00008 , A Thousandths we —PF| of a degree are 746 710 o 2 EE 2 ONES SR Soe a noe cmos een Ue eer 5 —2.5 SH 0 5 | 720 —2.5 —1.5| —10 674 10 | —5 GEES ee ae 8 a : 2.5 aes 5 =, 5 = =F 0 0 668 0 0 646 20 es a ee a le nt Pe eee 0 0) 0 30 —5 ‘ —5 —l 0 30 fs 650 10 622 15 0 20 Ea ed ks eee a ee —12.5 12S 3s —10 0 0 5 —10 ey 602 —15 10 627 10 10 5 (SASS Seeaees Sla t Sgaieee) aee Rae Seema —15 5 d Ads —10 a 15 593 0 606 2.5 5 10 25 25 —7.5 —15 10 —2.5 12.5 —10 —17.5 25 5 589 |}—12.5 —5 25 12:5 585 25 | —10 0 5 10 10 ee 0 0 0 — hs it — 5 15 25 —25 —1.5 25 —h 570 0) 15 573 —) —5 0 7.5 10 15 13 10 —5 558 5 25 —I0 10 5 12.5 —22.5 5 555 pss 5 =o 120 —5 Fred J. Bates | ANILIN CVYANIN FUCHSIN FUCHSIN LITMUS eae Alcoholic ‘Water Water ey Alcoholic solu-| Alcoholic solu-| solution ewe solution | solution gth| * ; ilengt tion concentra-| tion concentra- | concen- concen- | concen- tion 0.100019 | tion 0.000025 | tration tration } tration 0.000012 0.0013 | 0.00008 Thousandths ra A=PH) ofa degree ie —10 —15 1a) 543 15 —15 5 Peay tl 0) 5 0) I S75 —2) iS SN 10 =) 528 0 531 —2.5 15 —2.5 20 10 0 —2.5 0) 0 —— ined 20 20 15 520 125 ae D5 517 10 5 —15 5) Tes 509 —7.5. 0) ales 510 nels io dep 10 5 10 A) 0 5 (5) 498 (0) 5 —2.5 £06 —5 10 15 0 0 PS. —15 | 10 ES 25 5 489 10 3) 496 10) —20 15) —15 5 10 494 10 480 —5) B2O ne) 0 (0) —10 == 40) LbaeS 0 —2.5 20 20 5 5 —5 5 472 17.5 215. 485 10 2:5 —10 15_ 5 —35 a5 20 464 —20 0) —40 35 30 272 % ; 7 ra i> “Oe . Magnetic Rotary Dispersion of Solutions 9 In addition to the above wave-lengths the solutions were ex- amined for many other points in the spectrum. The method be- ing a differential one and not dependent on successive readings of the absolute rotation of solvent and solution, the phenomena sought for, if present, could be observed directly. Hence a solu- tion could be examined in a very short time. The differences between the solvent and solution, for wave-lengths not greatly absorbed by the latter, could have been made much smaller, had it been necessary, by increasing the sensibility of the polarizing system. The differences noted are irregular and apparently are merely experimental errors. With a polariscope giving consecutive set- tings within 0.005°, considerable difficulty would naturally be ex- pected in eliminating the various sources of error sufficiently to make this sensibility available. Even after the elimination of double refraction and mechanical errors there are the effects of bleaching and perfect compensation for the absorbing solution to be met. Unless the liquid in both the solution tube and the compensating cell is perfectly fresh a difference in rotation of perhaps 0.01° will be observed. At the time, for instance, the data for the anilin blue solution at wave-length 517 pu were taken there was actually present a difference of 0.01° between solution and solvent. Upon looking for it at some subsequent time it would perhaps have vanished. The above data seem to show that there is not sufficient anomalous magnetic rotary dis- persion in the solutions studied to be measured with the means_ available. Mr. Williams, fellow in physics, in a forthcoming paper has studied the dispersion of alcoholic fuchsin solutions by means of channeled spectra. He finds no anomaly in the dis- persion curve of a fuchsin solution thirty-six times the concen- tration 0.000024. Hence, if we accept the various formulae, for example Becquerel’s where 27e . an OW ak @ which have been presented as criteria for the magnitude of the 273 10 Fred J. Bates anomalous rotation, it is evident that we can not expect to ob- serve an anomaly in dilute solutions. The writer wishes to acknowledge his indebtedness to Pro- fessor D. B. Brace for assistance which has made this investiga- tion possible. 274 4) uoissiusura SE RTC SENS OTEReLdEEe ae me | Heian st Z ieeeeeu innit Sasenenscerssesaye/aaatiie; XS ~ a_i Hy - ‘oa a6 OE ot Sa = eu) a! may se ert ‘SS SraeeD | a LS eee eeey see rere nea ra fate te NS ) i: a HE i peal = PH aaiiite aE the hydranth body to assume the final form and size while still within the 298 tr a Regeneration in Hydromedusae 25 perisarc. The proximal tentacles develop first, the development beginning at the distal end. 6. In Tubularia tenella the first part of the regeneration is similar to that in T. crocea. The hydranth body is separated from the stem by a constriction before the tentacles begin to develop. Tentacles appear as rather short buds, the final form and size being assumed by new growth after the hydranth emerges from the perisarc. 7. In Tubularia larnyx there is a tendency for hydranths to regenerate at both ends of the stem, and there is no difference in the time of their development. The first stage of regeneration is the emergence of the coenosarc from the old perisarc, and the secretion of a new perisarc around the protruding part. A few annulations are always present close to the old perisarc. Circulation of fluid and deposition of pigment are similar to that in T. crocea. Tentacle anlagen are much longer than in T. crocea, and often twined spirally around the coenosare. Ten- tacles assume their final form and size while still within the perisarc. Gonads may begin to form before the hydranth emerges from the perisarc. | 8. Eudendrium ramosum, E. dispar, Pennaria tiarella, and Tubularia crocea were successfully grafted, union taking place equally weil, whether the distal ends or the proximal ends were joined, or whether distal end was grafted to the proximal end. No union of Eudendrium ramosum and E, dispar was secured, nor of Exudendrium and Pennaria. 9. In Tubularia crocea the distal tentacles form by the sepa- ration from the entoderm layer of a rod or column of entoderm cells which are surrounded by the ectoderm to form the tentacle. Proximal tentacles form by a complex folding, involving both ectoderm and entoderm, the ectoderm gradually surrounding the entodermal fold and separating the tentacle from the hydranth body. 10. Distal tentacles of Tubularia tenella form as in T. cocea. Proximal tentacles develop partly by evagination of the ento- derm and later by a new growth after the hydranth emerges. from the perisarc. , 299 26 George Thomas Hargitt 11. The formation of the distal tentacles of Tubularia larnyx is about the same as in T. crocea and T. tenella, Proximal ten- tacles form by a combination of evagination and folding, along the whole length, the tentacles being of mature size and form when the hydranth emerges from the perisarc. Nematocysts are found quite commonly in the entoderm and in the debris of the enteric cavity of T. tenella and T. larynx. 