‘ . bye wee tea iVe a) auit ey oars Poeee ren a es ye tet VS eH fi JS BA / WH Ned baad tad bed KN Lh rove avaye vol VY VON | ea ere NU ev sees ¥ eA wy See, Ee ee covanahviiocoond he iS L_ \, hdl =a ded dh A Pe LULU (£ G C Qe ‘aq a@ “aCe € GC ‘wal veys h Vig | Vilsyiueune S, ever ‘es fez, Ran meanoa Yomuuvy il i Ie) We | i A ih \ L SSRN aS, ee v dl ey (a | NOACOU eee vise IGG AI A AS) LA vill © eh) ViPS Fi Neh iy © 1 vw, A et | oh Adi v) Ve Gane we yl eu Sfe ged YSIS, | errr a | 4 | | | AN h i S VIII { hd he wo WW Wevuel ye AN A h_A. \= See Nhe 1 5S Se eeu Gea c my C(( CC ce re 4) f '. " A, y ¥ We A re f" | ey i A SS DSI AA IIS VECeR ee fy A) rh“) —) —) We) | hed] WF} LAAAS Sr Be ie war cere ee WdUy DALE IMSS INS MY) IOI ery hw A te thy hdvdhvs) ) ul i) i mat J i WA] RS w w= = ~ WOW ivi, © oe eed tag nn Cou Seeoyasesece yo’, ly WE ECU el lal Sy edeevryeel Mee | AA ALAA SONS See asada revere i Meuse =A} YU ~w~ \ gue vssuceteue 3 \ A a — i\ .v, Meveverenvunnree.’ L REESE A) Vesey, VJs y “N WW Wieliv ied dtc , : v i VNC MELY Ivey Batic hae THE BIOLOGICAL BULLETIN PUBLISHED BY © THE MARINE BIOLOGICAL LABORATORY Editorial Board GarY N. CALKINS, Columbia University E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University - LEIGH HOADLEY, Harvard University _M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University E. E. JUST, Howard University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University W. M. WHEELER, Harvard University EDMUND B. WILSON, Columbia University ALFRED C. REDFIELD, Harvard University Managing Editor VOLUME LXXI AUGUST TO DECEMBER, 1936 Printed and Issued by LANCASTER PRESS, Inc. PRINCE & LEMON STS. LANCASTER, PA. il THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50. Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain: Wheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2. Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Mass., between June 1 and October 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the re- mainder of the year. Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. LANCASTER PRESS, INC., LANCASTER, PA. CONTENTS Nol.) Aucusr, 1936 PAGE Tuirty-EIGHTH REpoRT OF THE MARINE BIOLOGICAL LABORATORY 1 "WATERMAN, A. J. The Membranes and Germinal Vesicle of the Egg of Sabellaria “UIE ary Ae cna ety ck oie, Cn a RC RSE ui Cts Cid bea 46 TyLeR, ALBERT Oowhenccscties on Diterentiation, Lllr. 2.42. eee seme 59 TYLER, ALBERT OnetherPncioeressom Ditterentiation, UV. . i o.-2 naar 82 Harvey, ETHEL BROWNE Parthenogenetic Merogony or Cleavage without Nuclei in Ar- eet INTC LECT eN tA rere ea sas atee lobo s «AC ee id es 0h 0 alos eee ergs ee 101 Cor, W. R. Sequencer unctonal sexual Phases im Mleredo qeeer er). aa 122 RAKESTRAW, Norris W. ‘ The Occurrence and Significance of Nitrite in the Sea ....... 133 FisH, CHARLES J. The Biology of Oithona similis in the Gulf of Maine and Bay of EN GTOVOR RRP, cs s\n E ‘4 5.48 ie SCRE ae ae PPE PUR rel ne, ts 168 Beams, H. W., anp R. L. Kine The Effect of Ultracentrifuging upon Chick Embryonic Cells, with special reference to the “ Resting’’ Nucleus and the Mi- OUI SPC |S) ta ee wee ne oe ey sce ac: Done 5 4 ab ole 188 DELAMATER, ARLENE JOHNSON The Effects of Heavy Water upon the Fission Rate and the Life Cycle oi the Ciliate~Unolepiusmobilisy 5:2. - a oe eee ar 199 Younc, Rogpert T., AND Denis L. Fox The Structure and Function of the Gut in Surf Perches (Em- biotocidae) with reference to their Carotenoid Metabolism ... 217 MErrzC. WW Effects of Mechanical Distortion on the Structure of Salivary Gland. Chronitosomes » 4. 1c. eee eee renee nn ae 238 GEorRGE, W. C. The Role of Blood Cells in Excretion in Ascidians .......... 249 ill iv CONTENTS No 2. OcrToser, 1936 PAGE Parker, G. H. The Reactivation by Cutting of Severed Melanophore Nerves imythte oehish: Maistelis 72.) jae ae) | wee ee ree) AsprAmowlTz, A. A. Physiology of the Melanophore System in the Catfish, Ameiurus 259 SmitrH, G. M., anp C. W. Coates Cutaneous Melanosis in Lungfishes (Lepidosirenidae) Cow .es, R. P., anp C. E. BRAMBEL A Study of the Environmental Conditions in a Bog Pond with Special Reference to the Diurnal Vertical Distribution of Gon- WOStOMIUMM! SEMIEM) «Neca ere ct ate eeskals aos ces opens eae tearsren ila 286 HEILBRUNN, L. V. Protein Lipid Binding in Protoplasm Mazia DANIEL, AND JEAN M. CLARK Free Calcium in the Action of Stimulating Agents on Elodea Gelb es PG etter rare ee evel reac ee ley ters ar sic ee caer ee 306 ZOBELL, CLAUDE E., anD D. QUENTIN ANDERSON Observations on the Multiplication of Bacteria in Different Vol- umes of Stored Sea Water and the Influence of Oxygen Ten- Rp EAE eb bios Na 55 28/9) Sion wand « Solid Siaiphaces esa. g eve te 2 rae ais eo eter eee ees 324 TRAGER, WILLIAM The Utilization of Solutes by Mosquito Larvae ............. 343 Con. Wok: Environment and Sex in the Oviparous Oyster Ostrea virginica 353 GRAFFLIN, ALLAN L, Renal Function in Marine Teleosts. IV. The excretion of in- Oceanic phosphate rt dieses a ee ke ails eee eee 360 -Koonz, Cari H. Some Unusual Cytological Phenomena in the Spermatogenesis of a Haploid Parthenogenetic Hymenopteran, Aenoplex smithii G@Rackard’)" ie oa vs cele oes eeeentpe sos 2 ac ae cae B75 PROGRAM AND ABSTRACTS OF SCIENTIFIC MrrtiIncs, Marine Bio- LOGICAL LABORATORY, 1936 No. 3. DercemBer, 1936 WATERMAN, A. J. The Effect of Alcohol on Gastrulation in Arbacia ...........- 413 CoonFIELD, B. R. Regeneration in Minenuopsis leidyie Agassiz ee crete 421 CONTENTS Vv BowEN, R. E. Sensory Hairs of the Dogfish Ear GieR, HERSCHEL T. The Morphology and Behavior of the Intracellular Bacteroids Out IRORS IES ath inten Sete a evens Rin cre aca eer ee eee 433 Saxton, JoHN A., Jr., Mary M. SCHMECKEBIER AND Rosert W. KELLEY Autogenous Transplantations of Pigmented and Unpigmented Date Skiie itl (Guineame tesmues ei sn eh Reese mean ee) «tse ota 453 Guyver, M. F., anp Peart FE. CLaus Recovery Changes in Transplanted Anterior Pituitary Cells Strained: any thew Wipra-Cetmbniiticee |) Aner seis aeash aee 462 CaRoTHERS, E. ELEANOR Components of the Mitotic Spindle with Especial Reference to the Chromosomal and Interzonal Fibers in the Acridide ..... 469 Lackey, JAMEs B. Some Fresh Water Protozoa with Blue Chromatophores ..... 492 ERIKSSON-QUENSEL, INGA-BRITTA, AND THE SVEDBERG The Molecular Weights and pH-Stability Regions of the Hemo- CHAIMIAS) SL Se Shes ee eee meme rent en 5 ob Circo wc 498 UISSIDY ESSAYS er eA ee ace NT LSet Davee a eaelspene Bee. is Soar S > Volume LXXI_- ec a ae ees Pees Number 1 ee THE ~ : BIOLOGICAL BULLETIN THE MARINE BIOLOGICAL LABORATORY Editorial Board # : : Gary N. CALKINS, Columbia University E. G.-CONKLIN, Princeton University FRANK R. LILLIE, University of Chicago E. N. HARVEY, Princeton University : CARL R. MOORE, University of Chicago - SELIG HECHT, Columbia University GEORGE T. MOORE, Missouri Botanical Garden LEIGH HOADLEY, Harvard University T. H. MORGAN, California Institute of Technology - M. H« JACOBS, University of Pennsylvania G. H. PARKER, Harvard University H. S. JENNINGS, Johns Hopkins University W. M. WHEELER, Harvard University _E. E. Just, Howard University EDMUND B. WILSON, Columbia University ALFRED C. REDFIELD, Harvard University Managing Editor AUGUST, 1936 Printed and Issued by LANCASTER PRESS, Inc. _ PRINCE & LEMON STS.” LANCASTER, PA. THE BIOLOGICAL BULLETIN THE BIoLocIcAL BULLETIN is issued six times a year. Single © numbers, $1.75. Subscription per volume (3 numbers), $4.50. Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain: Wheldon & Wesley, Limited, 2, 3. and 4 Arthur Street, New Oxford Street, London, W.C. 2. Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Mass., between June 1 and October 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the remainder of the year. INSTRUCTIONS TO AUTHORS Preparation of Manuscript. In addition to the text matter, manuscripts should include a running page head of not more than thirty-five letters. Footnotes, tables, and legends for figures should be typed on separate sheets. Preparation of Figures. ‘The dimensions of the printed page (4% x7 inches) should be borne in mind in preparing figures for publication. Draw- ings and photographs, as well as any lettering upon them, should be large enough to remain clear and legible upon reduction to page size. Illustrations should be planned for sufficient reduction to permit legends to be set below them: In so far as possible, explanatory matter should be included in the legends, not lettered on the figures. Statements of magnification should take into account the amount of reduction necessary. Figures will be reproduced as line cuts or halftones. Figures intended for reproduction as line cuts should be drawn in India ink on white paper or blue-lined coordinate paper. Blue ink will not show in reproduction, so-that all guide lines, letters, etc. must be in India ink. Figures intended for reproduction as halftone plates should be grouped with as little waste space as possible.. Drawings and lettering for plates should be made directly on heavy Bristol board, not pasted on, as the outlines of pasted letters or drawings appear in the repro- duction unless removed by an expensive process. Methods of reproduction. not regularly employed by the Biological Bulletin will be used only at the author’s expense. The originals of illustrations will not be returned except by special request. Directions for Mailing. Manuscripts and,illustrations should be packed flat between stiff cardboards. Large charts and graphs may. be rolled and sent in a mailing tube. : Reprints. Authors will be furnished, free of charge, one hundred re-_ prints without covers. Additional copies may be obtained at cost. Proof. Page proof will be furnished only upon special request. When cross-references are made in the text, the material referred to should be ~ marked clearly on the galley proof in order that the proper page numbers may be supplied. — Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. Vol. LXXI, No. 1 August, 1936 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY iV INE BIOLOGICAL TABORATORY Tuirty-E1GHTH REPORT, FOR THE YEAR 1935— Forty-EIGHTH YEAR I, TRUSTEES AND EXECUTIVE CoMMITTEE (As oF AuGusT 13 1035). 2nd’. 450 ret rie aera ee eiPusinn te ih o7 ne) Mamie 1 TEVATRON T1010) Dae ea ee PMN AUD nae 5 Mere NCAAO EN(CORORAULON® 0050/4 ais a's aldve « sheamsuahe wenene Gusto tenes 3 ewe ya leAnShOnMnEIE CORPORATION .. 22). 6 ics soe oaledlae ersten 4 IN eee SPORT On mE URIDASURER. 6... .\c)as aa Baye sacseevaraliel atetecy a) eee 5 Wee cb BORN COE Ein UBIARTAN .. 5). alc st tke ete eeainer aan aepaneers 9 VAP eseORm (Or SEER MIOMRICTOR: ....°oatct-cre dvalus oa «lore see epee eee 10 SS eMC PR, aig us! Sie lG. ciatidlenst's Gudea sae Sone 10 Addenda: lela ines Sante OS) 4 6's) «su. ois -de capslel o eae ae er ee 14 2 eolrivestt@uarseand Students, 1935)\9.. satel ae 16 Seababulaneview or Attendance .i.)s..c0aoeeu nose 26 4. Subscribing and Cooperating Institutions, 1935 .... 27 Spebvenimevivectures, 1935). . f:.0. 0s Ae ee 28 Omshonermoccutiie Papers, 1935 32.5. 0552 4.0) eee 28 PeNieninenswen the Corporation, 2s.) 0-2 eee eee 33 i Wisi Bis EX OFFICIO Frank R. Lituie, President of the Corporation, The University of Chicago. MerRKEL H. Jacogs, Director, University of Pennsylvania. Lawrason Ric6s, Jr., Treasurer, 120 Broadway, New York City. CHARLES PACKARD, Clerk of the Corporation, Columbia University. EMERITUS H. C. Bumpus, Brown University. E. G. Conxuin, Princeton University. C. R. Crane, New York City. H. H. Donatpson, Wistar Institute of Anatomy and Biology. R. A. Harper, Columbia University. M. M. Metcarr, Waban, Mass. G. H. Parker, Harvard University. ae 2 i) MARINE BIOLOGICAL LABORATORY W. B. Scort, Princeton University. W. M. Wuee.er, Harvard University. E. B. Witson, Columbia University. TO SERVE UNTIL 1939 W.C. Atesz, The University of Chicago. Gary N. CaLxins, Columbia University. B. M. Duccar, University of Wisconsin. L. V. HeireruNN, University of Pennsylvania. L. IrvinG, University of Toronto. W. J. V. OsterHout, Member of the Rockefeller Institute for Medical Re- search. A. H. Sturtevant, California Institute of Technology. LoraNpbE L. WoopruFF, Yale University. TO SERVE UNTIL 1938 E. R. Crark, University of Pennsylvania. Otto C. Guiaser, Amherst College. Ross G. Harrison, Yale University. E. N. Harvey, Princeton University. H. S. Jennines, Johns Hopkins University. F. P. Knowtton, Syracuse University. FRANZ SCHRADER, Columbia University. B. H. Wiurer, University of Rochester. TO SERVE UNTIL 1937 W.R. AmBerson, University of Tennessee. H. B. Goopricu, Wesleyan University. I. F. Lewis, University of Virginia. R. S. Liti1e, The University of Chicago. T. H. Morean, California Institute of Technology. A. C. RepFIELD, Harvard University. C. C. SpempeL, University of Virginia. D. H. TENNENT, Bryn Mawr College. TO SERVE UNTIL 1936 H. B. BiceLow, Harvard University. R. Cuamsers, Washington Square College, New York University. W. E. Garrey, Vanderbilt University Medical School. CaAswELL Grave, Washington University. M. J. Greenman, Wistar Institute of Anatomy and Biology. A. P. MatHews, University of Cincinnati. C. E. McCune, University of Pennsylvania. C. R. Stocxarp, Cornell University Medical College. EXECUTIVE COMMITTEE OF THE BOARD OF TRUSTEES Frank R. Litiir, Ex. Off. Chairman. MerKkeEt H. Jaccss, Ex. Off. Lawrason Rices, Jr., Ex. Off. G. N. CaLxins, to serve until 1936. ACT OF INCORPORATION 3 HEILBRUNN, to serve until 1936. KNow ton, to serve until 1937. I 13, N 18}. Wier, to serve until 1937. Tue Lisprary COMMITTEE E. G. ConKuLin, Chairman. WILLIAM R. AMBERSON. Cy Omiserin,, LU: C. C. SPEIDEL. -A. H. STURTEVANT. WILLIAM R. TAYLOR. THE APPARATUS COMMITTEE . V. HEILBRUNN, Chairman. R. AMBERSON. J. Epwarps. E Z0er . GARREY. HARVEY. . JACOBS. N H = Il. ACT OF INCORPORATION No. 3170 CoMMONWEALTH OF MASSACHUSETTS Be It Known, That whereas Alpheus Hyatt, William Sanford Stevens, William T. Sedgwick, Edward G. Gardiner, Susan Minns, Charles Sedg- wick Minot, Samuel Wells, William G. Farlow, Anna D. Phillips and B. H. Van Vleck have associated themselves with the intention of forming a Corporation under the name of the Marine Biological Laboratory, for the purpose of establishing and maintaining a laboratory or station for scien- tific study and investigation, and a school for instruction in biology and natural history, and have complied with the provisions of the statutes of this Commonwealth in such case made and provided, as appears from the cer- tificate of the President, Treasurer, and Trustees of said Corporation, duly approved by the Commissioner of Corporations, and recorded in this office; Now, therefore, I, Henry B. Pierce, Secretary of the Commonwealth of Massachusetts, do hereby certify that said A. Hyatt, W. S. Stevens, W. T. Sedgwick, E. G. Gardiner, S. Minns, C. S. Minot, S. Wells, W. G. Farlow, A. D. Phillips, and B. H. Van Vleck, their associates and suc- , cessors, are legally organized and established as, and are hereby made, an existing Corporation, under the name of the MARINE BIOLOGICAL LABORATORY, with the powers, rights, and privileges, and subject to the limitations, duties, and restrictions, which by law appertain thereto. Witness my official signature hereunto subscribed, and the seal of the Commonwealth of Massachusetts hereunto affixed, this twentieth day of March, in the year of our Lord One Thousand Eight Hundred and Eighty- Eight. [SEAL ] EBINRY © BR: PIERCE, Secretary of the Commonwealth. 4 MARINE BIOLOGICAL LABORATORY III. BY-LAWS OF THE CORPORATION OF THE MARINE BIOLOGICAL LABORATORY I. The annual meeting of the members shall be held on the second Tuesday in August, at the Laboratory, in Woods Hole, Mass., at 11.30 A.M., daylight saving time, in each year, and at such meeting the members shall choose by ballot a Treasurer and a Clerk to serve one year, and eight Trustees to serve four years. There shall be thirty-two Trustees thus chosen divided into four classes, each to serve four years, and in addition there shall be two groups of Trustees as follows: (a) Trustees ex officio, who shall be the President of the Corporation, the Director of the Laboratory, the Associate Director, the Treasurer and the Clerk; (6) Trustees Emeritus, who shall be elected from the Trustees by the Corporation. Any regular Trustee who has attained the age of seventy years shall continue to serve as Trustee until the next annual meeting of the Corporation, whereupon his office as regular Trustee shall become vacant and be filled by election by the Cor- poration and he shall become eligible for election as Trustee Emeritus for life. The Trustees ex officio and Emeritus shall have all rights of the Trustees except that Trustees Emeritus shall not have the right to vote. The Trustees and officers shall hold their respective offices until their successors are chosen and have qualified in their stead. II. Special meetings of the members may be called by the Trustees to be held in Boston or in Woods Hole at such time and place as may be designated. III. Inasmuch as the time and place of the Annual Meeting of Members is fixed by these By-laws, no notice of the Annual Meeting need be given. Notice of any special meeting of members, however, shall be given by the Clerk by mailing notice of the time and place and purpose of said meeting, at least fifteen (15) days before such meeting, to each member at his or her address as shown on the records of the Corporation. IV. Twenty-five members shall constitute a quorum at any meeting. V. The Trustees shall have the control and management of the affairs of the Corporation; they shall present a report of its condition at every annual meeting; they shall elect one of their number President of the Cor- poration who shall also be Chairman of the Board of Trustees; they shall appoint a Director of the Laboratory; and they may choose such other officers and agents as they may think best; they may fix the compensation and define the duties of all the officers and agents; and may remove them, or any of them, except those chosen by the members, at any time; they may fill vacancies occurring in any manner in their own number or in any of the offices. They shall from time to time elect members to the Corporation upon such terms and conditions as they may think best. VI. Meetings of the Trustees’shall be called by the President, or by any two Trustees, and the Secretary shall give notice thereof by written or printed notice sent to each Trustee by mail, postpaid. Seven Trustees shall constitute a quorum for the transaction of business. The Board of Trustees shall have power to choose an Executive Committee from their own number, and to delegate to such Committee such of their own powers as they may deem expedient. REPORT OF THE TREASURER 5 VII. The accounts of the Treasurer shall be audited annually by a certified public accountant. VIII. The consent of every Trustee shall be necessary to dissolution of the Marine Biological Laboratory. In case of dissolution, the property shall be disposed of in such manner and upon such terms as shall be de- termined by the affirmative vote of two-thirds of the Board of Trustees. IX. These By-laws may be altered at any meeting of the Trustees, pro- vided that the notice of such meeting shall state that an alteration of the By-laws will be acted upon. X. Any member in good standing may vote at any meeting, either in person or by proxy duly executed. IW) Msi MeO (ON aMeslidt IRIE A SIURUO IR To THE TRUSTEES OF THE MARINE BIoLoGIcAL LABORATORY : ’ Gentlemen: Herewith is my report as Treasurer of the Marine Bi- ological Laboratory for the year 1935. The accounts have been audited by Messrs. Seaman, Stetson and Tuttle, certified public accountants. A copy of their report is on file at the Laboratory and is open to inspection by members of the Corporation. At the end of the year 1935, the book value of the Endowment Funds in the hands of the Central Hanover Bank and Trust Company as Trustee, was CEU CKO EMO a SCCUIRICS 5 oa sled cose o's ob due yen ded Se $ 884,683.36 (CAISIIM © 51 SION a anaemia Aa Ea So's 26,482.57 iLabraarsy Patna: SECU Oa eae retary nner Me mate ie 7 193,582.06 (Cala Ate Ee ae ee mera OMe nee Te. f 101.19 Mictaltbools valuCMeRr ee aay so. fo.c aunts tee been $1,104,849.18 The income collected from these Funds was as follows: Geral Ed OWUIN CME | Madeira ose) ok ig, Sete gw ehtete t Sain Mee ae $40,479.31 ILD OTROU EN aT RO IPS Os NSE, 33) she 2. an Rg ae LOPE donner A 7,471.45 $47,950.76 an increase of a little more than one thousand dollars ($1,000) over the income from these Funds in 1934. The income in arrears, some of which may never be collected, was on December 31, 1935: (COREG! TEC, Pe Cee te eter ere tan CL raked nye $13,944.05 ETAT OaN HE! LON ayctecag S255 Vicious. cote SRP RIS tay eps RI eee Re 4,150.00 The extra dividends from the stock of the General Biological Supply 6 MARINE BIOLOGICAL LABORATORY House have continued; the total dividends received from that Company amounted during the year to $10,922.00. Retirement Fund. and 15>, and one at 22> and 12>. Inthe first two columns of each group the time at which 50 per cent of the eggs have TABLE [| Ciona; times of attainment of various stages for three batches of eggs, each run at two different temperatures; the time for the cleavage stages is given in minutes, that for hatching, in hours; the coefficients, Q; and Qio are the ratios of the time at the lower temperature to the time at the higher temperature. 20.0° 15.0° Os 25.0° 15.0° Quo 22202 12.0° O10 PCO tench 47.2) 77.0] 1.63 | 34.9 | 77.0} 2.21 | 41.9 | 118.5 | 2.83 ARCs ter, shes Oyen. A: 69.0} 114.0} 1.65 | 50.5 | 114.5] 2.27 | 64.0 | 178.0] 2.78 BECOME, water ayy «yes a 9157) 14910) 15635 )) 66:17 15055) 72-28) | 78-5) 224.5) 2.80 MORCelle shee. eh aa s - 116.0 | 195.0} 1.68 | 86.0 | 194.5} 2.26 PE ChiIm oo snk TARA eZ RAN ele SLOP 2 Seas 2h oO lites reached the indicated stage is given. In the third column are the ratios (Q; or Oi) of these values for the lower and higher temperature. In Table II a summary of these ratios for a number of determinations is given. The data are presented for each pair of temperatures separately since, as pointed out in the previous section, each batch of eggs was run at only two different temperatures. The variation in time of develop- ment that is obtained with eggs from different individuals may be seen from the following figures for the time of first division in five separate iubaeyele BAO Oh (One 47.2 48.2 47.0 48.3 48.0 Av. = 47.7 ad. = 0.5 The variation that is obtained at other stages and at other temperatures is of the same relative order of magnitude. The ratios, as shown in Table II, vary less on the whole than do the values for the time of division. In other words, even when the time of development is different in two batches of eggs run at the same two temperatures, the 66 ALBERT TYLER ratios of the time at the different temperatures for each batch tend to be more nearly alike. It can be seen from Tables I and II that the temperature coefficients for the rate of attainment of the various cleavage stages are alike. We need not present the deviation measures to demonstrate this, since it TABLE II Ciona; temperature coefficients; Q; and Qio represent the time at the lower temperature divided by the time at the higher temperature. Qs Ono Quo 15.0°/20.0° 15.0°/25.0° 12.0°/22.0° 2-cell 1.63 2.21 2.83 1.66 2.30 2.91 1.64 2.24 2.85 1.66 (Bees) 2.86 1.63 1.63 DoS 2.86 4-cell 1.65 Dll 2.78 1.65 2.32 2.81 1.67 2.29 2.83 1.63 2.27 2.85 1.65 2.29 2.84 8-cell 1.63 2.28 2.81 1.64 Deo 2.85 1.62 1.63 2.30 2.83 16-cell 1.68 2.26 1.67 2.25 1.63 1.66 2.26 Hatching lois 2.50 1.76 2.61 1.74 1.74 2.56 can be readily seen from Table I that for the same batch of eggs there are no consistent deviations in the coefficients for different stages, and from Table II that for different batches of eggs the range of variation of the average coefficients for the different stages lies within the range of variation of the coefficients obtained for one stage. DEVELOPMENTAL RATE AND TEMPERATURE 67 The temperature coefficients for hatching, however, are apparently greater than for the cleavage stages. At the lowest temperature employed, namely 12°, no hatching was obtained in four separate experiments, although development appeared to be normal in all cases. _ In three other runs at this temperature cleavage was interfered with. As a possible interpretation of the effect of temperature on hatching, the following might be suggested. Let us assume that hatching de- pends upon the attainment of a certain concentration of a particular enzyme and that the effect of temperature on the rate of production of d Fic. 1. Dendraster embryos. a, beginning gastrula; 6, early prism stage, dorsal view; c, early prism stage, side view; d, late prism stage, ventral view; e, middle pluteus stage; f, late pluteus stage. the enzyme is the same as its effect on the rate of development. The rate at which the enzyme diffuses away from the embryo as well as its rate of production will then determine its concentration in the mem- brane. Since the physical process of diffusion would have a lower temperature coefficient than the chemical processes involved in the production of the enzyme, the rate of diffusion would not be slowed up as much, when the temperature is lowered, as the rate of production. This would lead to larger values for the temperature coefficjent of hatching as compared with that for development. For the failure to 68 ALBERT TYLER hatch at a low temperature at which development is normal, we might assume in addition that the enzyme must attain a certain minimum concentration in order to be effective. Considerably more information than has yet been obtained would be necessary to test these points. Dendraster excentricus In this form temperature coefficients for gastrulation, prism and pluteus stages as well as for cleavage stages were determined. Ina few cases hatching-time was also noted. Some of the stages examined are illustrated in Fig. 1. During gastrulation embryos that are 1/8, 1/4, 1/2, 3/4 or completely gastrulated can be readily distinguished. At the completion of gastrulation the sand-dollar embryo has become flattened on one side (the anterior side in consideration of the direction of TABLE III Dendraster; same description as Table I 20.0° 15.0° Os 15.0° 10.0° Qs 22.0° 12.0° | Qo DE CET yeas etiea'sir an ae Sey 55.0] 89.5 | 1.63} 90.0] 184.0] 2.04) 46.8 | 135.0} 2.88 A= GO I RUAN fit) 8 atest 86.5 | 138.0 | 1.60} 139.5 | 281.0] 2.01] 72.0 | 202.0} 2.81 S-Ce lake teh slat Moker ae 109.0 | 173.0 | 1.59) 174.0 | 336.0 | 1.93) 98.0 | 278.0 | 2.84 WG=celleenras. cin senate le 126.0 | 203.0 | 1.61 — | — Elatchingeuy. sae Goo: 7.6) 11.8) 1.55) — — | — 7.0 | 17.4] 2.49 1/8 gastrulated......... 12.7] 20.9] 1.65} 20.2} 39.8] 1.97) 10.9 | 30.4] 2.79 1/4 gastrulated......... 13.2) 21.5] 1.63} — — | — 1/2 gastrulated......... 14.2} 22.9} 1.61) 23.1] 46.0} 1.99) — = |= 3/4 gastrulated......... 15.0| 24.1} 1.61) 24.6} 49.2} 2.00} 12.6 | 36.0] 2.82 Early prism............ 15.9| 26.4] 1.66} 26.3) 53.0] 2.01 — Eatetprisme rs acnn esse 20.8] 33.1] 1.59) — — |— Early pluteus........... 23.5| 38.6] 1.64) 38.0} 77.5] 2.04; — — |— Middle pluteus......... 33.1) 54.6} 1.65} 55.0] 113.5] 2.06} — = | = Late pluteus........... — — |—] — — | — | 48.7 | 142.0 | 2.92 swimming of the pluteus), and may at that time be designated as'the early prism stage (Fig. 16 and c). Later, that side of the embryo becomes concave and the archenteric pouches appear. We may desig- nate this stage as the late prism (Fig. 1d). Following this the aboral (ventral) arms appear and the gut becomes tripartite, forming the early pluteus. The next stage is just before the appearance of the oral (dorsal) arms, and may be called the middle pluteus stage (Fig. le). Before this stage the anterior border of the oral lobe is convex and as the oral arms form it becomes concave. This change takes place fairly rapidly and so the time at which this stage is reached can be determined with some accuracy. For the late pluteus stage (Fig. 1f) the length of the oral arms serves as a simple criterion in identifying embryos that are comparable. Table III gives the data for single batches of eggs, one run at DEVELOPMENTAL RATE AND TEMPERATURE 69 20.0° and 15.0°, one at 15.0° and 10.0° and one at 22.0° and 12.0°. The ratios of the time at the lower temperature to that at the higher temperature are given in the third, sixth and ninth columns of the table (Qs; and Qi). In Table IV the values of these ratios for five pairs of temperatures are given. Each value is, of course, obtained TABLE IV Dendraster; same description as Table II 15.0°/20.0° 10.0°/15.0° 12.0°/22.0° 15.0°/25.0°| 8.0°/18.0° 2-cell stage 1.63 2.04 2.88 2.21 3.91 1.64 1.61 1.95 De hS 2.18 3.85 1.59 1.61 2.00 2.79 Mohs 1.63 2.06 2.82 1.62 2.01 2.82 2.21 3.88 4,8 and 16-cell 1.60 2.01 2.81 2.26 3.82 stage 1.62 1.92 Dil Boll's 2 Ja) 1.58 1.94 2.80 2.85 1.59 2.78 1.60 1.96 2.80 2.24 3.82 Gastrulation 1.65 1.97 2.79 Dele 3.83 1.63 1.66 1.99 2.82 Deg aps 3.94 (OL > 1262 2.00 2.79 1.61 1.61 1.93 2.76 1.58 1.59 2.00 1.59 1.60 2.00 2.03 1.60 1.99 2.79 2.20 3.98 Prism 1.66 2.01 2.79 BN 3.90 1.59 2.08 3.82 2.02 1.63 2.04 2.79 Mell 3.86 Pluteus 1.64 2.04 2.85 3.86 1.65 2.06 2.92 4.01 2.03 2.76 3.95 2.00 2.82 1.65 2.03 2.84 3.94 from a single batch of eggs run at the two temperatures. In some cases values for a number of stages (as is the case in Table I) were obtained from a single batch of eggs. It can be readily seen from the tables that the temperature coefficients for the various cleavage stages are the same. Furthermore, the coefficients for the later stages fall 70 ALBERT TYLER TABLE V Urechts; same description as Table I. 20.0° 15.0° Os 22.0° 2-02 O10 CAM 7, oko 74.5 120.0 1.61 65.0 180.5 2.78 4-cell...... 104.0 162.0 1.56 88.0 249.0 2.83 8-cell...... 129.0 206.5 1.60 1iS}55) 319.0 2.81 NO-cellaee 158.5 254.5 1.61 39.5 385.0 2.76 within the range of values obtained for the cleavage stages. The range of temperatures (8.0° to 25.0°) covered is practically that outside of which normal development can no longer be obtained. If the figures obtained for Dendraster are compared with those for Ciona it may be seen that they are very much alike. The average values for all the cleavage stages are given in Table XIII along with those of the other forms investigated. This similarity in the coef- ficients for Dendraster and Ciona may best be discussed after consider- ing the other forms. ' Urechts caupo - Only the data for the cleavage stages of Urechis eggs are presented here, inasmuch as they present the same difficulties as Ciona for ob- taining reproducible results in the later stages. The results given in Tables V and VI show that for the eggs of this form too the temperature TABLE VI Urechis; same description as Table II. 15.0°/20.0° 10.0°/15.0° 12.0°/22.0° 2-cell 1.61 1.95 2.78 1.59 2.08 2.76 1.58 Dats) 2.84 1.60 1.64 1.63 1.61 2.06 2.79 4, 8 and 16-cell 1.56 1.97 2.83 1.62 2.04 2.79 1.65 2.11 2.81 1.60 Presta 1.63 2.78 1.59 2G 1.61 2.82 1.61 2.04 2.79 DEVELOPMENTAL RATE AND TEMPERATURE 711 coefficients of the rate of attainment of the various cleavage stages are alike. It may also be readily seen that the values obtained with Urechis eggs are quite similar to those obtained with eggs of Dendraster and Ciona at the same temperatures. Strongylocentrotus purpuratus Gastrulation and prism stages as well as cleavage stages of the Strongylocentrotus eggs were examined. The extent of the gut cavity TABLE VII Strongylocentrotus; same description as Table I. 20.0° 15.0° Os 20.0° 7.5° Ons DEM, asthe es Rea ee iM 77.0 | 105.0 1.36 76.3 | 227.0 2.98 Aalst a De ealins ehanetes Otay iat eee et 126.0 | 174.0 1.38 125.0 | 383.0 3.06 Wiosoastrulated ine cm sce ae ss: 25.0 S2 mii 1.31 24.7 18 2.98 HAicastrulated sa. 5 aq) oe ome 28.8 37.0 1.29 28.5 89.0 3.12 Ss Aseastruilateds, acorn oan se 31.4 42.0 1.34 — _ — Prism with skeletal rods....... 49.0 59.0 1.34 48.0 | 140.0 2.92 TaBLeE VIII Strongylocentrotus; same description as Table II. Os O1.5 15.0°/20.0° 7.5°/20.0° Dank Cella eta vient he caste eke 1.36 2.98 1.30 2.89 1.31 2.97 1.38 3.08 1.30 3.06 1233 3.00 Gastrulation sy valcke seus cea roe ace 1.31 2.98 Si 3.04 1.29 Soll 1.39 3.09 Sy 1.35 3.06 rather than the extent of invaginated cells was used as a criterion for determining the stage of gastrulation. This was done because the primary mesenchyme that covers and extends forward from the gut does not change very markedly in appearance as gastrulation pro- ceeds. Although the results with Strongylocentrotus vary more than with the forms previously presented, it may be readily seen from Tables VII and VIII that no consistent differences occur in the 72 ALBERT TYLER TABLE IX Lytechinus; same description as Table I. 15.0° 10.0° Os DECOR RA ee ts aa\ 1 hoe aressie a che 137.0 281.0 2.05 /Sreactnmlatedinnr yn ely sil 30.2 63.5 2.08 W2ieastrulated . Jyh) sc. 0. 35.0 73.0 2.09 Barly spristds..Sonsb.sek Gee: 58.5 116.5 1.99 Bate sprisine 2st 6G he eee 73.0 151.0 2.07 temperature coefficients of the cleavage and the later stages. The average values of the temperature coefficients for Strongylocentrotus differ considerably from those for Ciona, Dendraster and Urechis at the same temperatures, as may be seen in Table XIII. TABLE X Lytechinus; same description as Table II. Os 10.0°/15.0° Clea va sec iene SUA tice 12 yh ca vag eek) ae 2.05 1.91 1.87 1.94 Gasenwlationet a weeeme ee see yak ie cca mk rely geile keller 2.08 2.04 2.09 2.07 PFI STIN Mac eee Tea a tee eats vee ee 1.99 2.10 2.07 2.05 Lytechinus anamesus The time of attainment of the two-cell stage, gastrulation and prism stages was determined for the eggs of this sea urchin at only one pair of temperatures. The results are presented in Tables IX and X. The data are less extensive than for the other forms; neverthe- less it can be seen that the temperature coefficients for the later stages are very much like those of the two-cell stage. A comparison of the average value with those of the other forms is given in Table XIII. DEVELOPMENTAL RATE AND TEMPERATURE 7S) Number of Cells in Sea-urchin Gastrule Raised at Different Temperatures From the similarity of the temperature coefficients of the gastrula and the early cleavage stages, it is to be expected that the number of cells present in gastrule raised at different temperatures should be the same, provided, of course, that the later cleavage stages have the same temperature coefficient as the earlier ones. Since the data in the early literature (see above) showed different numbers of cells present in embryos raised at different temperatures, this point was re- investigated. Two methods were used in estimating the cell number, that of Marcus (1906) and that of Koehler (1912). Marcus’ method con- sisted in counting the number of nuclei (mz) in an optical section at a great circle. Then the total number of cells (JV) is given by 2 n( 2m) = 4777, nN where ¢ is the radius of the embryo. It is assumed that the surface outline of each cell is a square. This equation reduces to n* N=— T and it is unnecessary to measure the radius as Marcus did. Koehler counted the number of nuclei in a small area of the surface. The number of nuclei per unit area (m/s?) multiplied by the surface area of the embryo gives the total number of cells nN N => Arr’: 2? S where the radius, 7, is measured to the middle of the wall. It would be better, however, to measure the radius to a point one-third the distance between the inner and outer surfaces of the wall, since the centers of the nuclei correspond to about that position. For embryos that have gastrulated corrections must, of course, be made for the invaginated cells, if the total cell number is desired. The measurements presented here were made on embryos that had just begun gastrulation. The number of mesenchyme cells could be esti- mated roughly and the small invagination could be considered to occupy the same area as if the cells were still on the spherical surface. The surface area was calculated from the formula for an oblate-sphe- roid. Neither of the two methods gave very accurate results. Koehler claims that his method is better since one is not troubled with having the optical section coincide with the greatest circle and with higher and 74 PRI IE VIRAL eye TABLE XI Strongylocentrotus; number of cells in beginning gastrule. A B 20.0° 75? 