12. Regeneration of Eudendrium ramosui is first indicated by a knob-like protuberance from which the tentacles bud. The ten- tacles start their development as evaginations, involving both ectoderm and entoderm. The entodermal cell at the apex of the. evagination divides, and all cells resulting from this division may continue to divide till the entodermal core of the tentacle is formed, the cells being arranged in a single row. Proliferation of cells in the developing hydranth is by mitosis, amitosis also occurring. Increase in the surface of the ectoderm is brought about partly by a change in the form of the cells and partly by cell division. The hypostome develops as an outgrowth of the developing hydranth, cell division also taking place. The proxi- mal end of the hypostome is blocked by a mass of entoderm cells till a comparatively late period. The mouth opening forms when the hydranth is otherwise completely regenerated. 13. The early appearance of regenerating hydranths of Pen- naria is similar to that of Eudendrium. Nematocysts are very abundant in the ectoderm of the regenerating hydranths. The layers of ectoderm and entoderm are much thicl-ened and the cells are more abundant, being the result of cell proliferation. Mitosis is quite abundant; and amitosis is also found. 14. Amitosis is quite abundant in the tissues of the regener- ating hydranths studied. When not followed by cytoplasmic division (as seems to be the case sometimes) it may be for in- creasing the nuclear surface as an aid in metabolism. Amitosis may be the result of special conditions. The explanation of amitosis in regenerating hydroids can not be definitely deter- mined on the present evidence. 300 . Regeneration in Hydromedusae 27 BIBLIOGRAPHY Agassiz, L, 1862. Contributions to the Natural History of the United States, vol. 1V. Boston. . ALLMAN, G. J. : 1871. Monograph of the Gymnoblastic Hydroids. London, Bicxrorp, E. E. 1894. Notes on Regeneration and Heteromorphosis of Tubu- larian Hydroids. Jour. Morph., vol. IX. Driescu, Hans. 1896. Notes on Regeneration and Heteromorphosis of Tubu- larian Hydroids-Bickford. Arch. f. Ent.-mech., Bd. II. 1897. Studien iibér das Regulationsvermdgen der Organ- ismen. -1—Von der regulativen Wachthums- und Differensirungs-fihigkeiten der Tubularia. Arch. f. Ent.-mech., Bd. V. 1899. Studien iiber, etc. I1—Quantatative Regulationen ber der Reperation der Tubularia. Arch. f. Ent.-meck., Bd. IX. 1901. Studien iiber, etc. LI—Ergidnzende Beobachtungen an Tubularia. Archiv. f. Ent.-mech., Bd. XI. 1902. Studien iiber, etc. VIIl—Zwei neue Regulationen bet Tubularia. Arch. f. Ent.-mech., Bd. XIV. Hareitt, C. W. 1897. Recent Experiments on Regeneration. Zool. -Bull., vol. I. 1899. Experimental Studies on Hydromedusae. Biol. Bull., vol. 1c t9ot. Variation Among Hydromedusae. Biol. Bull., vol. I. Logs, J. 1892. Untersuchungen sur physiologischen Morphologie der Thiere. UW —Organbildung und Wachsthum. Wurz- burg. 1893. On Some Facts and Principles of Physiological Morphology. Biol. Lect., Woods Hole. 301 28 George Thomas Hargitt Morcan, T. H. . 1901. Regeneration in Tubularia. Arch. f. Ent.-ntech., Bd. XI. 1901a. Factors that Determine Regeneration in Antennularia. Biol. Bull., vol. IT. 1901b. Regeneration, New York. 1902. Further Experiments on the Regeneration of Tubu- laria. Arch. f. Ent.-mech., Bd. XIII. PEEBLES, F. 1900. Experiments in Regeneration and in Grafting of Hydrogea.. Arch, {, Ent.-mech., Bd. X. 1902. Further Experiments in Regeneration and Grafting of Hydroids. Arch. f. Ent.-mech., Bd. XIV. Ranp H. W. 1899. Regeneration and Regulation in Hydra viridis. Arch, f. Ent.-mech., Bd. VIII. STEVENS, N. M. 1901. Regeneration in Tubularia mesembryanthemum. Arch. f. Ent.-mech., Bd. XIII. 1902. Regeneration in> Tubularia mesembryanthemum, I. Arch. f. Ent.-mech., Bd. XV. 1902a. Regeneration in Antennularia ramosa. Arch, f. Ent.- mech., Bd. XV. WEISMANN, A. 1883. Entstehung der Sexualzellen bei den Hydromedusae. Jena. . , WILSON, (Dees 1896. The Cell in Development and Inheritance. New York. EXPLANATION OF FIGURES. Fig. 1. Eudendrium ramosum grafted by distal ends. Fig. 2a. Pennaria grafted by proximal ends. Fig. 2b. Pennaria grafted by distal ends. st.—stolons. 302 F we Regeneration in Hydromedusae 29 Fig. 3. Tubularia crocea. Distal end showing the tentacle anlagen. | Fig. 4. Tubularia crocea grafted by proximal ends. Fig. 5. Tubularia crocea, distal end in sand, new hydranth having regenerated. The unshaded portion is the old stem. Fig. 6. Pennaria, distal end in sand showing regeneration. Unshaded portion is the old stem. Fig. 7. Pennaria, stolon in sand showing regeneration. Un- shaded portion is old stem. Fig. 8. Pennaria lying flat on the bottom of the dish, with re- generating hydranth. Unshaded portion is old stem. Fig. 9. Pennaria and Eudendrium grafted. st—stolons from Pennaria. Fig. 10. Pennaria showing holdfasts and hydranths. This drawing is made from parts of two different stems. Figs. 11, 12. Regenerated hydranths of Tubularia larynx. Figs. 13a, 13>. Regenerating hydranths of Tubularia larynx, showing the tentacle anlagen. Developing gonads in black. Figs. 14a, 14b, 14c. Early stages in the development of the hydranths of Tubularia tenella. Figs. 14d, 14e, 14f. Hydranths of Tubularia tenella after emer- gence from the perisarc. Fig. 15. Pennaria anchored to the bottom of the dish by the distal end, showing regeneration. Unshaded portion is old stem. Figs. 16-24. Tubularia crocea. Fig. 16. Early stages in the folding of the ectoderm and ento- derm to form the proximal tentacle. 600. Figs. 17, 18. Later stages of folding. 600. Figs. 19, 20. Ectoderm folding around the entoderm to cut off the proximal tentacle. 500. Fig. 21. Promixal tentacles separated from the hydranth body. 500. Figs. 22, 23. Two stages in the development of the distal ten- tacles. Column of entoderm cells being formed. Ectoderm folds around this column to cut off the tentacle. ><600. Fig. 24. Separation of a mass of entoderm cells in a column. 303 30 George Thomas Hargitt Ectoderm gradually surrounds the column to form the tentacle. Amitotically dividing nuclei shown. 600. Figs. 25-31. Tubularia tenella. Figs. 25, 26. Early stages in the evagination of the entoderm to form the proximal tentacle. 6o0. Fig. 27. Evagination of the entoderm complete and the ecto- derm folding around the entoderm to cut off the tentacle. 500. Fig. 28. One proximal tentacle cut-off and another with the entoderm entirely surrounded by ectoderm but not yet separated from the hydranth body. 9500. Fig. 29. Longitudinal section through the developing proximal | tentacle, showing the evagination of the entoderm and the for- mation of the tentacle. 600. Fig. 30. Entoderm cells being forced away from the enteric cavity to form the entodermal column of the distal tentacle. < 600. Fig. 31. Entodermal column of distal tentacle separated from the entoderm layer and the ectoderm beginning to surround it. 600. Figs. 32-43. Tubularia larynx. Fig. 32. Proximal tentacles separated from the hydranth body, others in the process of separation. In one tentacle the ento- derm is not completely surrounded by the ectoderm. 600. Fig. 33. Earlier stages in the development of the proximal ten- tacles. EEntodermal column nearly complete, and the ectoderm just beginning to surround it. ><600. Figs. 34, 35. Very early stages in the development of the proxi- mal tentacles, only a few cells pushed away from the enteric cavity. ><600. Figs. 36-38. Early stages in the formation of the distal ten- tacles, the entodermal column shown in different stages of de- velopment. 600. Figs. 39, 40. Entodermal column being surrounded by ecto- derm to complete the tentacle. 600. Fig. 41. Entodermal column, entirely outside of the entoderm layer, being surrounded by ectoderm. 600. Fig. 42. Completion of the process shown in Fig. 41. ><600. 304 Regeneration in Hydromedusae 31 Fig. 43. Single entoderm cells lying almost entirely within the ectoderm. A condition sometimes found in the early stages of development of the distal tentacles. Figs. 41, 42 are really the completion of this process. 600. Figs. 44-60. Eudendrium -ramosum. 44-57 600; 58-60 X 290. Fig. 44. Very early stage. Layers of ectoderm and entoderm much thickened. Figs. 45-47, 49. Different stages in the early evagination of the entoderm. Figs. 48-50. Transverse sections through completed or par- tially completed tentacles. Fig. 51. Transverse sections through tentacles in different stages of development. The two more or less distinct series of anlagen, represent the development of the two series of mature tentacles whose bases are alternately elevated and depressed. The mass of polygonal cells are entoderm cells massed together in the region of the hypostome. Figs. 52-56. Progress of development of two tentacles, from the condition in which only a single cell has been pushed away from the enteric cavity (fig. 56), to the complete separation of the tentacles (fig. 52). Figs. 57, 58, 60. Longitudinal sections through the tentacles in various stages of development. Figs. 58-60 show different stages in the development of the hypostome. In Fig. 60 a considerable amount of debris is present in the enteric cavity. In figures of Eudendrium ramosum there often appear to be two cells in the sections of the completed tentacle. This is due to the fact that the section is not exactly transverse but has been made somewhat obliquely. Where three nuclei seem to be present this is due to the appearance of the nuelei of the underlying cells. Figs. 61-67. Pennaria tiarella, Figs. 61,62. Nuclei in both ectoderm and entoderm in the early stages of mitosis. > 1000. Fig. 63. Nearly completed amitotic division of an ectodermal cell. >< 1000. 305 32 : George Thomas Hargitt Fig. 64. Nucleus of ectoderm cell in an early stage of amitotic divisien. The cytoplasm has not yet begun to divide. 600. Fig. 65. Transverse section thisugh the hydranth region of a young budding hydranth, which was forming normally in its natural habitat. >< 1000. Figs. 66,67. Sections through the developing tentacles. ><600. 306 PLATE | GTN. del PLATE I! G&.TH. det. u Bees de 2 San e he, ; = 1 ; ” t ai “5 ant PLATE III PLATE IV PLATE V a x A i any me as PLATE VI @) cx 7] X C8 a) f NS ey iC, : as : > aS, G ae sweet PLATE VII PLATE VIII oe nese e PLATE IX II.—Some Peculiar Double Salts of Lead BY JOHN WHITE Although it has long been known that the sulfate of lead is markedly soluble in aqueous solutions of alkaline salts of certain organic acids, such as the acetates and tartrates of ammonium and sodium, it does not appear to have been generally remarked that this property of solubility applies also to other of the diffi- cultly soluble lead salts. In the course of some other work dealing with compounds of lead, it was observed that lead iodid dissolves to an appre- ciable extent in a concentrated aqueous solution of sodium acetate, yielding upon evaporation a white crystalline crust, which was found to contain both lead and iodin; subsequent experiments have shown that the chlorid and bromid of lead behave like the iodid under similar conditions. Upon search, it was found that the literature of the sub- ject is very meager, this property having been previously _ observed by only a few investigators, and that none of the text-books of chemistry make mention of it. The first notice of it is contained in an article by Poggiale,* in which he describes a complex salt as being formed by heating together lead chlorid and basic lead acetate; to this he gives the formula PbCls. PLO C2ff3O2-- 1 5F20. The published data concerning this salt is such as to lead to the conclusion that it was very impure. Later, Carius’ pre- pared a somewhat similar compound by dissolving lead chlorid in’ a water solution of lead acetate, to which he ascribes the formula 14nn. Chem. (Liebig), LVI, 234 (1845); Compt. rend. XX, 1180. 2Ann. Chem. (Liebig), CX XV, 87 (1863). 307 a John White Cl (2F13O2 xe GO ue | OHO, 3/20: Judging from the results obtained during the present investiga- tion, there is every indication that Carius’s results and formula are substantially correct. About ten years later, Tommasi,’ ap- parently in ignorance of the work of Carius, succeeded in dissolv- - ing the iodid of lead in potassium acetate solution, and from this obtained a crystalline compound to which he gave the formula Be 2Pb | CoO, + KR OHs0%. It will be seen later in this article that, although Tommasi did not ascribe the correct formula to his compound, he in all probability had a substance of definite composition. In the same article he calls attention to the fact that the iodid is soluble in the acetates of other bases, but he was unable to isolate any of these and so to prove that definite compounds had ‘been formed. Observations similar in character to the above, but which led to nothing of moment, were also made by Nickles? and Field.