20.0° eae 680 680 776 688 769 799 780 815 623 680 705 863 651 738 690 752 799 708 802 885 799 644 801 829 769 597 854 780 680 608 806 809 721 682 777 803 lower nuclei that might or might not be in the great circle. But in Koehler’s method the unit square of surface in which the nuclei are counted is actually a spherical surface and if a large area is taken, then too many nuclei are counted, whereas if a small area is taken the error due to the curvature is negligible but the counting becomes more difficult because a considerable fraction of the nuclei are cut by the boundaries of the area. In Table XI the cell numbers are given for eight beginning gastrule raised at 20.0° and eight of the same batch raised at 7.5° C. The counts were made on slides of preserved embryos, only those in side view being chosen. The first two columns (A) were obtained from counts of the nuclei in optical section. The second two columns (B) are from counts of nuclei in a unit area of surface. Figures 2a and b illustrate the nuclei in optical section and Figs. 2a’ and 0b’ the surface nuclei of a 20.0° and a 7.5° embryo. These two embryos happened to show the same number of nuclei (45) in optical section and gave a total (including 35 mesenchyme cells) of 680 cells. The cell numbers from the counts of the surface nuclei are 776 and 688 respectively. The other values presented in the table vary considerably, yet there are no consistent differences between the warm and cold series. DISCUSSION Numerous methods have been employed by biologists to express the relation between temperature and the rate of biological processes. An extensive review is given by Bélehradek (1935). Although we are not especially concerned here with establishing any particular relation, DEVELOPMENTAL RATE AND TEMPERATURE 75 it is of some interest to examine how well two of the commonly em- ployed formule fit our data. We shall consider first the Arrhenius equation which has been extensively applied by Crozier and his col- laborators (1924—) to various biological processes including the rate of development. If the logarithm of the velocity of the particular process is plotted against the reciprocal of the absolute temperature, then, according to the Arrhenius equation, a straight line should be expected. In the graph of Fig. 3 the data for the first cleavage of o= —-0-O-60 Oo © Cro One b oO>6d b/ ey OK a @QQO ? OO Lo Oo 0-0 O O—O Fic. 2. Strongylocentrotus beginning gastrulz: a, optical section of embryo raised at 20.0° C.; a’, nuclei in unit area of surface of same embryo; 8, optical section of embryo raised at 7.5° C.; 6’, nuclei in unit area of surface of same embryo. Ciona, Dendraster and Urechis are so plotted. Only the values for the first division were used since these were the most reliable ones obtained. The figures from which the graph was obtained are given in Table XII. These figures do not represent averages of all the first cleavage values but were taken from cases where the cleavage time obtained at one temperature was practically identical in two separate runs in which the temperature of the second bath had been changed. This was done in order to cut down the variability that would be obtained with different batches of eggs and, of course, because only two temperature baths were available at one time. The graphs of Fig. 3 are certainly not straight lines. Small portions of these graphs might be considered as such. But in order to conclude that the 76 AEB ER Yar Ea: TABLE XII Time for first cleavage. Ciona Dendraster Urechis SOME ee ene aS ke 248.0 OO Pere he senet ene 184.0 262.0 WAH Re Bes Nike nes tenes ota 117.9 135.0 183.5 ILS) coke teeta day enero 77.0 89.5 120.0 S(O) eee ate Bch oneReee 56.5 63.5 89.5 DOOR a bras Gly Suheee ake 47.7 54.75 74.5 DOO Arias eset teats 42.2 46.75 64.5 DSO PM day serie Nes Neem tote : : 35.1 40.5 Arrhenius equation applies we would need to make the now prevalent but as yet unproved assumption of different limiting reactions of a catenary series at different temperature intervals. If such an as- sumption were to be made, it would be desirable to have considerably more points on the graph. le pA =10,700 S S 8 [.=10,540 =16,250 = Q. 1.20 S M= 16,680 S S M=22,950 0.95 w Ciona E S S< 8 ) s Dendraster (lL) SI S M=25,|50 0.70) 350 0.00355 0,00340 Yr (Abs) 0.00345 0.00: Fic. 3. First cleavage data of Table XII plotted as logarithm of the rate against the absolute temperature. There is a rather simple relation between temperature and the rates of biological processes that was first applied to developing eggs by Krogh (1914). This relation is simply that the rate is directly pro- DEVELOPMENTAL RATE AND TEMPERATURE 77 portional to the temperature. Support for this view has been ob- tained by a number of investigators, and recently Ephrussi (1933) has shown that the effect of temperature on the rate of cleavage of sea- urchin eggs is fairly well expressed by it. In Fig. 4 the data of Table XII are plotted as the reciprocal of the time for first cleavage against the temperature. The points fall fairly well on a straight line for most of the temperature range, as they should if the rate is directly proportional to the temperature. At the lower temperatures devia- tions occur in the graphs for Dendraster and Urechis. The graph for Ciona, the data for which are more reliable than in the case of the other forms, is a very good straight line. Dendraster Urechis 100 TempERATURE (CenT.) 15.0 200 250 Fic. 4. First cleavage data.of Table XII plotted as rate against centigrade temperature. The significance of this linear relation between the rate of a bio- logical process and temperature is quite obscure. Bélehradek employs a modified equation which may be written v= a(t — a)? in which v is the velocity, ¢ the temperature and a, a and 0 constants. When the exponent 6 = 1, this becomes the linear relation. It is assumed that the temperature coefficients give a measure of the rela- tive viscosity although just how the constants of the formula are 78 AEB Rabe ale Be related to viscosity is not clear. Support of this view is given by Bélehradek and Mladek (1934), who found higher values for the temperature coefficients of respiration of more viscous meal-worm larve than of less viscous ones. Bodine and Thompson (1935) showed, however, that this is not necessarily the case, inasmuch as dehydrated grasshopper embryos gave the same values for the temper- ature coefficient of respiration as normal embryos. If the temperature coefficients for the five forms investigated are compared, it is seen that four of them, Ciona, Dendraster, Urechis and Lytechinus, are very similar, whereas Strongylocentrotus gives values that are quite different. In Table XIII the mean values for the cleavage stages are presented. An examination of the values for the temperatures 15° and 20° (first column of table) shows that the coefficient for Strongylocentrotus is much lower than for the others. At the temperatures 7.5° and 20° the figure for Strongylocentrotus in TABLE XIII : Temperature coefficients for cleavage. (Mean values and their standard devia- tions taken for all the cleavage data.) Os Oro Os Quo QO1e.5 15°/20° 15°/25° 10°/15° 12°/22° 7.5°/20° Cionane a ee 1.65+0.018 | 2.270.033 2.84+£0.035 Dendraster.....| 1.61-+0.02 | 2.22+0.032 | 1.99+-0.053 | 2.80+0.038 4.7* Urechis. 54.205 1.61 0.026 2.050.079 | 2.790.029 Lytechinus..... == 1.94 —— Strongylo- centrotus....| 1.330.037 oo 3.00 +0.076 * Calculated value. this table is 3.0, whereas the value for Dendraster in the temperature interval 8° and 18° is 3.9 (Table IV) and the extrapolated value for (eS ands2 Op risieede In view of the striking difference shown by Sirongylocentrotus, it is hardly worthwhile at present to try to account for the similarity shown by the others. The closeness of taxonomic relationship evidently has no bearing on the results, inasmuch as the two other echinoderms, Dendrasier and Lytechinus, differ from Strongylocentrotus, whereas they are like Ciona and Urechis. It also appears that the eggs of one species may have different temperature coefficients, as Ephrussi (1933) has shown for the results of Horstadius (1925) on the rate of cleavage of Paracentrotus lividus eggs in summer and in winter. This particular case is taken by Ephrussi to illustrate the rule of Skrabal (1916) that fast reactions have lower temperature coefficients DEVELOPMENTAL RATE AND TEMPERATURE 79 than slow reactions. Zawadowsky and Sidoroy (1928) present values for the cleavage of sea urchins (three species), frog and Ascaris eggs that are in agreement with that rule. The data presented in this paper, however, do not fit into that scheme even in the case of the three echinoderms alone. For example, Strvongylocentrotus has a slower rate of cleavage than the other two echinoderms and also a lower temper- ature coefficient. It is the relation between the temperature coefficients for cleavage and for organogenesis that is of particular interest here. That-the various early cleavage stages have the same temperature coefficients is perhaps to be expected since the early cleavage stages represent the same sort of process. This statement must be qualified, however, for development up to the first cleavage since in addition to the kind of events occurring in the later cleavages there is also fertilization, fusion of pronuclei, extrusion of polar bodies, etc. to be considered. The temperature coefficients for the various phases of the first cleavage have been investigated by Ephrussi (1927) and for fertilization alone by Chase (1935). Different values are found by Ephrussi for various stages from fertilization to first division in Paracentrotus, but this would not prevent the overall values for the various cleavage stages (including the first) from being the same. Fry (1936) finds, on the other hand, the same temperature coefficient for the mitotic phases as for cleavage in Arbacia. The kinds of changes that occur in the later stages of differentiation (i.e. from gastrulation on) are manifestly different from the cleavage stages. The interpretation that may be made of the identity of temperature coefficients throughout develop- ment depends, of course, on the particular meaning assigned to temperature coefficients in general. But this involves some contro- versial questions that are beyond the scope of this paper. Our results show that we cannot within the range of temperatures used dissociate cleavage and organogenesis by means of temperature. It would be of interest to know whether other agents or extremes of temperature might effect such a separation so that the rates may be affected differently or that a process such as gastrulation may be made to proceed without the accompanying cell division. SUMMARY The effect of temperature on the rate of development of eggs of Ciona intestinalis, Dendraster excentricus, Urechis caupo, Strongylo- centrotus purpuratus and Lytechinus anamesus has been investigated. It is found that not only do the various early cleavage stages have the same temperature coefficient, but the values for the later stages of differentiation are the same as for cleavage. 80 MESS eS AP APSA IER In Ciona, the coefficient for hatching time is different from that foi cleavage. An interpretation of this based on the production and diffusion of a hatching enzyme is suggested. Counts of the number of cells of gastrule of Strongylocentrotus show no significant differences in embryos raised at different temperatures. The data are fairly well expressed by the assumption of a linear relation between the rate of development and temperature. The values of the coefficients for all the forms investigated, with the exception of Strongylocentrotus, are very closely alike, although the rates of development are quite different. Strongylocentrotus eggs, which develop more slowly than Dendraster or Lytechinus, give lower values of the temperature coefficient. LITERATURE CITED Attias, M., 1935. The effect of temperature on the development of Rana pipiens. Physiol. Zool., 8: 290. BELEHRADEK, J., 1935. Temperature and living matter. Protoplasma Mono- graphien, vol. 8. BELEHRADEK, J., AND J. MLADEK, 1934. Richesse en eau du bioplasme et la grandeur du coefficient thermique des oxydations. Protoplasma, 21: 335. BERRILL, N. J., 1929. Studies in tunicate development, Part 1. General physiology of development of simple ascidians. Phil. Trans. Roy. Soc. London, B, 218: Site BIERENS DE Haan, J. A., 1913. Uber homogene und heterogene Eseinnveuseaiuels- ungen bei Echiniden. Arch. Entw.-mech., 36: 474. BopineE, J. H., AND V.:THomepson, 1935. ienmperstne coefficients and viscosity. Jour. Cell and Comp. Physiol., 6: 255. Borine, A. M., 1909. On the effect of different temperatures on the size of the nuclei in the embryo of Ascaris megalocephala, with remarks on the size- relation of the nuclei in univalens and bivalens. Arch. Entw.-mech., 28: 118. Cuase, H. Y., 1935. The effect of temperature on the rate of the fertilization re- action in various marine ova. JBvzol. Bull., 69: 415. CroziER, W. J., 1924. On the possibility of identifying chemical processes in living matter. Proc. Nat. Acad. Sct., 10: 461. Crozier, W. J., 1924. On biological oxidations as function of temperature. Jour. Gen. Phystol., 7: 189. Crozier, W. J., 1926. The distribution of temperature characteristics for biological processes; critical increments for heart rates. Jour. Gen. Phystol., 9: 531. DriescuH, H., 1900. Studien iiber das Regulationsvermégen der Organismen 4. Die Verschmelzung der Individualitat bei Echinidenkeimen. Arch. Entw.- mech., 10: 411. Eprurusst, B., 1927. Sur les coefficients de température des différentes phases de la mitose des oeufs d’Oursin (Paracentrotus lividus, lk.) et de l’ Ascaris megalo- cephala. Protoplasma, 1: 105. Epurussi, B., 1933. Contribution a l’analyse des premiers stades du développement de l’oeuf. Action de la température. Arch. de Biol., 44: 1. ERDMANN, RH., 1908a. Experimentelle Untersuchung der Massenverhaltnisse von Plasma, Kern und Chromosomen in dem sich entwickelnden Seeigelei. Arch. f. Zellforsch., 2: 76. ERDMANN, Ru., 1908). Kern- und Protoplasmawachstum in ihren Beziehungen zueinander. rg. Anat. und Entw.-gesch., 18: 844. DEVELOPMENTAL RATE AND TEMPERATURE 81 Fry, H. J., 1936. Studies of the mitotic figure. V. The time schedule of mitotic changes in developing Arbacia eggs. Bzol. Bull., 70: 89. GODLEWSKI, JUN. E., 1908. Plasma und Kernsubstanz in der normalen und der durch aussere Faktoren veranderten Entwicklung der Echiniden. Arch. Entw.-mech., 26: 278. Gray, J., 1928. The growth of fish. III. The effect of temperature on the develop- ment of the eggs of Salmo fario. Brit. Jour. Exper. Biol., 6: 125. HeERTWIG, O., 1898. Ueber den Einfluss der Temperatur auf die Entwicklung von Rana fusca und Rana esculenta. Arch. f. Mtkrosk. Anat., 51: 319. Horstaptus, S., 1925. Temperaturanpassung bei den Eiern von Paracentrotus lividus Lk. Buologia Generalis, 1: 522. KOEHLER, O., 1912. Uber die Abhangigkeit der Kernplasmarelation von der Temperatur und vom Reifezustand der Eier. Arch. f. Zellforsch., 8: 272. Kroeu, A., 1914. On the influence of the temperature on the rate of embryonic development. Zetéschr. f. allg. Physiol., 16: 163. Marcus, H., 1906. Uber die wirkung der Temperatur auf die Furchung bei Seeigel- eiern. Arch. Entw.-mech., 22: 445. MorGan, T. H., 1901. The proportionate development of partial embryos. Arch. Entw.-mech., 13: 416. Morgan, T. H., 1903. The gastrulation of the partial embryos of Sphzrechinus. Arch. Entw.-mech., 16: 117. PETER, K., 1905. Der Grad der Beschleunigung tierischer Entwicklung durch erhéhte Temperatur. Arch. Entw.-mech., 20: 130. SKRABAL, A., 1916. Cited by Ephrussi, 1933. TYLER, A., 1933. On the energetics of differentiation. A comparison of the oxygen consumption of “‘half’’ and whole embryos of the sea-urchin. Publ. Staz. Zool. Napoli, 13: 155. TYLER, A., 1935. On the energetics of differentiation. II. A comparison of the rates of development of giant and of normal sea-urchin embryos. JBzol. Bull., 68: 451. ZAWADOWSKEY, M. M., AND C. M. Siporov, 1928. Die Abhangigkeit der Entwicklung der Eier von Ascaris megalocephala, Ascaris suilla und Toxascaris limbata von der Temperatur. Arch. Entw.