® The paucity of the literature upon the subject and the uncer- tainty of the results obtained suggested the advisability of under- taking a thorough investigation with the object of determining the general character of this and allied reactions and the nature of the compounds, if any, which are formed. Some of the results thus far obtained are recorded in the following pages. The iodid was selected as the best of the halogen salts of lead to experiment upon in the beginning, for, although the resultant compound might prove less stable, any change of color could be more easily detected than with either the bromid or chlorid. This has proved to be the case, for in practically - every case the salt obtained was white; any decomposition would give rise to lead iodid which could be easily recognized by its 1Ann. chim. phys. [4], XXV, 168 (1872); Bull. soc. chim., XVI, 337. 2Compt. rend., LVI, 388 (1863). 87, Chem. Soc. (London), XXVI, 575 (1873). 308 "eats * Some Peculiar Double Salts of Lead e ‘color. The present paper will describe the results cbtained with the iodid. THEORETICAL Cariust describes the preparation of a class of compounds of bs : the general sors Pb< COoHsO» where x may be either chlorin, bromin, or iodin. To obtain these, he heated together, in a sealed tube, lead acetate and some alkyl-halid, e. g., ethyl, methyl, or methylene chlorid. On heating to a proper temperature, reac- tion takes place, yielding an alkyl-acetate and what he calls chlor- acetin of lead in a crystallized state. Schorlemmer? prepared in like manner lead hexylacetochlorid, thus substantiating Ca- rius’s work; the same has been done in this investigation by repeating the experiments of Carius. In attempting an explanation of this reaction, it is possible to make use of the results of von Ende’s* observations upon the ionic dissociation of the halogen salts of lead. In this he has shown that these salts dissociate in two stages, e. g., PbCl2,= PbC-+-C’ and then POC’ =Pb"-+-C'; the first stage of toniza- tion takes place much more readily than the second. It seems reasonable to suppose that this same argument applies to the other salts of lead, in which case the formation of compounds like PILE, HO», be readily understood, for: PI AO POO On" + GH! . This first dissociation would take place with comparative ease, the second less readily, being hindered furthermore by the presence of acetic acid—glacial acetic acid being always used as a solvent; the positive ion PbQ@H;O2, could then com- bine with the more negative Cl’ ion, forming the compound GeO; The reaction is probably reversible, that is, es 24nn. Chem. (Liebig), CXCIX, 142 (1879). 3Ztschr. Anorg. Chem., XXVI, 129 (1901). 399 4 John White lead chlorid if heated with ethy! acetate would yield the same compound, were it not for the fact that the chlorid is soluble in ethyl acetate to so slight an extent as to make it impractical. In the endeavor to overcome this objection, a solution of sodium: acetate was used instead of the ethyl acctate, with the result that a reaction took place, presumably yielding the above compound, but only as an intermediate product, for the excess of sodium acetate required to bring about a reaction was ‘so large that a further action took place, presumably caused by the halid-acetate first formed uniting with the metallic acetate, yielding a double compound of the general type Pb making use, also, in the period up to July 15, of the Courrier de Provence seven times,® the Journal de Versailles 1The account of the session of May 25 fills nearly two pages in the Péci¢; Bailly gives the substance of it in less than half a page. The Pvrocés-verbal devotes twenty-three pages to the account of the first sessions of the commis- sioners selected to consider the matter of how the credentials should be verified ; Bailly disposes of it in three pages. Bailly’s account of the ses- sion of May 26 is a skilful condensation and combination of the Pyocés- verbal of the conference of the 25th of May and the Réc7¢ of the 26th. The account of the session of the 27th in the J/émoires is pieced together from extracts taken literally from the Aécit. 2 Mémoires de Bailly, 1, pp. 79,110. 8 Jbid., I, pp. 84, 135. 4 Jbid., I, pp. 43, 87, 98. 5 The account of the session of the 12th of June is taken from the Aéci¢ for the first part of the day and from the Pvocés-verbal—that began on this day—for the last part. A comparison of the /Pvocés-verbal for June 22 (I, No. 1), with the A/émoztres (I, pp, 199-202) will furnish a good example of Bailly’s dependence upon his source. 8 Wémoires de Bailly, 1, pp. 147, 212, 278, 284, 807, 315, 333, 348 The Mémoires de Bailly 19 four times,’ the Journal de Paris once,? the Point du jour seven times.* His own additions are numerous and often important. Mentioning only those of at least a paragraph in length, they are: an account of the critical session of the evening of June 16 (I, pp. 150-156) ; a correction of the Procés-verbal (1, p. 167) ; observation on the use of the word décret by the assembly (1, p. 171) ; observation on the expression classes privilégiées employed in the king’s letter of June 17 (I, p. 175); remark on the’ vote par téte made June 18 (I, p. 175) ; incident of Madame de Tessé (I, p. 175); disorder in the assembly (I, p. 176) ; Bailly present in the street when the clergy voted to join the commons (I, pp. 178, 179); members of the assembly threatened with violence because of their opinions (I, pp. 179, 180) ; the events of June 20, in which Bailly was so prominent a figure (I, pp. 180-194, the personal recollections of Bailly being combined with the ac- count of the Procés-verbal) ; the incident of the return of Martin d’Auch to the assembly (I, pp. 192, 193, 194); letter from the king, June 21 (I, pp. 196, 197) ; the union of the clergy with the commons, June 22 (I, pp. 198, 203) ; arrangements for the Royal Session, call upon the guard of the seals, and midnight interview of Bailly with members of the liberal nobility (I, pp. 204-206) ; account of the Royal Session, in which Bailly supplements from memory the accounts of the Procés-verbal and of the Courrier de Provence (1, pp. 206-223) ; admittance of the public to the hall of the commons (1, pp. 225, 226); attempt of the people to force an entrance into the hall (I, p. 233); attempt on the part of Bailly to suppress applause in the assembly (I, p. 247) ; valu- able personal recollections on the conduct of the majority of the nobility and the minority of the clergy (1, pp. 247-264); the affair of the French Guards (I, pp. 266-268) ; personal notes on Villedeuil and Breteuil (I, pp. 307-310); recollections of the 14th of July, running through the accounts taken from the sources (I, pp. 359-395) ; the king in the assembly (I, pp. 7-11). With the 15th of July, the last period of the Mémoires opens. 1Jbid., 1, pp 246, 304, 312, 317. 2 Tbid., I, p. 343. 8 Jbid., I, pp. 147, 192, 247, 253, 263, 275, 278. 349 20 Fred Morrow Fling It deals with the incidents of the mayoralty of Bailly up to Octo- ber 1, 1789. The chief sources are the Procés-verbaux of (1) the electors of Parist (to July 25), of (2) the first assembly of the representatives? (July 25-September 18), of (3) the second as- sembly of the representatives of the commune (September 18- October 1). The most of the material is drawn from these rec- ords, all of which were accessible in printed form when Bailly wrote. He constructs his journal by selecting from these records what appears to him important enough to be emphasized and especially the incidents with which he was connected. When his journal contains little under a given date, it is not due to the failure of his memory, but to the lack of interesting material in the sources that he is consulting.* For the most part he selects and condenses, the condensation being much greater in this sec- ond volume than in the first; at times he follows the text closely, putting the narrative into the first person, but without quota- tion marks.* In the second volume, he refers only incidentally to what is taking place in the assembly at Versailles, taking his material from the Procés-verbal or the newspapers.® For the events in Paris, he refers frequently to the papers, quoting the Gazette de Versailles,° the Patriote francais,’ the Révolutions de 1The Procés-verbal des électeurs was largely composed aprés coup. An account of its composition is given in detail in the third volume (pp. 1-53). The substance of this account is found in Flammermont, La journée du rg juillet 1789, pp. X-XV. Bailly was secretary from April 26 to May 21, Du- veyrier from May 22 to July 30, 1789. 2 The exact titles of these Procés-verbaux of these two assemblies have been given above. Bailly claimed (II, p. 147) that the Procés-verbal of the first assembly was not as reliable as that of the electors, ‘‘comme ces fvo- cés-verbaux n’ont été redigés que longtemps aprés, les rédicteurs ont mis ce quwils ont voulu.’ 3“ Tes jours présents ne me fournissent rien pour mon compte; je n’ai pas grand’chose non plus 4 dire de l’assemblée des représentants, ses procés- verbaux montrent le vide de ses séances’’ (II, p. 220). *Compare especially pages 476 and 477 of volume one of the Proces-ver- bal des électeurs with pages thirty-two and thirty-three of the second volume of the J/émoires. >These passages are not numerous, Saadl only with such important ques- tions as the adoption of parts of the constitution, the discussion of the Au- gust decrees, or the declaration of rights, and are treated very briefly. 6 Mémoires de Bailly, Il, p. 197. 7 Ibid., Il, pp. 201, 229, 243, 260, 277. 350 The Mémoires de Bailly 21 Paris,’ the Chronique,* the Point de jour,* the Journal de Ver- sailles,* the Courrier de Provence,’ and the Journal de Paris.® He does not, however, recognize his full indebtedness to the papers, using their material and often their language without any acknowledgment.‘ He undoubtedly had before him when he wrote his correspondence with Necker during August and September, 1789,5 and possibly some of the papers or records of the comité des subsistances of Paris.® His personal recollec- tions do not form as large a part of the second volume as they did of the first; they relate chiefly to the work of the committee engaged in procuring food for Paris and to the friction between Bailly and the assemblies of the commune. His running com- ment upon the acts of the assemblies, and his interpretations of events are valuable, and the historian can not afford to overlook them. They are so interwoven with the narrative that it does not seem advisable to indicate them in detail. In the first two volumes of the Actes de la commune de Paris are found the 1Jbid., II, pp. 80, 81, 219, 244, 319, 325, 368, 379, 385 392, 404. 2 Tbid., 11, pp. 304, 315, 325, 330, 368, 372, 379, 392, 400. 3 Jbid., Il, pp. 36, 256. 4Jbid., Il, pp. 69, 169, 253. 5 Tbid., II, p. 185. 8 Jbid., II, pp. 320, 332. 7In the second volume, for example, Bailly makes use, without citing his source, of the Courrier de Provence, the material on pages 37 and 38 be- ing taken from volume one, page 445 of the Courrier; of the Point du jour, the passage, ‘‘ M. l’abbé de Montesquiou a reconnu que les membres de la minorité du clergé s’étaient trompés, et qu’ils en faisaient l’aveu a la nation avec plaisir,’’ (I, p. 218 of the paper) is literally reproduced on page 39 of volume two of the A/émoires ; of the Point du jour (1, p. 232), on page 41; of the Révolutions de Paris (1, No. 2, p. 12), on page 77; of the same pa- per (I, No. 5, p. 36) on page 265; of the Point du jour (III, p. 46) on page 375. These are a few examples of the necessity of collating every passage in Bailly with the sources from which it might be drawn before crediting it to Bailly as independent evidence. 8 Bailly refers to this correspondence on pages 235, 288, 356, and 371 of volume two. The letters have been published in volume four, pages 172 to 195 of the Histoire parlementaire de la révolution francaise by Buchez et Roux. A comparison of the contents of the letters with the text in the Mémoires will make clear that Bailly had the letters before him. %There is only a possibility that Bailly had before him when he w ote the records of the comité des subsistunces. He said of these records (II, p. 358), ‘‘ Des registres, ily en avait peu, et ils n’étaient point parfait- ement en ordre.”’ 35! 22 Fred Morrow Fling proces-verbaux used by Bailly, and in his excellent éclaircisse- ments, the editor, M. Sigismond Lacroix, has indicated most of the passages in Bailly that add anything to the records. What is the conclusion of the whole matter? Or, in other words, what is the value to the historian of the Mémoires de Bailly? It is a record made by a competent eye-witness, but by a witness who followed carefully the best sources in construct- ing his journal, supplementing these sources by his personal rec- ollections. It is a compilation and yet not a compilation. Bailly could not recall the events and the days when they occurred without the use of the procés-verbaux and the newspapers, but when he had recalled them he seldom trusted to his memory — for the order of the facts and often employed in his journal the very language found in the sources that he had consulted. By so doing, he gave his approval to the account, practically saying so it occurred and not otherwise. This approval certainly has some value, just as his corrections of the record have a value, but it is not the value that we attribute to the account of an independent witness. The Mémoires may be safely used by those to whom the sources from which Bailly drew are not accessible; the