-mech., 113: 536. ON THE ENERGETICS OF DIFFERENT TATION: LV. COMPARISON OF THE RATES OF OxYGEN CONSUMPTION AND OF DE- VELOPMENT AT DIFFERENT TEMPERATURES OF EGGS OF Some Marine ANIMALS ABBE RD MYCE (From the William G. Kerckhoff Laboratories of the Biological Sciences, Cahforma Institute of Technology, Pasadena, California) The results of these experiments show principally that, for the eggs of certain marine animals, the total oxygen consumed in reaching the same stage of development at different temperatures is the same. THEORETICAL PART It has been assumed (Tyler, 1933, 1935) that energy is required for the differentiation as well as the maintenance and growth processes of development and that this energy is supplied by reactions involving oxygen. The terms maintenance, growth and differentiation have been used in various ways by different investigators. By maintenance I mean the processes by which the ege or embryo is kept alive in a “resting condition.” An unfertilized egg illustrates this. Also a fertilized egg in which development has been reversibly stopped would illustrate maintenance. But it cannot be assumed that the maintenance energy requirement of a developing egg is equivalent to the metabolism of an unfertilized or of a fertilized egg in which development has been stopped, for the maintenance requirement of a fertilized egg is prob- ably different from that of an unfertilized egg; it very likely changes throughout development and under different conditions. In other words, it takes different amounts of energy to keep different kinds of cells alive under the same conditions, and the same kind of cell alive under different conditions. By growth we mean the processes by which food material is con- verted into protoplasm. This may not be the most convenient way in which to define this term in the case of the early development of a ma- rine egg. It is, however, more consistent with the generally employed definitions of growth. The difficulties arise in distinguishing between the food materials and the protoplasm of an egg. We recognize, how- 82 EMBRYONIC RESPIRATION AND TEMPERATURE 83 ever, certain constituents such as the yolk, fat, and some pigments as essentially non-protoplasmic materials. In the case of most small eggs these materials are distributed throughout the cytoplasm. In the case of large eggs, such as those of birds, a region which is preponderantly protoplasmic may be distinguished from one which is chiefly yolk. The embryo arises in the protoplasmic area and evidently contains a negligible amount of food material. Thus in studies on the growth of bird or fish embryos it is essentially the conversion of food materials into protoplasm that is being measured. In the case of unicellular or- ganisms growth studies likewise give a measure of the conversion of food materials of the medium into protoplasm of the cells. For small eggs it has been the custom either to consider growth as not occurring until actual increase in mass by absorption of materials from the out- side takes place, or to use the term growth as loosely synonymous with development. It is evident, however, that there is growth in small eggs during the period before actual increase in mass takes place. The dif- ference is that the nutritive material is intracellular. Distinction should also be made between growth and what we may call storage. The increase in size of the young oocyte as it is trans- formed into the ripe egg may be considered as illustrating storage since it is principally food materials that accumulate. It may seem that this manner of regarding growth leads to a paradox; we have cases of no increase in mass in which we do not consider growth to be occurring. But the ordinary definition of growth as increase in mass is meaning- less unless the system is specified. Our definition considers the pro- toplasm to be the growing system. We define growth in this manner in order to be able when necessary to utilize the general results of growth studies on various kinds of living things. Differentiation may be regarded as the changes in form that occur during development, the so-called morphogenetic and histogenetic changes. Strictly speaking, this is one aspect of what may be called differentiation. The production of the typical chemical substances characteristic of the various tissues must also be included. However, this also fits into our definition of growth. There is no point at present in attempting to classify it specifically with growth or with the changes in form. It rests, of course, with the actual experimental evidence to decide what the exact relationships are. The relation of cleavage to differentiation has been considered in the previous paper. In the early experiments of Lillie (1906), a sort of differentiation without cleavage has been described. But it is ques- ‘tionable whether in this case typical differentiation processes have taken place. This is not merely a matter of the use of words, for if the kinds 84 ALBERT TYLER of changes described by Lillie are comparable with those undergone in normal development, such embryos would furnish important ma- terial for the analysis of the form changes and histochemical changes. It has, of course, been long known that cleavage and growth may go on without differentiation (e.g., anidian chick embryos, Dareste, 1882, Grodzinski, 1933). But that should not lead us to expect the reverse to be true. In distinguishing between maintenance, growth and differentiation we do not mean to imply that these are independent processes in de- velopment. They are undoubtedly interrelated, but we can effect a separation conceptually and to some extent experimentally. An excel- lent review of normally occurring and experimentally produced cases of dissociation of the various embryological processes is given by Need- ham (1933), and justification for the general point of view presented here is contained therein. Concerning the effect of temperature on the rate of respiration and on the rate of development, the different processes involved should be considered. If as the temperature is lowered the requirements for these processes are decreased in the same proportion, we should expect the decrease in rate of development to be the same as the decrease in rate of oxygen consumption. This would not tell us much about maintenance, growth and differentiation, but it would mean that the efficiency with which the available energy is utilized remains the same at different temperatures. If, however, the requirements for the various processes changed in different ways with change in temperature we should not expect such a simple relation. Unfortunately we cannot state a priori in what manner the various processes should be affected by temperature. But it is possible to get at this experimentally. Thus it should be possible to learn the manner in which maintenance is affected by studying the effect of temperature on the metabolism and length of life of unfer- tilized eggs. The duration of the fertilizable condition must also be investigated, since strictly speaking we must regard maintenance as not merely the process of keeping the cells alive but also of keeping them in a condition in which they are capable of functioning normally. Such an investigation involves certain difficulties which remain to be overcome before the results can be presented. This work concerns only the effect of temperature on the rate of development and rate of oxygen consumption. Several such investi- gations have been made in the past. The results of Loeb and Wasteneys (1911) are the most often cited and generally accepted. They found the temperature coefficient of the rate of development to be larger. than EMBRYONIC RESPIRATION AND TEMPERATURE 85 that of the rate of oxygen consumption. At low temperatures, then, the total oxygen consumed in reaching a given stage of development would be greater than at higher temperatures. Loeb and Wasteneys worked with sea-urchin eggs and considered only the early cleavage stages. But as we have seen in the previous papers, the temperature coefficients for the cleavage and the later stages are the same. They determined their temperature coefficients for respiration by running a batch of eggs for some time at one temperature and then changing the temperature on the same batch of eggs. Thus, the eggs are not in the same stage of development and, as Ephrussi (1933, p. 115) points out, the results are not strictly comparable, since as development proceeds the rate of oxygen consumption rises. Ephrussi, working also on sea-urchin eggs, gets different results. Using the time from the two-cell stage to hatching as the criterion for rate of development, he finds that the total oxygen consumpticn is the same at different temperatures, in the interval from 10° to 23°. Above 23° the total oxygen consumption is greater than at the lower tem- peratures, but Ephrussi states that it is not certain that development would proceed very far at these higher temperatures. When, however, he takes the interval from fertilization to the two-cell stage, he finds the total oxygen consumption to be the same only in the interval between 16° and 23°. Above 23° the total oxygen consumption is higher and below 16° it is lower than at the other temperatures. He supposes that these differences only occur in the early stages and are obscured in the longer time interval and increasingly greater oxygen consumption up to hatching. Measurements of the rate of cleavage and rate of oxygen consump- tion at different temperatures have been made by Fauré-Fremiet (1924) on eggs of Sabellaria. From his results it appears that the total oxygen consumption in reaching the same stage of development is different at different temperatures. The same sort of experiment has been performed on eggs of the frog by Barthélemy and Bonnet (1926), using the bomb calorimeter. They found that between 8° and 21° there was the same expenditure of energy in reaching the stage at which the external gills disappear. In insects, Krogh (1914) and more recently Crescitelli (1935) have determined the rate of pupal development and of respiration at different temperatures. They obtained conflicting results. Krogh used pupz of the beetle, Tenebrio, and found that in the interval of 20° to 33° the total CO, produced during pupal life was the same at different temperatures. Crescitelli used pupz of the moth, Galleria, and found in the interval of 20° to 40° the total O, consumed or CO, produced to 86 ALBERT a YLER be different at different temperatures. He gets a minimum total gas exchange at 30° C. with regular increases to each side of that tempera- ture. Dobzhansky and Poulson (1935) present some data on pupe of Drosophila pseudoobscura from which it appears that the total oxygen consumption at 25° and 14° is of the same order of magnitude. Simi- lar results are obtained by Brown (1928) with Ascaris eggs and by McCoy (1930) with hookworm eggs. If in the case of the sea-urchin egg the total oxygen comctined at the lower temperatures were much greater than at the higher tempera- tures, as the results of Loeb and Wasteneys show, we should expect a rather interesting result. The sea-urchin embryo dies within a limited time after entering the pluteus stage unless food is supplied. If, then, we allow sea-urchin eggs to develop at different temperatures on their own initial food supply, the retardation in time of death will not be the same as the retardation in rate of development but will be the same as the retardation in rate of oxygen consumption, since that measures the rate at which the original food supply is being depleted. For example, if we take Loeb and Wasteneys’ figures of 7.3 and 2.0 for the coefficients of rate of cleavage and rate of oxygen consumption for 7°-17°, and assuming (as they do) that the values for cleavage hold throughout development, then development will be slowed up 7.3 times at 7° as compared with 17° but the length of life will be prolonged only 2.0 times. Thus, even allowing for the short time in which the pluteus remains alive without undergoing any developmental changes, we should not expect the embryos raised at 7° to have developed as far at the time of death as those at 17°. In fact, it appears from the figures that the embryos raised at 7° would be expected to die in the late prism stage. Thus below a certain temperature embryos may develop normally but would die before being able to feed, since in the prism stage the mouth is not open. We have investigated this particular point and have found no such precocious dying at the lower temperatures. As is evident in the previous paper, embryos raised at the lower temperatures develop as far as those raised at the higher temperatures. While this sort of evidence shows that enormous differences in the temperature coefficients of development and respiration do not occur with our material, it might well be that smaller differences exist. Our experimental evidence, how- ever, gives no significant differences. EXPERIMENTAL PART Material and Methods The eggs of four marine animals were used; namely, the sand-dollar, Dendraster excentricus, the sea-urchin, Strongylocentrotus purpuratus, the ascidian, Ciona intestinalis, and the gephyrean worm, Urechis caupo. EMBRYONIC RESPIRATION AND TEMPERATURE 87 The method of determining the rate of development at different tem- peratures is given in the previous paper. One or more determinations of the temperature coefficient of the rate of development were made on each set of eggs used for the respiration experiments. For this purpose eges from the same fertilized lot that were used for oxygen consump- tion measurements were placed in the temperature baths at the same time as the manometers. These egg suspensions were of about the same concentration as the suspensions in the manometer vessels and were subjected to the same shaking. The eggs in the manometer vessels, used for the respiration measurements, were also examined at the end of each run to check their condition as well as the stage of development. Fic. 1. Type of manometer vessel used in these experiments. The respiration was measured by means of the Warburg manometer. The vessels used are of the type illustrated in Fig. 1. They are quite small, the calibration volumes ranging from 2.8 to 3.0 cc., which makes them particularly suitable in cases in which only small quantities of material can be obtained. The manometer capillaries were also small (about 0.6 mm. diameter) and the T connection for the arm that hangs into the bath was made much smaller than usual, thus reducing the gas volume outside of the bath. These vessels have been checked against the standard conical type. The amount of fluid with which the vessels can be filled depends upon the rate of shaking. As much as 1.5 cc. of fluid can be used with moderate shaking (about twenty round-trip shakes per minute). In most of the experiments, however, the vessels were filled with 1.1 cc. of fluid, which gives vessel constants for oxygen at 20° ranging around 0.18. As has often been pointed out (see especially Whitaker, 1933), the rate of shaking is of considerable importance when measurements on marine eggs are being made, for one is con- cerned not only with insuring adequate diffusion of oxygen but also with avoidance of injury to the material. We have determined the optimum shaking for our material and find slow rates of 40 to 70 round-trip shakes per minute at 3.5 cm. amplitude to be satisfactory for dilute egg suspensions (up to 0.5 mg. of nitrogen in 1.1 cc., which for Strongylocentrotus corresponds to a volume concentration of 1: 100). For more concentrated suspensions stronger shaking is necessary, not only because of the greater oxygen consumption, but also because at 88 AEBERT @ VEER the same rate of shaking the contents are less disturbed when a con- centrated egg suspension is employed. For suspensions giving 2.5 mgm. nitrogen in 1.1 cc. the shaking was 70 to 100 round-trips at 4.5 cm. amplitude. The quantity of eggs employed was determined after each run by measuring the nitrogen content by a micro-Kjeldahl method. Since this determination could be made to 0.01 mgm. N, the error involved when the quantity of eggs corresponds to 1.0 mgm. N is less than 1 per cent. In some instances the actual number of eggs used was determined by counting. With the higher ratio of fluid te gas volumes obtained in our vessels the solubility coefficient of O, (Bunsen’s coefficient, a) is of slightly more importance than in the usual type of vessel. Values calculated from the formula of Fox (1909) range from 0.024 at 25° C. to 0.033 at 7.5° C. This term therefore contributes only about 2 per cent to the vessel constant and an error of as much as 10 per cent in the values of a would mean an error of 0.2 per cent in the vessel constant. It is evident then, especially since we are interested in the relative values of oxygen consumption, that errors from this source are negligible. The manometer readings were made to the nearest 0.5 mm. To de- termine the pressure changes to which the oxygen consumption data of the following tables correspond it is simply necessary to multiply by the mgm. N of the egg suspension in the vessel and divide by the vessel constant. For example, in Table I, Xo,/mgm. N for the first 20° vessel is 6.7 cu. mm. O, in one hour. Multiplying by the mgm. N and dividing by the vessel constant gives 11.5 mm. as the pressure change. Since the reading is made to 0.5 mm. the error is less than 2.5 per cent. This, however, is one of the smallest values presented in the table. Ii we take the figure 128.8 cu. mm. O, from Table If for 5 hours respiration at 20°, this corresponds to a pressure change of 357.5 mm. and the reading error is less than 0.1 per cent. The measurements were practically all made in duplicate. For the same batch of eggs two vessels were run simultaneously at one tempera- ture and two at a different temperature. Different batches of eggs may vary somewhat in their absolute rates of respiration as well as in their rates of development. To avoid this source of error it is preferable to compare only the results on the same batch of eggs. Strongylocentrotus In Table I the results of three experiments with eggs of Strongylo- centrotus in the early cleavage stages are given. The first column of the table gives the temperature at which the measurements are made; EMBRYONIC RESPIRATION AND TEMPERATURE 89 the second column shows the nitrogen content of the eggs used in each vessel; the third, the age of the eggs as the time from fertilization at 20° C. of the first reading of the manometer; the fourth and fifth give the length of time (from the first reading) during which the oxygen consumption is measured and the corresponding oxygen consumption in cu. mm. O, per milligram of egg nitrogen; the sixth and seventh col- umns again give the respiration time (also from the first