historian will use the work only when it corrects or supp!e- ments the sources upon which it is based or when he wishes to show the point of view of Bailly himself in 1792. Before any passage of the Mémoires is attributed to Bailly it must be col- lated with all the procés-verbaux and newspapers that may have served him as sources of information; if found in none of these, it may safely be treated as an independent bit of information. Failure to do this in the past has led historians to quote Bailly when he is not a source’ and to charge him with drawing from Louis Blanc makes use of Bailly’s A/émotres in volume three of his FTistoire de la 1 évolution francaise, but it seems to be a mere accident when he cites matter found in Bailly alone. An examination of the passages of the J/émoires cited by Blanc in his footnotes will show that many of the pas- sages had been taken by Bailly from the pvocés-verbaux or from the news- papers. In these cases, naturally, the J7émoires should not have been cited. Louis Blanc’s critical work was not of the highest order. 352 The Mémoires de Bailly 23 works that did not exist at the time when he wrote.!' There are mémoires and mémoires. The most of those upon the French revolution were composed as Bailly composed his journal, al- though, as a rule, the writers did not show the same good judg- ment in selecting and in using their sources; some drew largely upon the memory—Grace Dalrymple Elliott, for example, in her Journal of My Life during the French Revolution?—and are extremely unreliable, at times, absolutely worthless. Just what the value of each mémoire is can be determined only after a careful critical study. 1The editors of the edition of the A/émoires published at Paris in 1822, state in a footnote to page 311, that the details of a certain incident related by Bailly on pages 310, 311 were drawn from the JJoniteur of September ~ 15, 1789. Asthe Moniteur of that date did not exist until after the death of Bailly, it was not possible for him to make use of it. The history of the origin of the numbers of the J/onifeur that precede November 24, 1789, isa sealed book to many of those who make use of them. * Journal of My Life during the French Revolution, by Grace Dalrymple Elliot, London, 1859. It would be somewhat difficult to find a work writ- ten by an eye witness that contains as many errors to the page as are con- tained in the first chapter of this work. 353 IV. —On the Representation of Numbers as Quotients of Sums and Differences of Perfect Squares BY ROBERT E. MORITZ Let the simple continuant! whose elements a@,, @p41,. . . @—1, @%; r>¢, that is, when it involves elements which do not exist in the continuant under consideration, and thus no attempt seems to have been made to extend formula [1] to values of s¢. Yet it frequently occurs that general expres- sions based upon [1] are to be specialized in a way which necessitates s to assume values outside the limits ~ and ¢, as when we expand a continuant with reference to the Ath element, and then desire to let k=1 or 2. Thus constituents like 19 or 2-1 may occur, which vitiate the results and require that values of s which give rise to such expressions be treated as separate cases. This is quite unnecessary. Since fg,5 is meaningless as a con- tinuant for a<7 or 4>42, it may be given any meaning we please. We shall therefore define it in a purely formal way by the funda- mental equation [7]. 1For definition and fundamental theorems see Chrystal, 4/gedra, Part II, p- 466. 355 2 Robert E. Morite Put in [1] for s, ~—1z and we have Bri = brea pritprr—2 prt: This equation is obviously satisfied for f,,-1==1, pr,7—-2==0. For s=r—z2, we have Prt = prr-2pr-1tthrr-s prt hence #;-3 must equal 1. For s=r—3, Prt= bros pr-attbrr—s prt which is satisfied for p,,-4 = 0, Droz = prt. Proceeding similarly we find successively Prr-1 = 1, Prva = 0, . . Dr v-oetn = 1, pr roe =O, Dr-2,t = pr—-4t = prot =... + Prt. ' Similarly by putting s==/, ¢+1, etc., we obtain Ptit = 1, Pree = 0,.- ~ , Pep+oe—1t = 1, Deport = O, Prt = pris = prt) + = Prt so that if the continuant under consideration is f,;, y<¢4, we have Prr-2hH = 1, prr—2e = 0, pr—2kt = pre 1 ce & positive. [2] Pivera—-1,t el Purest a= ©, Prt42k = prt ) By definition p,*=/( a, )=a,, hence for s=r [1] becomes Pri = & prorttprt2,ts [3] and similarly for s=/—1 we have Pri = "% prerthrt-2 ita The relation between continuants and continued fractions, which has been deduced in various ways,* is most readily established by repeated application of [3] and [4]. For [3] gives us successively *We could with equal reason write ~;,7 instead of J,, but I wish to reserve the symbol #y7,r for a later use. 1S. Giinther, NMidherungswerthe v. Kettenbritchen, S. 31, Habilitations- schrift, Erlangen 1872. 356 Quotients of Sums and Differences of Perfect Squares 3 : 1 prt | Prasat = Or+ Peas foea! posta | Prine = argit 5a r+1,¢ r+2,¢ r+1 Privat ] Prist Bi-1t | pi = 4-1 ak an Sel ee aia SERS ST ES Now f; = a@ and fr444 = 1 by [2], consequently ig 1 ae t Pads 1 , a rele ae (s] ; 1 a} 1 1 Vet career at Dirichlet! has introduced the symbol (a,,a,+1,... , a) to represent the fraction [5]; we may therefore write pri SS (Ay, ss, Rem 6 Pca eee a), 7S [6] Similarly we obtain from [4] Prt SA st penietiaty Ori} ret. [7] pr,t-1 Finally, since by the definition of a continuant Prt ane Pur [8] we have also 1Dirichlet, Werke, Bd. 2, S. 141. 357 A Robert E. Moritz re (Qnfrti.++14), 1 Pin-1, OF Pin-1—fon#9, and hence, A pure recurring continued fraction can not be equal to a pure quadratic surd number. : If ag = 0, [3] gives us Pon = p22, Pon-1 = f2,n-1; and [13] goes over into — Pin-1—f2,n 4 ny her 2m) +4pinpon-1_ . [14] pin 2P i,m Now by [3] and [4] Pin=1—fan = % £2n-1+p3,n-1—(An Pn-12+hn-22) hence in order that the rational constituent in [14] may vanish we _ must have 5 eg ay p2n—-1+f3,n-1—4n pn-1,2—fn-2,2=9, or 2 ig I ie 1 Pan 2—P3,.n—-1 | [15] | P2n-1 Furthermore 2 n-1/f2n-2=(an-1,- - - , 42) and fog 1/f3,n1—= (az. . .,4n-1) hence fon-1 is greater than either fo,,-2 or f3.n~-1 OF their difference J2n-2— P3.n—-1, ( fon—2—f3,n—1)/f2,n-1 IS a proper fraction, and [15] is satisfied only when simultaneously We now treat p2n-2— p3,n-1 eee as we rive Pin 1—f2,n and find that J» »-2— f3,n-1 can vanish only, if az = An, and £3 n~-3— Pin—2=0, and this in-turn leads to the general condition a@g41—=a@p,_,. t 359 6 Robert E. Moritz Moreover, if these conditions prevail, the rational part of [14] vanishes, for in that case, A1,n—1==fn-11==f2,n. We have therefore: The necessary and sufficient conditions, that [13] may represent a pure quadratic surd, are ao—0O, 4—Aa4n, a =An—-1, a3—=an_2, owe, (si »Apijt—An_p. With these conditions, the recurring continued fraction becomes X=(0,4,a2,...,42,a,x). The last few terms of this fraction are 1 ui 1 Wea rete. eva ah a Sagi a ce Be Cane aC | ORS Rey AE NET TaN TCT a,-|--——_— a,+——— __. 