reading but continued over a longer time) and the corresponding oxygen consump- tion. In the first experiment the Q, is 1.36. This was determined from the time between the first and second cleavage for eggs of the same fertilized lot that were placed in the 20° and 15° baths at the same time TABLE I Strongylocentrotus; oxygen consumption at different temperatures; early cleavage stages. Temp mem. Ni i cee Respiration a Respiration oe NE minutes hours hours 20 0.31 45 1.0 6.7 5.0 40.5 20 0.25 45 1.0 6.8 5.0 41.0 15 0.40 45 1.36 6.2 6.8 39.1 15 0.36 45 1.36 6.6 6.8 38.9 20 1.71 35 1.0 6.4 3.5 24.4 20 1.74 35 1.0 6.3 3.5 24.1 10 1.75 35 2.3 6.5 8.1 23.8 10 1.78 35 2.3 6.1 8.1 23.4 20 2.38 50 1.0 6.4 2.5 17.8 20 2.40 50 1.0 6.3 225) ES 7.5 2.39 50 3.2 6.2 8.0 17.6 7.5 2.38 50 Si 6.3 8.0 17.7 as the manometers and were also at about the same concentration as the eggs in the manometers. The one-hour interval at 20° therefore cor- responds to 1.36 hours at 15° and the five-hour interval at 20° to 6.8 hours at 15°; that is, in those time intervals the same stage of develop- ment would be attained at the two temperatures. Comparing then the oxygen consumption for the corresponding time intervals at the two temperatures we see that for both the shorter and longer runs the 20° figures are greater. The differences are, however, within the limits of error. It is evident, too, that for 5 and 6.8 hours the differences are much smaller than for 1 and 1.36 hours, which is to be expected if the Q, obtained for the rate of development is reliable, since the errors in the oxygen consumption measurements decrease in the longer run. 90 ALBERT TYLER In the second and third experiments larger amounts of material were used and the differences in total oxygen consumption are much smaller. The Q,, of the rate of cleavage in the second experiment is 2.3 and the Q,.., in the third experiment is 3.2, both determined, as before, from the time between the first and second cleavage. It may readily be seen that the differences in total oxygen consumption at the two temperatures are no greater than the differences between the duplicates. The two values for one temperature may even lie between the two for the other temperature, as in the third experiment (20° and 7.5° for 2.5 and 8.0 hours). It may be concluded then that in the early cleavage stages and be- tween 20° and 7.5°, the total oxygen consumed in reaching the same stage of development is the same at different temperatures. We may next see whether this holds for the later stages. In the preceding paper it was shown that the later stages of development were affected in the TABLE II Strongylocentrotus; oxygen consumption at different temperatures; gastrula stage. Tes. “asain, INT Pune) guar Respuation — Respeaon a “&. hours hours hours 20 0.51 25 1.0 23.6 5.0 128.8 20 0.37 25 1.0 24.0 5.0 129.1 15 0.42 25 1.31 20.9 6.55 119.5 15 0.34 25 1.31 21.6 6.55 120.7 same manner by temperature as were the cleavage stages. It is pos- sible, however, that the rate of oxygen consumption might be differently affected by temperature in the early cleavage stages and in the later stages. One experiment is presented in Table II and another in Table III showing that this is not the case at least for the temperatures 20° and 15° C. Similar experiments are presented in connection with the other forms investigated. In Table II the differences between the average total oxygen con- sumption at the two temperatures appear to be fairly large compared with the small differences between the duplicates. These differences are, however, most likely due to the value of O, being too small. Only one determination of QO, was made for this batch of eggs and it was taken for the time from fertilization to gastrulation, which is less accu- rately determined than cleavage time. In Table III the differences in oxygen consumption at 20° and 15° are quite small. The figures in EMBRYONIC RESPIRATION AND TEMPERATURE 91 this table ‘are arranged differently from those in the other tables and should be compared horizontally. The time interval is given only for the 20° runs. The time interval at 15° is 1.37 times that at 20°. This figure is the OQ, as determined from the time from the first to second cleavage and from first cleavage to beginning of gastrulation at the two temperatures. Table III also illustrates the rise in rate of oxygen consumption that occurs as development proceeds. Thus the rate at 21 to 24 hours is three times that at 0 to 3 hours. Of course, even in the first three hours the rate is rising, as may be seen in Table I. It is because of the in- creasing rate of oxygen consumption that we do not compare the rates at different temperatures, but rather the total oxygen consumed by the Taste III Strongylocentrotus; oxygen consumption at different temperatures; cleavage to gastrulation. PRMD esha) chevesusretere 20° 20° 5? 15° TEMS INGS soci Sees 0.53 0.72 0.61 0.86 Time of first reading. 45 min. 1834 hrs. 45 min. 1834 hrs. Time interval XOo X02 X02 X02 at 20°C. mgm. N mgm. N mgm. N mgm. N hours OR SR Si caa ch: 25.4 24.7 SOMO cae eat 34.1 33.8 (GAO A eae aad 41.8 42.9 1S eae 53.7 55.0 LH ee 65.0 67.2 68.1 68.8 DMD A ieee. ty Poll 74.8 74.9 ee eggs in undergoing the same developmental changes at the different temperatures. A technical point is also illustrated in Table III. It is possible that over long periods of time conditions change in the vessels so that.an abnormal respiratory rate is being measured towards the end of a long run. Most of the experiments reported here are of relatively short duration and would not be expected to show such effects. The experi- ment of Table III shows, however, that reliable data may be obtained throughout a run as long as 24 hours. In this experiment two vessels, one at 20° and one at 15°, were started at 45 minutes after fertilization and run for 24 hours (32.9 hours at 15°). The main batch of eggs was divided between two flasks, one kept at 20° and the other at 15° C. At 18 hours after fertilization (24.7 hours at 15°) embryos from these flasks were washed by gentle centrifugation, and two more respiration vessels, one at 20° and one at 15°, were started. It may be seen from 92 ALBERT TYLER AB EEL, Dendraster; oxygen consumption at different temperatures; early cleavage stages. hen seayesaal, BY eS Respiration aoe Bespiration — PCy minutes hours hours DD, 1.28 50 1.0 9.7 3.0 33.5 Dy, 0.82 50 1.0 10.7 3.0 34.6 12 0.92 50 2.8 11.0 8.4 35.9 12 0.82 50 2.8 11.7 8.4 36.6 25 0.69 70 1.0 18.3 3.0 59.1 25 0.53 70 1.0 19.5 3.0 65.8 15 0.62 70 Dols 14.8 6.6 54.9 15 0.78 70 2.2 14.1 6.6 53.0 the table that there are no significant differences in the oxygen consump- tion for the 18 to 21 and 21 to 24-hour intervals between the two vessels that had been run continuously up to that time and the two that had been freshly started. Dendraster It is more difficult with eggs of this form to obtain large quantities in which one hundred per cent are fertilized. In two experiments, how- TABLE V Dendraster; oxygen consumption at different temperatures; blastula and gastrula stages. Temp. ment NI Des Re oon oe Respiration oo “Co hours hours hours 20 0.89 13.0 1.0 34.4 3.0 109.1 20 0.94 13.0 1.0 34.2 3.0 108.1 15 0.76 13.0 1.65 33.5 4.95 105.2 15 0.78 13.0 1.65 34.3 4.95 107.5 15 0.87 12.0 1.0 16.4 3.0 51.0 15. 0.98 12.0 1.0 15.9 3.0 49.5 10 0.95 12.0 1.95 16.8 5.85 49.6 10 0.75 12.0 1.95 17.4 5.85 52.0 DD, 0.21 8.5 1.0 24.4 3.0 75.8 22 0.19 8.5 1.0 D5 3.0 77.1 12 0.21 8.5 2.75 23.2 8.25 77.6 12 0.20 8.5 2.75 23.7 8.25 78.7 22 0.66 10.0 1.0 27.3 3.0 86.4 22 0.36 10.0 1.0 28.1 3.0 87.0 12 0.55 10.0 2.8 29.4 8.4 92.7 12 0.36 10.0 2.8 31.6 8.4 93.8 bt PRI EMBRYONIC RESPIRATION AND TEMPERATURE 93 ever, fairly large quantities of eggs were obtained in which the fer- tilization was 95 to 97 per cent. The oxygen consumption data are pre- sented in Table IV, one experiment being at 22° and 12°, the other at 25° and 15°. The Q,,’s are 2.8 and 2.2 respectively, determined from the time between first and second cleavage. In the first experiment the oxygen consumption is evidently the same at the two temperatures. In the second experiment the values are higher at 25° than at 15°. The differences appear to be significant, but more data would be needed at these temperatures to establish this point. It should be noted that 25° is just about the upper limit at which normal development is obtained. In Table V the results of four experiments on blastule and gastrule are presented, one at 20° and 15°, one at 15° and 10° and two at 22° and 12°. Since only top swimmers are taken, it is not necessary to have 100 per cent fertilization at the start. The time of the first read- ing (column 3 of the table) represents the age of the embryos at 20° and the corresponding stage of development may be obtained from the data in the preceding paper. The temperature coefficients of the rate of development were determined in these cases from the time between fertilization and the beginning of gastrulation, and the respiration times are, as usual, taken from the values of the temperature coefficients. The oxygen consumption data (Xo./mgm. N) again show the amount consumed during the same developmental period at the two temperatutes of each experiment. As was the case with Strongylocentrotus, we find here too no significant differences in the total oxygen consumed at the different temperatures. An experiment at 20° and 15° covering the entire period from fer- tilization to gastrulation is presented in Table VI. In this experiment AAsicis, WAL Dendraster; oxygen consumption at different temperatures; from one-cell stage to end of gastrulation. . Dae 1000 X Ceara 1000 X Temp. Ne, aif eae ee a Reoeieeren xo =. Respiragon ool ses AO minutes hours hours 20 1653 40 6.0 Do 15.0 7.9 20 2791 40 6.0 2.0 15.0 7.5 15 2581 40 9.6 2.0 24.0 7.4. relatively few eggs were used in each vessel and the number was deter- mined at the end of the run by counting them all. There were very few immature eggs in this lot, fertilization being about 98 per cent. In one of the 15° vessels the sea water slopped over into the alkali well. But 94 ALBERT YEER Taste VII Urechis; oxygen consumption at different temperatures; early cleavage stages. Time of first | Respiration X02 Respiration X02 Temp. mgm. N reading time mem. N time mgm. N SiG minutes hours hours 22 1.19 60 1.0 4.4 4.0 18.6 22 1.40 60 1.0 4.9 4.0 Die? 12 1255 60 Boll 5.9 10.8 23.0 12 2.26 60 Dail 5) 10.8 21.1 20 1.03 60 1.0 4.0 4.0 17.3 20 1.36 60 1.0 3.9 4.0 17.3 15 1.66 60 1.62 5.0 6.5 18.2 even without the duplicate run at 15° it may readily be seen that the total oxygen consumed at the two temperatures is the same. TasB_LeE VIII Ciona; oxygen consumption at different temperatures; cleavage and later stages. Remy naa. AY Fone cupst Respiation hs Respiretion oe CN OF minutes hours hours 20 0.22 60 3.0 28.2 9.0 119.5 20 0.25 60 3.0 29.3 9.0 120.1 15 0.25 60 5.0 28.9 15.0 116.3 15 0.33 60 5.0 Dies 15.0 113.0 Urechis Two experiments, one at 22° and 12° and one at 20° and 15°, were performed on eggs of Urechis. In both 100 per cent fertilization was obtained, as is usual with eggs of this animal. The results are pre- sented in Table VII. The Q,, for the rate of cleavage in the 22°-12° experiment is 2.7 and the Q, in the 20°-15° experiment is 1.62. In both experiments the total oxygen consumed at the lower temperature is somewhat higher than at the higher temperature. However, considering the variation between the duplicate vessels, and also the error associated with the temperature coefficient determinations for ‘development (see preceding paper), we cannot regard these differences as significant. Ciona Two experiments at 20° and 15° were run with eggs of Ciona. In the first (Table VIII) a relatively large number of eggs were employed EMBRYONIC RESPIRATION AND TEMPERATURE 95 and the amount determined as usual from the Kjeldahl nitrogen. In the second (Table IX), fewer eggs were used and the amount deter- mined by counting. shAaie) 1G Ciona; oxygen consumption at different temperatures; cleavage and unhatched tadpole stages. Tavis INawoneree Time of first | Respiration 1000 XO Respiration 1000 Xo reading time No. of eggs time No. of eggs 2G: minutes hours hours 20 1417 60 3.0 1.9 12.0 IBall 20 1273 60 3.0 Mell 12.0 13.3 15 1052 60 5.0 Mall 20.0 12.9 15 1042 60 50 « 2.0 20.0 12.8 There are some technical difficulties in obtaining large quantities of Ciona eggs for respiration measurements. Generally several ani- mals must be used and the eggs must be pooled in order to have com- parable lots in each vessel. In removing Ciona eggs it is very difficult to be certain that no sperm from the same animal is obtained at the same time. This introduces no serious difficulty when eggs from one indi- vidual are used since Ciona is a self-sterile hermaphrodite. However, to control the time of fertilization when several animals are used it is necessary to mix all the eggs simultaneously and inseminate at the same time. The material of the first experiment was prepared in this man- ner from four individuals; that for the second experiment came from one individual. In both experiments fertilization was 100 per cent, which is the rule for Ciona eggs removed from the oviduct. The Q,’s as determined from the time from first to second cleavage were 1.67 in both cases. The duplicate runs in these experiments checked very nicely, the biggest difference being 3 per cent. Since the error associated with the tem- perature coefficients is also of this order of magnitude, it may readily be seen that the differences in total oxygen consumption for the longer runs at the different temperatures are well within the limits of error. For the shorter readings in both experiments the duplicates at the two temperatures overlapped. We may conclude then that in Czona, too, the total oxygen consumed in reaching the same stage of development is the same at different temperatures. 96 ALBERT TYLER Discussion In these experiments it is seen that, within the range employed, the effect of temperature on the rate of development parallels its effect on the rate of oxygen consumption. We know, however, that at tempera- tures at which development is inhibited respiration goes on at a meas- urable rate. We might therefore expect striking differences as such temperatures are approached. But certain complications enter when we consider inhibiting temperatures. As the temperature is lowered we first reach a point at which cytoplasmic division is interfered with whereas nuclear division proceeds. This point is reached rather abruptly and not after an infinitesimally slow rate of development is attained, as is evident from the fact that in some eggs of a given lot cytoplasmic division is inhibited whereas in others it goes on at an easily measurable rate. For example, in a batch of sand-dollar eggs raised at 7.5° C., about 80 per cent failed to divide while the rest under- went the first cleavage at about 270 minutes after fertilization. The nuclear activity that proceeds when cytoplasmic division has been in- hibited must be regarded as a sort of developmental activity. Should we, however, expect the rate of oxygen consumption to drop to zero at temperatures at which all developmental activity has ceased? This is not necessarily the case if there is any significance to the concept of maintenance. If the inhibition by low temperature is reversible then the blocked eggs may be expected to show a basal metabolism characteristic of that temperature. The magnitude need not be the same as that of the unfertilized egg at the same temperature, since, as was pointed out above, the maintenance metabolism of different kinds of cells probably differs. The recent investigations of Whitaker (1931, 1933) bear on this point. Whitaker found that unfertilized eggs of Chetopterus and Cumingia respire at a higher rate than do the fer- tilized eggs. If then it were assumed that at any temperature the maintenance component of the developing egg is equivalent to the me- tabolism of the unfertilized egg, negative values would be obtained for the requirements of the developmental processes. A similar situation arises in one of the interesting series of experiments of Runnstrom (1935). He finds that the respiration of unfertilized sea-urchin eggs is much more strongly increased on addition of pyocyanine than that of the developing eggs. The eggs may be fertilized in the pyocyanine solution, but the increase in the rate of respiration is very small com- pared with that in normal sea water. Addition of HCN depresses the rate of the fertilized eggs in pyocyanine but, strangely enough, in- EMBRYONIC RESPIRATION AND TEMPERATURE 97 creases that of the unfertilized eggs, so that at certain concentrations the latter show a higher respiratory rate. The identity of the temperature coefficients for the rate of develop- ment and the rate of oxygen consumption may be interpreted to mean that the maintenance, growth and differentiation requirements are similarly affected by temperature. It might also mean, although it seems more unlikely, that the proportionate requirements for mainte- nance and growth change but their sum remains the same fraction of the total at the different temperatures. I{ maintenance is assumed to be the same sort of process, although different in magnitude, in de- veloping eggs and in unfertilized eggs, then, according to the former interpretation, we should expect the respiration of the unfertilized egg to have the same temperature coefficient as that of the fertilized egg. Measurements have been made of the temperature coefficient of the respiration of unfertilized eggs of the sea-urchin by Rubenstein and Gerard (1934). They find it to be different from that of fertilized eggs. The values for the coefficients of the unfertilized egg are higher than for the fertilized eggs, and the big rise in respiratory rate that occurs