2a, + —_—_—— Spm ate ay+ : tT as + With this equation [14] may be written | P2,2 4==(0,%,82, 0°. . 22,20), | ( ) 12, ) och 1) pis 2 [16] or its reciprocal Aid = (41,02, «4, 20)) = ae [17] * * =e where the stars denote the beginning and end of the recurring elements, and for the sake of convenience £1 and #22 have been written for f(a, ,a2,...,@2,a,) and p (ag,a3, . . . ,a3,@2 ) respectively. We now need only to take the result just obtained, namely, that every recurring continued fraction which represents a pure quadratic surd takes the form [17], in conjunction with the theorem that every pure quadratic surd can be expressed as a recurring continued fraction, to arrive at the equations. Dae pa NP NTs p22 beg 218 "yaa Aus ’ o/h [18] for every non-quadratic rational number N=L/M, that is: or 360 Quotients of Sums and Differences of Perfect Squares 7 Every non-quadratic rational number can be expressed as a quotient of two simple reciprocal* continuants, whose elements are positive integers, and of which one is formed by omitting the initial and final elements of the other. Moreover, since every quadratic surd can be expressed as a continued fraction, with unit numerators and positive integral de. nominators, in one way only, the representation of non-quadratic rational numbers by means of quotients of continuants is unique, there is a one-to-one correspondence between the domain of non- quadratic rational numbers and the totality of simple reciprocal continuants with a single cycle of positive integral elements. To each number N of this domain corresponds one definite con- tinuant which we may call its continuant, and each number of the domain belongs to one or the other of two classes according as its continuant is of even or odd order. We may attach the order of its continuant to the number itself, and speak of num- bers as having odd or even order, which of course has nothing to do with the odd- or evenness of the numbers themselves ; 6, 19, 28, 51 are of an odd order, 10, 13, 58, 97 of even order. If we admit continuants of more than a single cycle of ele- ments, the one-to-one correspondence ceases to exist. For any pure quadratic surd is represented equally well by each of the forms Vr CLS DF sa a, a, inte -»@2, 201), = (a1,a2, 724 ts . 42,241), * * = (a1, a2, PES SAG CEP eRe eS . 42,241), * * Be EE ERs | city DOA Seto ag SABI Ne yo GALT, Oe feu, ie OB, 2AL), * * etc., consequently to the same number L/M corresponds each of the continuants p(ai,a2, . . .,@1,a2), DU Bibs sols ey SAY gs i. ORAL), WGA a aur tg ZON; 3 fey @Alyo a > |b Oa1), \ [ro] p(a1,a2, bh Regi T es ve sok QAI at oe SM ty sw a gAByQ), | etc., J 1A reciprocal continuant is one in which a4 = a@n-4 for every value of &, 361 8 Robert E. Moritz containing respectively one, two, three, four, etc., cycles of ele- ments. We will use the shorter symbols p(ai, . . .,@1),fi(a1, . . .,@1), po(ai, + . .,@1),p3(a1, . . .,a1), etc., to represent the above con- tinuants, and generally write Prlay, - + -sOp) for the continuant ip which the element 2a, recurs 2 times between the initial and final elements a) and ay. We may then write ee PGi. % G1) b> PuQeiy: «3.5 )@1) uM PAI ors a2) pr(aa, ae pag Rit ck eX) i PAOD ak = “po as, 2h og a pene tS Es Cage 42) | [20] Let us consider the first of the continuant quotients [20] for a. number whose order is even; in that case its continuant has two equal middle elements, az, and by means of [1] we obtain PGI ie AL) = Paty Lede DP (Aedl sean ee) ; PCa’ w}- -,Qk-1) p(ae-1, = Dn See =r (ai Co an) +p? (aa, One -4k-1); and likewise Daa; is aa) = f"(az, rik .,4,) +P? (a2, Me ee) hence, I. Every number whose order is even can be expressed as a quotient of sums of squares of positive integers. On the other hand, the continuant of a number whose order is odd has a single middle element, a, and in that case Pia esa) = PMI, Gita SOE EAT ues ep + PCA, : 20. Ag) PASS | eee (CAT, Oe) DAA Lee pam + pla, .-.,@e-1) pn, . 5 @e=8)e =F) [ax pan pS NAEA 1) ete PLANS a ans) | 4 Oe Ce 5 fg Pet yo ee Cae a) | Sas [ Pa, ons ., 2p ) = OE, enn yet) ; 362 Quotients of Sums and Differences of Perfect Squares 9 since by [2] app(ay, .. .,@e-1) + pla, .. -,@e-2) = pla, . - -,@z) and . App(ai, .. .,42-1) — pla, .. -,@2) = Pla, . ..,@z=2). Likewise P(@s, so +)@2) = <{ Pas, > 2 .,@z) —P(a1,. . sana) |. ‘In the quotient of p(ai,.. out, and we have a1) by PC aa .,a2), a divides II. Every number whose order is odd can be expressed as a quotient of differences of squares of positive integers. If the continuant p(ai, . . continuants [19] is of odd order. first, third, fifth and 2a, in the second, fourth, etc. .,a1) is of odd order, each of the other The middle element is ag in the The steps that - led to Theorem II., lead likewise to each of the equations :— eS L>M, order odd, L _ play, eat) _ pa, 2 yA) f(a, . - ,4@p-2) M~ p(aa,. . .,a2) Pas, .. .,a2)—f" (aa, .. 242) ed MURAL, ta 2 2g) _ (a, .. .,2a1)—f?(a1, . . .,a3) ’ pilaa,..'.,@2) p*(aa, . . .,2a1)—f? (a2, . . .,a8) __ polar, . . -»%) _ pr(a, oces Ge) Pt (iy <2 Se2) pr(aa,...,a2), pir(aa,.. .,az)—pi*(ae, . . .,ae—2) [21] __ pen( ay - - M) __ Pri (a, . + 05 @p)—pn' (a, . . .,ap-2) P2n( aa, . . -,@2) pn? (a2, : Ap )—pn? (aa, hk Sep) Prt. -+1 1) __ Pr (Qa, » « +5241) fr? (A, « « - 543) Pan+i(42,..+,42) pr’ (az, se .» 241) —py (aa, Lak 5B) 363 IO Robert E. Moritz This result may be stated thus: III. Every number whose order is odd may be expressed in an unlimited number of different ways as a quotient of differ- ences of squares of positive integers. | If the continuant A(a;,...,@,) is of even order, the succession of | continuants [19] have alternately even and odd order. middle elements of the first, third, fifth, etc., is az, ‘second, fourth, etc., have each 2a, for the middle element. we have successively The while the Hence Se L>WM, order even, y L£ pa, vite at) _ Pa, uo” az) +f7(a1, RR Re tS M plaza. . .,a@2) fp (aa, . . «,@x) +f? (a2, . . .,@p-1) pila, Sere ab) Lp Cay A .» 