upon fertilization at ordinary temperatures disappears at higher temperatures. It would appear then that maintenance in unfertilized and in developing eggs is not the same kind of process. There are, of course, a number of assumptions in this argument. The assumption that the same kinds of processes should show identical temperature co- efficients is probably the safest of these. But it is also assumed that the respiration of an unfertilized egg is a measure of its maintenance requirement. This is the way in which basal metabolism studies on adults are regarded. But it is quite conceivable that very little, if any, energy is required for maintenance, and that most of that supplied by the oxidations is quite unnecessary. The other assumption is made because, in the absence of any real information as to the nature of maintenance, it appears to be the simplest. For example, the oxygen consumption might be regarded as representing some sort of destruc- tive process in the unfertilized egg (which, in a certain sense, it is, since the metabolites are being destroyed), but calling it destruction is merely a change in terminology. The real point is whether or not the same kind of process occurs in developing eggs. One conclusion that has been drawn from the dissimilarity of the temperature coefficients of development and respiration (shown chiefly by Loeb and Wasteneys) is that Child’s theory of metabolic gradients cannot hold (Crozier, 1926). But regardless of the particular merits of that theory, it is evident that the experimental facts are inadequate 98 Jblaleiecdy A) \eILID Is to justify that objection. The recent experiments of Brachet (1934) appear, on the other hand, to favor the view that the most “active” regions have the highest metabolic rate. He found the respiratory rate of the dorsal lip of the blastopore in the frog to be higher than that of the rest of the egg, and Child (1929) regards this region as a metabolic center. However, to conclude that the inductive properties of the dorsal lip of the blastopore are due to the high metabolism is certainly premature. As Brachet himself points out, injury to an egg or tissue raises the rate of oxygen consumption, but it seems quite certain that simple pricking of the egg would not give induction. We may await with interest experiments on amphibian gastrule in which the effect of local application of various reversible dyestuffs affecting the rate of respiration is tested. SUMMARY The effect of temperature on the oxygen consumption of the eggs of four marine invertebrates, Strongylocentrotus purpuratus, Dendraster excentricus, Urechis caupo and Ciona intestinalis, was investigated. With Strongylocentrotus eggs the measurements were made in the early cleavage and gastrula stages as well as continuously up to the gastrula stage and at the temperatures 20°, 15°, 10°, and 7.5° C. With Dendraster eggs the measurements were made in the early cleavage, blastula and gastrula stages as well as continuously throughout those stages and at the temperatures 25°, 22°, 20°, 15° and 12° C. With Urechis eggs the measurements were made in the early cleavage stages and at the temperatures 22°, 20°, 15,° and 12° C. With Ciona eggs the measurements were made in the early cleavage stages and up to hatching and at the temperatures 20° and 15° C. In all cases, with the exception of one run with Dendraster, the total oxygen consumed during the same developmental period at the different temperatures is the same within the limits of error of the measurements. The exceptional Dendraster experiment showed a somewhat higher total respiration at 25° as compared with that at 15° C. Omitting this case, which needs to be further investigated, it may be concluded that at least in the range of temperatures investigated no optimum exists at which development is accomplished with a minimum oxygen consump- tion. I am indebted to Mr. W. D. Humason for assistance in collecting some of the data presented here and to Professor T. H. Morgan for his valuable suggestions. EMBRYONIC RESPIRATION AND TEMPERATURE pe LITERATURE CITED BartHeLEMY, H., AnD R. Bonnet, 1926. L’énergie de croissance. IX. Influence de la température sur l’utilisation de l’énergie au cours du développement de l’oeuf de Grenouille rousse (Rana fusca). Bull. Soc. Chim. biol., 8: 1071. BracHet, J., 1934. Etude de l’oeuf de Grenouille (Rana fusca) au cours du développement. 3. Metabolisme respiratoire et “centre organisateur” de la gastrula. Arch. de Biol., 46: 25. Brown, H. W., 1928. A quantitative study of the influence of oxygen and tem- perature on the embryonic development of the eggs of the pig ascarid (Ascaris suum Goeze). Jour. Parasitol., 14: 141. Cutty, C. M., 1929. The physiological gradients. Protoplasma, 5: 447. Cuitp, C. M., 1929. Lateral grafts and incisions as organizers in the hydroid, i Corymorpha. Physiol. Zoél., 2: 342. CRESCITELLI, F., 1935. The respiratory metabolism of Galleria melonella (Bee Moth) during pupal development at different constant temperatures. Jour. Cell. Comp. Physiol., 6: 351. Crozier, W. J., 1924. On biological oxidations as function of temperature. Jour. Gen. Physiol., 7: 189. Crozier, W. J., 1926. The distribution of temperature characteristics for biolog- ical processes; critical increments for heart rates. Jour. Gen. Physiol., 9: 531. DaresTE, C., 1882. Recherches sur la production des monstres, dans l’oeuf de la ~ poule, par l’effet de l’incubation tardive. Compt. Rend. Acad. Sci., 95: 254. DoszHANSkKyY, TH., AND D. F. Poutson, 1935. Oxygen consumption of Droso- phila pupz. II. Drosophila pseudodbscura. Zeitschr. vergl. Physiol., 22: 473. Epurussi, B., 1933. Contribution a l’analyse des premiers stades du développe- ment de l’oeuf. Action de la température. Arch. de Biol., 44: 1. Fauré-Fremiet, E., 1924. L’oeuf de Sabellaria alveolata L. Arch. d’ Anat. Micr., 20: 211. Fox, C. J. J., 1909. On the coefficients of absorption of nitrogen and oxygen in : distilled water and sea water, and of atmospheric carbonic acid in sea water. Trans. Faraday Soc., 5: 68. ; GropzInsk1, Z., 1933. Uber die Entwicklung von wunterktthlten Huhnereiern. Arch. Entw.-mech., 129: 502. Krocu, A., 1914. On the rate of development and CO, production of chrysalides of Tenebrio molitor at different temperatures. Zeitschr. allg. Physiol., 16: 178. Lituig, F. R., 1906. Observations and experiments concerning the elementary phe- nomena of embryonic development in Chetopterus. Jour. Exper. Zodll., Semlos: Logs, J.. AND H. WasTENEYS, 1911. Sind die Oxydationsvorgange die unabhangige Variable in den Lebenserscheinungen? Biochem. Zeitschr., 36: 345. McCoy, O. R., 1930. Cited in Needham’s Chemical Embryology. NEEDHAM, J., 1933. On the dissociability of the fundamental processes in onto- genesis. Biol. Rev., 8: 180. RUBENSTEIN, B. B., AND R. W. Gerarp, 1934. Fertilization and the temperature coefficients of oxygen consumption in eggs of Arbacia punctulata. Jour. Gen. Physiol., 17: 677. RUNNsTROM, J., 1935. On the influence of pyocyanine on the respiration of the sea urchin egg. Biol. Bull., 68: 327. Tyrer, A., 1933. On the energetics of differentiation. A comparison of the oxy- gen consumption of “half” and whole embryos of the sea-urchin. Publ. Staz. Zool. Napoli, 13: 155. 100 ALBERT TYLER Tyier, A., 1935. On the energetics of differentiation. II. A comparison of the rates of development of giant and of normal sea-urchin embryos. Buol. Bull., 68: 451. Tyrer, A., 1936. On the energetics of differentiation. III. Comparison of the temperature coefficients for cleavage and later stages in the development of the eggs of some marine animals. Biol. Bull., 71: 59. Warsure, O., 1926. Stoffwechsel der Tumoren. Berlin. Wuitaker, D. M., 1931. On the rate of oxygen consumption by fertilized and unfertilized eggs. II. Cumingia tellinoides. Jour. Gen. Physiol., 15: 183. WHitTakeErR, D. M., 1933. On the rate of oxygen consumption by fertilized and unfertilized eggs. IV. Chetopterus, and Arbacia punctulata. Jour. Gen. Physiol., 16: 475. Wuauitaker, D. M., 1933. On the rate of oxygen consumption by fertilized and un- fertilized eggs. V. Comparisons and interpretation. Jour. Gen. Physiols, 16: 497. PARTHENOGENETIC MEROGONY OR CLEAVAGE WITH- OUT INUICIEIGI TUN AURIBVAICIVAL JETUUN CIOL yeVIb at ETHEL BROWNE HARVEY (From the Marine Biological Laboratory, Woods Hole, and the Biological Laboratory, Princeton University) Parts of unfertilized Arbacia eggs from which the nucleus has been removed, may be activated by parthenogenetic agents, and they will throw off fertilization membranes, and cleave quite like the fertilized nucleate eggs. This combination of parthenogenesis, in which the male nucleus is lacking, and merogony (development of a fertilized enucleate egg fragment), in which the female nucleus is lacking, may be termed “parthenogenetic merogony.” It is of particular interest since the resulting non-nucleate organism lacks chromosomes, genes and there- fore the hereditary qualities usually associated with the nucleus. MATERIAL—THE Parts oR FRACTIONS OF CENTRIFUGED EcGcGs The material has been obtained by breaking apart the, mature un- fertilized eggs of Arbacia punctulata by centrifugal force and using the non-nucleate parts. The method, as previously described (E. N. Harvey, 1931; E. B. Harvey, 1932) consists in centrifuging the eggs in a sucrose solution of the same tonicity and approximately the same density as the eggs. The eggs remain suspended in this solution during centrifugation and are free to elongate and break apart. ‘The stratifica- tion of the centrifuged egg is shown in Plate I, Fig. 1, and eggs in the process of stratification and breaking, as they are observed with the centrifuge microscope, are shown in Photograph 1. The eggs become dumb-bell shape and then usually break into two slightly unequal parts, a larger white half containing oil, clear layer, mitochondria and some yolk, and a smaller red half containing yolk and pigment (Figs. 2, 6 and Photographs 2, 4).1_ The nucleus is always in the white half under the oil—this is invariable in Arbacia punctulata. Both the halves will elongate with further centrifuging (Figs. 3, 7 and Photographs 3, 5) and break into quarters; the white half breaks into a larger clear quarter containing oil, clear layer and nucleus and practically 1 The sizes of the individual granules are: oil (spherical droplets) 0.6-1.04; mitochondria (fine granules) 0.6-1.0”; yolk (irregular or polyhedral granules) 0.7-1.1H; pigment (spherical granules) 1.1-1.64. The relative amount of formed bodies as ascertained by E. N. Harvey (1932) is: oil 1 per cent; mitochondria 4.8 per cent; yolk 27.2 per cent; pigment 5.5 per cent. 101 —ae . “SUISNJ1IJUID JOSUOC] YIM sojnueIs oy} JO supped Jayyinyz 0} onp st ajod jeyadrquas ay} ye / “SIyy Ul Bore eal oy < syIuse “ayeos 0} a]qGIssod se Aleyeanoge se epeut sydersojoyd pue soyojoysS eplon] esisWied WoOIF oie SSUIMeIP OY, “Syeotq - Hon ecient pue saajey oy} pue ‘(3 xX QOO'OT 3 SeMuIUE ¢ Jnoqe) 2d10F Jesnyrsz}UI9 Aq poyrzesys “wyojnjouUnd DIIDgG4p7 JO S30 pazipnseyun oLL J a1v1g pabnj i4jnaoad Pa Ie pabnjiajneg bby ajoum : ["\ (2) C 494e0) 2319 ers, ty : gee VIVINLONN ed = VWIOVEYV CEEAVAGE WITHOUL NUCLEI 103 no visible granules, and a smaller granular quarter containing all the mitochondria and some yolk granules (Figs. 4 and 5 and Photograph 3). The nucleus is invariably in the clear quarter. The red half egg breaks into a larger yolk quarter containing only yolk granules, and a very small pigment quarter containing all the pigment and a few yolk granules (Figs. 8, 9 and Photographs 1, extreme right, and 5). The halves and quarters are of a very definite and characteristic size and remarkably uniform in any particular batch of eggs (i.e. eggs from one female). The sizes of the halves and quarters (which | will call “fractions’”’) of typical batches are given in Table I; the table is based on the measurements of many eggs. from many typical batches and may be considered as a standard. The four parts or fractions which lack a nucleus are marked with asterisks; they are (1) granular quarter, (2) red half, (3) yolk quarter, (4) pigment quarter. There is no possibility of any of these fractions containing any portion of the female ‘nucleus, since in Arbacia this is always thrown intact under the oil cap. WWAwios IL Size of the Arbacia egg and its fractions obtained by centrifugal force. Approximate Parts Diameter Volume proportion of whole egg Le pe per cent WINOLEHE OOM nak Aa eur. Shoe dere e 74. 212,000 WWihitemiali Agel: pees 2) oe 62 125,000 60 Clear quarter............ 56 92,000 40 * Granular quarter........ 40 33,500 20 Pe ihvall PacMan rah sPuel in nag 56 92,000 40 pvolloquanten 1.0. e ee 52 73,600 33 * Pigment quarter........ SP 17,160 7 G@Nitclenstaey a teae acaae oy ese lca 11.5 796 0.4) * Non-nucleate. Though the great majority of batches of eggs conform quite closely to the sizes given in Table I, there are some consistent variations. A few batches occur each year in which there is a great disparity in the size of the halves, the red halves being very small (Photograph 7) ; this occurs uniformly throughout the batch irrespective of the speed of cen- trifuging. Then there occur a few batches which break into almost equal parts, through the mitochondrial layer (Photograph 8).? Then there occur in the course of the summer’s work one or two batches in 2 These figures for the relative size of the halves agree closely with those of Shapiro (1935) for slightly smaller eggs. 3 The measurements of Lucké (1932) to determine the distribution of osmoti- cally inactive material, include two sets (B and D) of this unusual type. 104 ETHEL BROWNE HARVEY which the eggs of one batch break at two levels, approximately half of the eggs break into a very small red half and large white half, and the others into a larger red half and a smaller white half (Photograph 9) ; this gives white halves (and of course the corresponding red halves) of two distinct sizes (Photograph 10). These variations from the standard type of breaking must be due to a difference in the physical structure of the egg, probably viscosity. Such would seem to be the case since the stratification varies with the position of the break; the mitochondrial layer is much better formed when the two halves are nearly equal (cf. Photographs 7 and 8). Apart, however, from these infrequent and consistent variations, the material is quite uniform, and there can be no question as to the structure of the various fractions, especially as to whether they are nucleate or not. MertTuop For the experimental work, the eggs are centrifuged in tubes a little larger than hematocrit tubes, of about 0.7 cc. capacity. The eggs in sea water are placed above a layer of 0.95 molal sucrose solution in the proportion of about 14 sea water. to 24 sucrose solution; this is isosmotic and isopycnotic with the eggs. After centrifuging for 3 to 4 minutes at about 10,000 « gravity, there are usually three distinct layers in the tubes, a layer of white halves above, a pinkish layer of elongate but as yet unbroken whole eggs a little below, and a layer of red halves at the bottom of the tube (Photograph 6, right). With further centri- EXPLANATION OF PHOTOGRAPHS 1-6 Centrifuged egg and egg fractions of Arbacia punctulata 1. Unfertilized Arbacia eggs stratifying and pulling apart as observed in the centrifuge microscope. Photographed while rotating at about 10,000 X gravity. At the bottom of the slide are the red non-nucleate halves. At the right is one of the red halves pulling apart into quarters; at the extreme right, the pO quarter above and the pigment quarter below. 2. White halves showing oil cap (black), clear layer, and granules (mito- chondria and yolk). The nucleus invariably lies under the oil cap. 3. White halves (at right and left) elongated by greater centrifugal force, and their two quarters, clear quarter (below) with nucleus under the oil cap, and granular quarter (above). 4. Red halves, non-nucleate, containing only yolk and pigment (black). 5. Red halves elongated by greater centrifugal force, separating into their two quarters, yolk quarter (larger and lighter) and pigment quarter (smaller and darker). 6. Centrifuge tubes (placed upright immediately after removal from the centri- fuge), showing masses of eggs and egg fractions in distinct layers. Right tube, three layers: (above) white halves; (center) elongate but as yet unbroken whole eggs; (bottom) red halves. Left tube, five layers: (1, at top) white halves; (2) unbroken whole eggs; (3) yolk quarters; (4) red halves; (5, at bottom) pigment quarters. EE Dea i pail PHOTOGRAPHS 1-6 105 ® PHortocraPHs 7-10 show unusual breaking of unfertilized Arbacia eggs. PuHo- TOGRAPHS 7-9 were taken while rotating on the centrifuge microscope. 7. One broken egg shows unusual breaking into a very large white half (above) and a very small red half (below). Mitochondrial layer not well formed. 8. Unusual breaking into almost equal halves. Mitochondrial layer well formed, and the break comes across this layer. : 9. Unusual breaking into two types in the same batch of eggs: (1) small white and large red halves at left and center; (2) large white and small red half at right. Nucleus shows well in some of these eggs. 10. Lower magnification of white halves of two distinct sizes, no intermediates, from such a batch as shown in Photograph 9. The larger ones appear darker be- cause of the mass of granules; these obscure the oil cap, which is plainly visible in the smaller type. 42. Plutei and blastule from whole and white half eggs, fertilized, 2 days old. Several white halves have developed into plutei half the normal size. Other white halves are still spherical blastule. (Low magnification.) 43. Parthenogenetic merogone from yolk quarter, 4 weeks old. Detailed drawing of same in Fig. 20, Plate Il. Highly magnified. 44 and 45. Fertilized merogones. 