241) — f(a, 4 Shag EBD) | Beil Gay ae naa) pa ea gest = +3201) —p" (42, eee) __ pra, Wee) ee CA, ous 42) +pr(ai, eee Ope) po ag, otis .,@2) pr (aa, * 4p) + pr a2, «e +» @p-1) | 22) a) See ‘ 5 eae ie:: whe ., a2) pu? (aa, a 4p) thr? (a2, Bil _ penti(a1,. . . a1) Detar, ee 5201) —pr7 (ar, o 43) | ponti(a2,...,@2) pn’ (a2, . . .,201)—pn® (aa, . = .,@8) J This gives in the following theorem: — pron( a, 2n( a2, Pure IV. Every number whose order is even may. be expressed in an unlimited number of different ways as a quotient of sums, as well as of differences, of squares of positive integers. The direct computation of these quotients for a given number — is for large values of the suffix very laborious, but recurrence- formulae may be deduced by means of which, from the con- stituents which enter into the quotients for a given suffix, the quotient for the next higher suffix may readily be computed. Suppose that for some number of even order, 364 | Sea Quotients of Sums and Differences of Perfect Squares 11 br» (a1, ¢ pm 42, ate has been computed, then N= .,@1) “ns Pat (Gr; es bylg@a) __ Pn (a1, Aso ,a2) pn’ (az, o. , ,Ap-1) -»@p-1) 4p) + pr? (ai, A 42) +fpn’ (aa, ee 5201) —py? (ai, ot. + 9@lg) N — Peet, 4 Eat P21 (42, ats - 42) pr? (aa, P56 may be readily calculated, for by [1] Pal 41, \.2.5221) == pal Gs, «2 1,82) P(2ai,,:. @%) PAI, «oo Oh) P 2a, 2 U3REP (2Gs;. «5 201)—fn?(@a, : «4-,@3) 1241) —=py( a3; . .,Ap) Pa (aaj, 3. bea, .. G64) P(2ai, » 3s ae-1) = pn(ar, nee 43) = pn(ay, ate . ap) pas, a Ap) TPpn( a ri -»Ap-1) (as, volte een) | Pn(a@2, . . .,48) = Pn Ga, 2-05 @k) PC O8,.« FH i2e) } +pn(ae, sy 8 .,ap—1)p (aa, oe »»A-1) . The next higher quotient N me Beetle, a ait) _ Ponta, - + Ap) +p nt1(Q, ss 6522-1) Pont2(a2,...,@2) frnti(as, .. .,ae)+p'n41(a2,.. . is obtained by employing the recurrence formulae Pni(a, oa -)@z) = pn(ay, Sys ., 24) p( ae, 4s ap) +Pn(a@1, .. .,42)p(as, . . .,@e) Pati(@a, . . .,@e) = pu(aa, . : 1,20,)P(a@2, . . .,az) +phn(ae, .. .,a2)p(as, .. .,a@e) Pnti(@y . . -:4e-1) = pn( ay, . . 52a) fh (a2, . . .,ae-1) +hn(a, + .,a2)p( as, » 274821) [24] Patil @2;-...,@e-1) = Pn(@s, . . 528) p(aa, ... .,@z-1) --pn( ag, + .,42)p( as, nyt jee) prla, sos .@2) = pn(a, a Ap) p (a2, vie -»&z) +hn( a, . . .,@%-1)p(aa, .. .,@%-1) Pu(Ga, . . -,42) = pn(aa, . ~ .;4%)p(a2, . : .,az) Ff —-pn(da; .'. »»@p-1)P( a2, . . »,@z-1) 365 12 Robert E. Moritz The values of the constantly recurring continuants A(2a,, .. .,@%), PC2ay, . . -5824) 5 Pas, Be) Plas > =), ag), Planeta P(a2, . . .,@z-1) are computed once for all, and then [23] and [24] ¢ are applied alternately in the computation of successive quotients. A different set of recurrence formulae similarly obtained can be employed in the computation of successive quotients when the number /V is of odd order. For example, POON, 12,2, 1eist 10) PCL EM 2 e-1 a) Pla, aes yl) —= 85, p(aa, oe abe ——s 8, . ‘plas, ... 2p). == 5; Pleas. 12,42) =e Play «+ +Ae-1) = 32, Pla2,..-,4z-1) = 3, p(4s, oa Rap) = 2, P(2a, a -,@k-1) ae 62, and now by successive substitution in [23] and [24] we readily obtain 113 a 8532? __ 16 009-489" _ 131 952°+-49 579" Reece 82+ 37° 1 5062— 46” ~—s 12 413% 4 664? __ 24 845 978°—758 918" __ 204 789 589?-+76 946 640” 2 337 313-— 71 393° 19 264. 984?-++ 7 238 531° _. 38,560 973 865°—1 £77 841 225? © 3°627-511 2827—" 110 301,982? 1 ee __ 317 833 574. 080?-- 119 421 234 8597 ~~ 29 899 267 5817+ 11 234 204 776" __ 59 846 656 284 458°—1 828 010 340 1187 __ 5 629 899 846 977°— 171 964 747 457” Of special interest is the inquiry under what conditions p(a,...,a1)/p(a@2,...,a2) represents an integer. Let us suppose all the elements but one of the continuant given. Let the variable element 1 occupy the th place. The problem then is, for what integral values of + will pla, oo yp, XV, Api14,.. -)Apti1,X,Ag_1,.. .,@) [25] p(aa, oe +p hh—-1,V, Apt, +. -,Api1,X Api)... ,Q2) represent integers? 366 Quotients of Sums and Differences of Perfect Squares 13 We expand the numerator of this quotient in terms of + by repeated application of [1] and find POY O41, %,Ae41, «Oe. % a1 1) = BY +2GL+C, where B= piper, C= pie-1PerepetPre1f1e—2 Pees C = 2p12-1Pia—2Patrapotbie-1 PateepetPie-2 Pee Similarly a 4 Or 4H etd, «oye, 0,2e—t, 0-22) = Adee 2k, where 2 2 A = prerpetiey, = p22-1Petpik+otpr,k-1f2,k—-2 Pee 2 2 F = $(2f2,0-1P2,2-2het1 ete tp2,e—1 Pipe, e-2tPok—-2 Det At It follows that the integral values of [25] for integral values of wv are given by the integral solutions of the equation 6 = Axry—Bx+2Hxy—2Gx-+2Fy—C = [26] Tor all values of k>1, this is a cubic form, the discussion of which can not be brought within the scope of a short paper, but for k=1, that is when the unknown element occupies the first place in the continyant of the numerator, we have, by virtue rar 2] B= Piobsa = Pap C—Piohas + PuoPy Poe = Paw C= 26,0P1,-1 Pant ProdsstPiaiPas = Pas A= Prabrs=, APD, Pao Par\ Poa = Oj 2F = 2p. Ps: Past PaoPaatPa—iPea = Posi since Par = Pio = 1, Poo =fPi-1 = 9, fo,-) = I- In this case we have therefore —$¢ = py? 2bo,3%—froV+f3,3 = 9; 367 14 Robert E. Moritg + apparently a quadratic form, but in essence linear. For since + and y are to be integers, +?—y will be integral, and putting this equal to g our equation becomes Poor +2host+h3,3 = 9, and it remains only to solve the congruence Post = —pyg(mod Pr2): [27] If this congruence has a root xo, which is always the casé ex- cept when 243, and f,, possess a divisor which is prime to fy, then _ be, a2,.. .,a2%) GREE a2) will represent an integer for every value X= Xo + A frp where A is any integers which will render -r) positive. 2) will then be given by — (20,34 +b, 8) dan ee eee ea and y) by The results just obtained are of importance in the theory of recurring continued fractions and may be used to construct with comparative ease a table much more extensive than the famous Canon Pellianum. For example, the integral values + which render DHT ky 2d, Tut) PCL lgarayt) integral are given by the congruence © 2p(1,2,1,1)*% == —(1,2,1) {mod p¢2,2,2,1,1) 368 ee Quotients of Sums and Differences of Perfect Squares that is by 14x == —4(mod 12). 1§ The smallest positive root of this congruence is 4, hence x is integral for every value X)\= 412A, X = 0,1,2,3,.... We have accordingly ~ SaAHNAOFWNWrO 10 and finally oN. 4 16 sae \ TeX 16 5 256 19 784 33 1600 47 2704. 61 4096 15 5776 89 7744 103 10000 ner 12544 131 15376 145 qPr= (*)51,1,2,1,1,2%) ). 369 TX, 21 275 817 1647 2765 4171 5865 7847 10117 12675 15521 Volumes Iand II of UNIVERSITY STUDIES are each complete in four numbers. Index and title-page for each volume is published separately. A list of the papers printed in the first two volumes may be had on application. Single numbers (excepting vol. I, no. 1, and vol. II, no. 3) may be had for $1.00 each. ~ A few copies of volumes I and II complete in numbers are still to be had. All communications regarding purchase or exchange should be addressed to THE UNIVERSITY OF NEBRASKA LIBRARY LINCOLN, NEB., U. S. A. JACOB NORTH & CO., PRINTERS, LINCOLN 3 9088 00876 4003