44. Fertilized yolk quarter with 2 daughter nuclei. Cell division only indicated by notch at left. Three hours after fertiliza- tion. 45. Fertilized red half with about 12 nuclei and no cell boundaries. Five hours after fertilization. PHotTocraPHs 7-10 106 CLE MVAGE Wil OU NUCEEL 107 fuging and using more of the sucrose solution, the red halves are often broken into quarters and there are five distinct layers in the tubes: (1) white halves, (2) elongate whole eggs, (3) yolk quarters, (4) red halves, (5) pigment quarters at the bottom (Photograph 6, left). One can decant off from the tubes the two upper layers of cells, which con- tain nuclei, and obtain a pure culture of non-nucleate fractions. ‘Thus is obtained the material for the study of parthenogenetic merogony, and it can be obtained in great abundance. The study has been made almost exclusively on living eggs and all photographs and drawings are from living material. I have not yet worked with the granular quarters of the white halves; these are more difficult to obtain as it re- quires prolonged centrifuging. The yolk quarters have proved to be excellent material for the observation of cytological details in the living eggs, owing to their lack of pigment. Even the pigment quarters, though only about one-fourteenth the volume of the whole egg, are activated and cleave. PARTHENOGENETIC AGENTS For parthenogenetic agents, I have used, for the most part, hyper- tonic solutions: sea water concentrated by boiling to half its volume, or sea water brought to a similar hypertonicity by the addition of NaCl in the proportion of 30 grams of NaCl to a liter of sea water. The addition of KCl to sea water (39 grams to a liter) is efficacious but renders the surface of the egg sticky. Also leaving the eggs in the sucrose solution for an hour or keeping them for 24 hours at 8° C. will initiate development. The eggs are left in the hypertonic solution for twenty minutes, and then returned to sea water. This procedure causes development in both nucleate and non-nucleate fractions, but the dif- ferent fractions, and also the whole eggs, from the same batch react quite differently. The same solution that activates the red halves may fail to activate the white halves or the whole eggs, and usually when the white halves develop the red halves do not. The batch, however, runs quite uniformly; if some of the red halves are activated, the ma- jority of them are; it is not sporadic. Stretched eggs are activated more readily than unstretched, probably because the membrane is thinner and materials can pass through more easily. The stretched whole egg is activated more readily than the normal uncentrifuged egg, and elon- gate fractions are activated more readily than the same fractions which have become spherical on standing. The eggs are treated, therefore, immediately after removal from the centrifuge. It may be that the sucrose solution in which the eggs are centrifuged helps the activation. 108 BP EE EB RO WEN IE era yi) DEVELOPMENT OF NON-NUCLEATE FRACTIONS (PARTHENOGENETIC MEROGONES ) The first sign of activation of any of the centrifuged fractions or whole eggs takes place while in the hypertonic solution; it is a “ setting ”’ of the egg. The egg retains its elongate shape and the granules do not redistribute. This is characteristic also of centrifuged whole eggs fertilized immediately after centrifuging; they retain their dumb-bell shape and the fertilization membrane follows the irregular contour of thekegon (EAB. larveys 1952) lin, however tiiey mane a iOtmmne Gal tzed! immediately after removal from the centrifuge, but are left unfertilized in sea water, they become spherical within an hour and the granules redistribute. Centrifuged whole eggs and nucleate fractions when acti- vated artificially are also “ set.” The non-nucleate fractions are likewise “set” (Photograph 11). In the best batches nearly 100 per cent are “set.” Then within ten or fifteen minutes, they usually throw off a fertilization membrane and the ectoplasmic layer forms. In some batches, the fertilization mem- brane does not form, but development proceeds nevertheless. If kept in the hypertonic solution, the eggs remain in the same condition for some hours with no further development. After treatment with hypertonic sea water for 20 minutes, the non- nucleate eggs are returned to sea water. Frequently the change from the hypertonic solution to the sea water causes a rupture of the fertiliza- EXPLANATION OF PLaTeE II Detailed camera lucida drawings of non-nucleate egg fractions (partheno- genetic merogones) from living material. The coarse stippling denotes yolk granules, the large solid dots denote pigment. (The times given are times after activation and return to sea water.) Fic. 10.. Formation of clear area (astrosphere) in non-nucleate yolk quarter. One hour after return to sea water. Fic. 11. Formation of monaster in non-nucleate yolk quarter. Two hours after return to sea water. Fic. 12. Formation of amphiaster. Fics. 13, 14. First cleavage. Fic. 15. Non-nucleate red half with many asters and no cell boundaries. “Crater blastula.” Still within fertilization membrane. Fic. 16. Cell boundaries have come in around the asters. Fic. 17. Embryo emerges from the fertilization membrane as individual cells which later disintegrate. Fic. 18. Three-day non-nucleate blastula from yolk quarter; large cells in- side, small cells outside. Fic. 19. Eight-day non-nucleate embryo from red half, vacuolated. Not di- vided into cells. Fic. 20. A. Four-week non-nucleate embryo from yolk quarter, not divided into cells, but with vacuoles and large spherical inclusions. 8B. More detailed draw- ing of the surface, showing possible cilia. Photograph 43 is of this same embryo, still living. 16 Prate IT 109 - ¥ e Bh 21 PwHorocrarHs 11-21 All are of living non-nucleate fractions (parthenogenetic merogones ) 11. One- and 2-cell stage of red halves; two cells nearly equal in size. Five hours after activation and return to sea water. 110 CLEAVAGE WITHOUT, NUCEET 111 tion membrane, owing to the swelling of the egg. After about an hour, a clear area may be observed in the living eggs, especially in the yolk quarters and in the yolk section of the red halves; this is an astrosphere, the beginning of a cytaster (Plate II, Fig..10). After another hour an aster 1s formed around this, forming a large and quite typical mon- aster (Fig. 11). After another hour one often observes two cytasters (Fig. 12), and this may be followed by a cleavage plane between the two asters in quite typical fashion (Figs. 13, 14). A spindle is not visible in the living material in these non-nucleate eggs nor indeed in any mitotic division of fertilized eggs. A thorough study of prepared and sectioned material has not yet been made, but a survey of a few slides has failed to reveal anything but asters in the parthenogenetic merogones; these are very granular along the rays. The activated non-nucleate fractions may divide into two equal or unequal cells (Photographs 11, 12, 13). In the best batches, about eighty per cent cleave. The cleavage plane in the red halves is not constant in position; it often comes in along the boundary between yolk and pigment, but this is not always so. It usually comes in across the narrow axis of an elongate egg, but often not the shortest axis. Very frequently, the egg is pinched in by the free edge of the now hardened fertilization membrane which had broken (usually at the centripetal pole) when the egg was taken from the hypertonic solution to sea water ; the cleavage plane usually comes in above this constriction (Photo- graphs 26, 36). This gives a very striking and characteristic 2-cell stage. Such a 2-cell stage is found also in elongate whole eggs treated parthenogenetically (Photograph 31), and also in elongate whole eggs fertilized immediately after centrifuging (Photograph 30). In the latter case, however, the constriction is the original narrowed portion 12. One- and 2-cell stage of red halves; two cells unequal; fertilization mem- brane present. Five hours after activation. 13. Two-cell stage of yolk quarter. Six hours after activation: 14. Four-cell stage of yolk quarter. Asters faintly visible, one in each cell. Five hours. 15. Six- and 8-cell red half. Four hours. 16. Eight- to 16-cell red halves. Five hours. 17. Eight- to 16-cell red halves. Cells still confined within fertilization mem- brane. Eight hours. 18. About 24-cell red half. Sixteen hours. 19. About 64-cell red half. Blastula in center breaking through the fertiliza- tion membrane, others still enclosed. Twelve hours. 20. Yolk quarter (not divided into cells), emerging from the fertilization membrane in amceboid fashion. Empty fertilization membrane at right. One and one-third days. 21. Yolk quarter, divided into many cells; about 500 in number. Two days. i, ETHEL BROWNE HARVEY of the egg preceding the break into two spheres which would have oc- curred with further centrifuging; it is not caused by a broken fertili- zation membrane as in the parthenogenetic eggs. In any of these cases, however, it is more usual for the cleavage plane to come in above (cen- tripetally to) the constriction and not through it, determined 1n all probability by the stratification. The separation between the two non- nucleate cells is a bona-fide cleavage plane and not merely a pinching in of the surface, as is evident from Photographs 26 and 27—the latter taken two minutes after the former. A second cleavage plane may divide the two non-nucleate cells into four, again between two asters (Photograph 14; asters are visible in Photographs 36 and 14), or it may divide only one of the first two blastomeres, giving a 3-cell stage. Or the non-nucleate egg may oc- casionally divide at once into 3 or 4 cells. By subsequent cleavages more cells are formed, and there occur fairly regular 8-, 12- and 16- cell stages, and intermediate numbers. Sometimes these are still en- closed in the fertilization membrane (Photograph 17); sometimes they form a loose cluster of cells not confined within a membrane (Ehotovcaphs Ws, lo 1824925, 28,7298 37436) 5) Cells boundaries sometimes disappear and new cell boundaries are formed. I have not found any indication of micromeres at the 16-cell stage of the non- nucleate egg, but these apparently do not form even in the whole eggs fertilized immediately after centrifuging, where the cleavage is quite irregular owing to the shape of the egg. The cleavages of the yolk quarters are more regular than those of the red halves, probably be- cause they lack the heavy red pigment and are more homogeneous and more nearly spherical. The pigment cells of the red halves are usually slower to divide than the yolk cells and are therefore, in general, larger. These same facts are true also of similar fertilized fractions. Even the pigment quarters, though only about one-fourteenth the volume of the whole egg, will cleave when activated. They often divide EXPLANATION OF PHOTOGRAPHS 22-29 All are of living non-nucleate fractions (parthenogenetic merogones ) 22-25. Series from one and the same red half egg. Times denote hours after activation and return to sea water. 26. Red half 1144 hours after activation. First cleavage plane (at left) begin- ning through shorter axis. The constriction at right is caused by the pinching in of the broken and now hardened fertilization membrane, and is not a cleavage plane. 27. Same egg 2 minutes later, showing further progress of cleavage plane. 28. At right, 3-cell stage of pigment quarter. At left, about 8-cell stage of red half. Six hours after activation. 29. At right, below, 4-cell stage of pigment quarter. Above, about 6-cell stage of red half. Six hours after activation. eS SZ Lie PHOTOGRAPHS 22-29 113 114 ETHEL BROWNE HARVEY into two equal cells, and by subsequent cleavages into as many as 16 cells (Photographs 28, 29, 37). These unfertilized activated quarters have in my experiments cleaved better than the fertilized pigment quar- ters; I have obtained the latter with 16 nuclei but without cell bound- aries. It should be especially noted that cleavage follows upon cleavage in the non-nucleate egg in a fairly regular sequence. Photographs 22—25 are a series from the same egg at various intervals. The cleavages are not so regular as in normal uncentrifuged fertilized eggs, where they occur with clock-like regularity, but they are almost as regular in oc- currence as in fertilized or parthenogenetic centrifuged eggs and in fertilized fractions. Indeed the similarity of the cleavage figures in the non-nucleate egg fractions to those in nucleate ones or in nucleate whole eggs is very striking. The cleavage of a centrifuged egg fer- tilized immediately after centrifuging is different from that of a normal fertilized egg owing to its elongate shape and to the uneven distribution of granules. The cleavage of an elongate stratified non-nucleate egg fraction is precisely like that of a similar nucleate fraction. Whether EXPLANATION OF PHotToGRAPHS 30-41 Two vertical columns arranged for comparison of nucleate (30-35) with non- nucleate (36-40) eggs at similar stages of development. 30. Two-cell stage of whole ege fertilized immediately after centrifuging. One hour after fertilization. 31. Two-cell stage of whole egg, unfertilized but artificially activated imme- diately after centrifuging. Four hours after activation. 32. Below, 2-cell stage of pigment quarter. Above, 8-cell stage of yolk quar- ter. Three hours after fertilization. 33. Sixteen- to 32-cell stage of whole eggs 3 hours after fertilization. At right, the two components from the first two cells have developed separately without fertilization membrane; with further development these would produce twins. 34. Early blastule 4 hours after fertilization. At left, whole egg. At right, white half, unusually large. Still within fertilization membranes. 35. Late blastula from yolk quarter, 1 day after fertilization. 36. Two-cell stage of non-nucleate red half. Note two asters in upper cell. Six hours after activation. 37. Above, at left, 8-cell stage of non-nucleate yolk quarter. Above, center, 2-cell stage of pigment quarter. Below, right, 8-cell stage of red half. All non- nucleate. Five hours after activation. 38. Sixteen- to 32-cell red halves, non-nucleate. Nine hours after activation. At right, the two components from the first two cells have developed separately ; with further development these would produce twin blastulz. 39. Early non-nucleate blastulz, 12 hours after activation. At left, red half. At right, yolk quarter. Still within fertilization membranes. 40. Late non-nucleate blastula from yolk quarter, 2 days after activation. 41. Left, pluteus from whole egg, 3 days after fertilization. Right, above, small blastula from non-nucleate yolk quarter, 3 days after activation. To show com- parative size and differentiation after 3 days. 42-45 follow Photograph 10. CLEAVAGE WITHOUT NUCLEI HS) NUCLEATE NON-NUCLEATE 7 Par thenogenetic Merogone 6 hrs Fertilized Parthenogenetic thr 4 hrs 4hrs 2 3 days Puorocrarus 30-41 116 ETHEL BROWNE HARVEY the nucleate egg or fraction has one or two nuclei, that is, whether it is fertilized, or parthenogenetic, or merogonic, makes little difference in the type of cleavage. Compare, for instance, the 2-cell stage of a fer- tilized whole egg (Photograph 30), a parthenogenetic whole egg (Photograph 31), and a non-nucleate red half (Photograph 36). The first cleavage, as mentioned previously, in an elongate egg usually comes in across the shorter axis in the less dense area. Later cleavages of nucleate and non-nucleate eggs are closely similar, as a comparison of Photographs 32-35 with those of Photographs 37-40 will show. Cleav- age stages of nucleate eggs are at the left and the corresponding stages of non-nucleate ones at the right. Even the frequent independent de- velopment of the first two blastomeres is similar (Photographs 33 and 38). The main difference is in the rate of cleavage. The non-nucleate eggs cleave much more slowly. Whereas it takes about fifty minutes (at 23° C.) for the first cleavage of a fertilized whole egg, whether normal or centrifuged, and the same time for the fertilized white half, it takes about six hours after activation for the first cleavage of a non-nucleate egg. The times for first cleavage as well as later ones are extremely variable in different sets of the non-nucleate eggs. A nucleate egg, treated parthenogenetically, 1s intermediate in rate; it takes about four hours after activation for the first cleavage. The merogonic egg fractions, i.e., fertilized enucleate parts of eggs, are also slow to cleave (Photograph 32; see also E. B. Harvey, 1932). In general, eggs with 2 nuclei (¢ and 2) develop most rapidly, those with one (0 or 2) less so and those without any nucleus most slowly; the very slow development of the fertilized clear quarter is an excep- tion (1932). It may be that the nucleus contains an enzyme which accelerates cleavage, but apparently certain granules must be present. Later cleavages of non-nucleate fractions (parthenogenetic mero- gones) result in the formation of a blastula. The blastula, if still con- fined within the fertilization membrane, may be observed breaking through the membrane after about twelve hours (Photographs 19, 39). It frequently happens that cleavage planes do not come in, especially in the dense red halves, and more and more cytasters appear as time goes on, giving the surface somewhat the appearance of the surface of the moon, pitted by many craters (Fig. 15). Exactly the same phenomenon was noted in the development of the fertilized red half-eggs (Fig. 20, 1932), only here nuclear membranes are present. Photographs 44 and 45 are of fertilized enucleate fractions showing a multinucleate condition without cell boundaries. Cleavage planes frequently come in, in these red halves both fertilized and non-nucleate, just before the breaking of the fertilization membrane, thus dividing the egg into many cells CLEAVAGE WITHOUT NUCLEI 117 (Fig. 16). The non-nucleate blastula sometimes emerges as isolated cells instead of as.a complete organism, and then goes to pieces (Fig. 17). Sometimes the cleavage planes do not come in even at this late stage, and the embryo emerges from the fertilization membrane as a compact mass of protoplasm, somewhat amoeboid in appearance, but perfectly viable (Photograph 20). The oldest cellular non-nucleate blastule obtained were from yolk quarters two and three days old (Fig. 18, Photographs 21, 40). These consisted of somewhat differentiated cells, the inside cells larger than the peripheral; the individual cells were filled with granules and small vacuoles or fat droplets. As many as 500 cells have been counted in these blastule. Staining im toto with aceto-carmine failed to reveal nuclei or chromosome plates which are well brought out in fertilized yolk quarters similarly treated. The em- bryos usually do not survive more than three or four days owing both to their low vitality and to the accumulation of decomposing eggs in the culture with resultant growth of bacteria and protozoa. Some of the non-cellular embryos which were isolated have survived longer ; an eight- day embryo from a red half is shown in Fig. 19, consisting of granular protoplasm filled with large vacuoles and some pigment. The oldest embryo was still viable after a month (Photograph 43, following Photo- graph 10; and Fig. 20). This, from a yolk quarter, was a mass of protoplasm filled with yellowish granules and small vacuoles and larger spherical inclusions, possibly individual cells. There were fine strands radiating from the protoplasm to and possibly through the thin en- veloping membrane characteristic also of normal free-swimming Dlas- tule; these are possibly short cilia, and the organism may have been slightly motile. Whether a non-nucleate egg can give rise to a true free- swimming ciliated blastula is still uncertain. It can, however, develop into a many-celled embryo having a certain amount of differentiation in that the cells are of different sizes; or into an embryo without cell boundaries but with differentiated protoplasm, perfectly viable a month after activation with parthenogenetic agents. The unfertilized and un- treated sea-urchin eggs live only two or three days at room temperature, according to Loeb and Lewis (1902); they certainly do not survive longer than this in mass cultures without special care. The striking similarity of the nucleate and non-nucleate eggs stops with the late blastula of one or two days (Cf. Photographs 35 and 40). I have seen no indication in the non-nucleate embryo of the formation of a gut or of the arms and skeleton characteristic of a pluteus. The normal fertilized egg has developed into a well-formed pluteus in three days, whereas the parthenogenetic merogone is still a spherical blastula. A photograph (41) of one of the 3-day parthenogenetic non-nucleate 118 ETHEL BROWNE HARVEY yolk quarters together with a pluteus of the same age from a normal fertilized egg shows the disparity in size and differentiation. The parthenogenetic yolk quarter of a month remained the same size with little change, whereas the normal pluteus in that time changes greatly in size and shape. Since the non-nucleate eggs are slower to develop in all stages, it is just possible that with longer time they might develop further than the blastula. Such a possibility is suggested by a com- parison of larve from fertilized half and whole eggs 2 days old. In Photograph 42 (following Photograph 10) are shown normal plutei together with half-size plutei from fertilized white half-eggs. There are also present many spherical blastulz of the half-eggs (which are slower to differentiate) not yet developed into plutei. As observation showed, many of these did develop into plutei later on. A comparison of Photographs 41 and 42 would at least suggest that the small blastula of the parthenogenetic merogone might also develop into a pluteus. DISCUSSION There has been considerable discussion as to whether cells without nuclei can divide, or whether asters can divide in the absence of chro- matin. Boveri (1918) threw doubt himself on his earlier conclusions that asters (but not cells) could divide without chromatin. Daleq (1931) has reviewed all the pertinent cases and has come to the con- clusion that they are all subject to criticism, and that cell division does not take place without chromatin, and this seems to be the case with his own experiments. Fankhauser (1929, 1934) believes, however, that the large number of non-nucleate cells, some with amphiasters and even spindles, in his merogonic Triton eggs, indicates their origin from other non-nucleate cells, though he believes that nucleate cells must be also present to influence division. The presence of accessory sperm nuclei in the developing amphibian egg prevents these results from being absolutely clear-cut. Wilson (1901) was the first to describe multiplication of asters in the entire absence of chromatin in the non-nucleate fragments (ob- tained by shaking) of To.ropneustes treated parthenogenetically, but _he found no cell division. His results were questioned by I*ry (1925), who maintained that in enucleate fragments accurately cut by hand and treated with parthenogenetic agents, asters might arise and irregular cleavage come in but the asters did not multiply. McClendon (1908), with a primitive micro-manipulator, very ingeniously removed the nu- cleus from the starfish egg and treated the egg with a parthenogenetic agent. He concluded that cells could divide without chromatin though CLEAVAGE WITHOUT NUCLEI 119 the segmentation was quite irregular. These results were not unequi- vocal since many of the operated eggs contained chromatin. There seems no doubt from the present observations that cell division can take place without nuclei. I have previously (1935) shown that the nucleus may be moved by centrifugal force from its normal position and cell division come in with no relation to its final position. Cleavage may also come in while the nucleus is still intact, as I noted previously in Parechinus (1935, Fig. 11), and as I have recently observed under certain conditions in Arbacia. The present work shows that the nucleus does not have to be present at all in cleavage. Asters arise de novo in non-nucleate eggs and become more numerous with time. Normal cleav- ages come in exactly as they do in nucleate eggs of the same sort and shape, and cleavage follows upon cleavage in a fairly orderly fashion. An egg fragment lacking both maternal and paternal chromosomes has given rise by repeated cleavages to an embryo containing about 500 cells with a certain amount of differentiation. An embryo arising from a non-nucleate egg has lived a month; it is probable but not certain that these have cilia and are free-swimming. The early stages of develop- ment can, therefore, take place without chromosomes. This means that the maternal cytoplasm is of great importance and has within itself the potentialities of determining at least the early stages of development. These potentialities must either (1) be given to the cytoplasm by the nucleus previously, or (2) be innate in the maternal cytoplasm and en- tirely non-nuclear. They might be given to the cytoplasm either at the time of the breakdown of the germinal vesicle, or else earlier by the chromosomes of a preceding generation. The former alternative cannot be tested experimentally since the egg and egg fragments cannot be arti- ficially activated, nor are they fertilizable, before rupture of the germinal vesicle. An influence on egg cytoplasm caused by the chromosomes of a preceding generation has been postulated for spiral twisting in snails and for certain characters in the silkworm and in corn. It would cer- tainly be difficult to prove experimentally in the case of the sea-urchin egg. The second supposition, i.e., that the potentialities of early develop- ment are innate in the cytoplasm, restricts inheritance by genes to later developmental stages, and it may very well be that only the more specific and differential characters are controlled by the genes, whereas the gen- eral and fundamental characteristics of living matter are cytoplasmic. If further work shows, however, that the development of the partheno- genetic merogones does not stop with the blastula, but that they will develop into plutei, the results will be difficult to harmonize with the accepted ideas of the mechanism of genetical inheritance. I have shown by previous studies (1932, 1933) that the visible gran- 120 ETHEL BROWNE HARVEY ules in the eggs of several species of sea urchins are not of great im- portance in development. Any part of the egg seems capable of de- velopment, and many parts have, when fertilized, given rise to quite normal plutei with skeletons. Any difficulty in development seems to be mechanical rather than structural, such as the breaking apart of the blastomeres in the clear quarters of the Arbacia egg, owing to the thin- ness of the membranes; or the absence of cleavage planes in many of the red halves, owing to the very dense granular cytoplasm. No par- ticular type of visible and moveable granules seems essential to develop- ment; these must be concerned with metabolism and respiration. It must therefore be the “ ground substance ” which is the material funda- mental for development—the matrix, which is not moved by centrifugal force and which, in the living egg, is optically empty. SUMMARY 1. Non-nucleate fractions of Arbacia eggs obtained by centrifuging can be activated by parthenogenetic agents, and develop; these may be termed “ parthenogenetic merogones.” 2. Asters are formed and cell division takes place without nuclei. 3. Cleavages of non-nucleate fractions are strikingly similar to those of nucleate fractions (or whole eggs) of the same stage of de- velopment. 4. Many cleavages, in sequence, result in the formation of a blastula of as many as 500 cells. 5. Blastula emerge from the fertilization membranes; one embryo was still viable after a month. ILIEIRT STACI Ofigcles, (CLAN SID) Boveri, TuH., 1918. Zwei Fehlerquellen bei Merogonieversuchen und die Entwick- lungsfahigkeit merogonisher und partiell-merogonisher Seeigelbastarde Arch. f. Entw.-mech., 44: 417. Datcg, A., 1931. Contribution a l’analyse des fonctions nucléaires dans l’ontogé: nése de la grenouille. I. Arch. de Biol., 41: 143. FANKHAUSER, G., 1929. Ueber die Beteiligung kernloser Strahlungen (Cytaster ) an der Furchung geschniirter Triton-Eier. Rev. suisse de zool., 36: 179 FANKHAUSER, G., 1934. Cytological studies on egg fragments of the salamander Triton. IV. Jour. Exper. Zool., 67: 349. Fry, H. J., 1925. Asters in artificial parthenogenesis. II. Jour. Exper. Zodl., 43: 49. Harvey, E. B., 1932. The development of half and quarter eggs of Arbacia punc- tulata and of strongly centrifuged whole eggs. Biol. Bull., 62: 155. Harvey, E. B., 1933. Development of the parts of sea urchin eggs separated by centrifugal force. Biol. Bull., 64: 125. Harvey, E. B., 1935. The mitotic figure and cleavage plane in the egg of Parechi- nus microtuberculatus as influenced by centrifugal force. Biol. Bull., 69: 287. CLEAVAGE WITHOUT NUCLEI 21 Harvey, E. N., 1931. The tension at the surface of marine eggs, especially those of the sea-urchin, Arbacia. Biol. Bull., 61: 273. Harvey, E. N., 1932. Physical and chemical constants of the egg of the sea- urchin, Arbacia punctulata. Biol. Bull., 62: 141. Loes, J.. anp W. H. Lewts, 1902. On the prolongation of the life of the un- fertilized eggs of sea-urchins by potassium cyanide. Am. Jour. Physiol., 6: 305. Lucx#&, B., 1932. On osmotic behavior of living cell fragments. Jour. Cell. and Comp. Physiol., 2: 193. McCuenpon, J. F., 1908. The segmentation of eggs of Asterias forbesii deprived of chromatin. Arch. f. Entw.-mech., 26: 662. Suapiro, H., 1935. The respiration of fragments obtained by centrifuging the egg of the sea-urchin, Arbacia punctulata. Jour. Cell. and Comp. Physiol., 6: 101. Witson, E. B., 1901. Experimental studies in cytology. I. Arch. f. Entw.-mech., 1275593 SEQUENCE) OF FUNCIION ASE GU AweP EASE SaliN TEREDO Wer RR) GO (From the Osborn Zoological Laboratory, Yale University) In a recent number of this journal certain questions have been raised by Grave and Smith (1936) as to the sequence of the functional sexual phases and the proportions of individuals in each phase in the successive age groups of the variety of Teredo navalis found so abun- dantly at Woods Hole and other localities on the coast of southern New England. Evidence is now available which will answer these questions with some degree of positiveness. In his earlier studies on the biology of this species at Woods Hole the senior author (Grave, 1928) stated that the females outnumber the males but he made no mention of the possibility of a change of sex. The recent investigations by Grave and Smith, however, indicate a se- quence of alternating phases of functional sexuality comparable with that of certain species of oysters. They thereby confirm in most re- spects the conclusions reported during the past few years on the sexual rhythm of this species (Coe, 1933, 1934a, 1935). Yet in regard to the proportions of individuals in each of the successive phases there is some divergence of opinion and it will be shown on the following pages that the reason for this discrepancy may be partly due to the failure of Grave and Smith to follow the life cycle through the entire sequence of sexual phases. Their published investigations cover only the two months from June 28 to August 27, 1935. It is quite illogical to treat the sperm-producing and egg-producing phases of the teredo as distinct categories of individuals, since they rep- resent merely successive stages in the life of a single individual. If the number of older animals in the functional female phase is found to be twice as great as in the alternate phase of sexuality, the explanation may be simply that the duration of one sexual phase averages twice as long as the other, always assuming that no selection, intentional or otherwise, as to ages is involved. Perhaps an exception should be made in regard to the relatively few individuals known as true males, in some of which the initial sexual phase may be retained for an indeterminate period. In still fewer cases the individual appears to be proterogynous. Protandry in Teredo has been indicated, although not recognized as 122 SEQUENCE OF FUNCTIONAL SEXUAL PHASES IN TEREDO 123 such, since the work of Quatrefages nearly a century ago, for he esti- mated that there were about twelve times as many females as males in the adult population. Nelson later suggested an even greater dis- parity of the sexes, with as many as five hundred females to one male among large individuals. Yonge (1926), on the other hand, found defi- nite evidence of protandry in two specimens of T. norvegica, while Sigerfoos (1908) had previously suggested protandry in Bankia gouldt. Substantial proof of a sequence of sperm-producing and egg-pro- ducing phases during the lifetime of the same individual has been pub- lished in the series of recent articles to which reference has been made (Coe, 1933a, 1933b, 1934a, 1935). In the first of these papers it was reported that an actual transition from the functional male phase to the functional female had been observed in about half of the small indi- viduals infesting ropes in New Haven Harbor. The second paper (Coe, 1933b) gave detailed descriptions of the transformations of the primi- tive bisexual gonad of the very young individual, through the initial male phase and thence to the female phase toward the end of the breeding season. No evidence was available at that time that the female phase might again revert to the male condition but this transformation of sex was reported, with details of the cellular changes involved, in the third paper of the series (Coe, 1934a). In a more recent brief summary of the sexual changes which the sev- eral types of individuals are believed to undergo the successive phases were schematized (Coe, 1935) as follows: evinle . Stemale =. / recuperation .. . female. 2. Male ... female . . . recuperation ... male... female. Se Niale. 2. recuperation . - . temale. HaeiVale (‘true male’) .”. . recuperation ... male... - (female?). 5. (Exceptional.) Female ... recuperation .. . (male?). To this series the paper by Grave and Smith (1936) adds one addi- tional aspect, namely, that the transformation from female to male phase may occur abruptly during the activities of the breeding season as well as in the recuperation period. | PRopoRTION OF INDIVIDUALS IN Eacu FUNCTIONAL SEXUAL PHASE AT SUCCESSIVE AGES AND AT DIFFERENT SEASONS The accompanying Tables I and II indicate the sexual condition of 2,599 individuals grouped according to their sizes and approximate or maximum ages at each season of the year. The gonads of about three hundred of these were cut and mounted as serial sections; for each of the others the gonad was teased and examined microscopically. Most 124 Wr Re COE of the collections were made at Woods Hole, Massachusetts ; others were obtained from Milford, Connecticut, through the courtesy of Dr. Victor Loosanoff, and a few from Barnegat Bay, New Jersey, through the kindness of Dr. T. C. Nelson. Examination of 388 additional large individuals during the third week in June, 1936, showed 29 true males, 81 bisexual males, 17 her- maphrodites, 95 females with nearly ripe ova, and 166 females carrying embryos. Omitting the hermaphrodites, these numbers would repre- sent a population having 29.65 per cent of the adults in the functional male phase, with 70.35 per cent functioning as females at the time of the examination. These data were obtained too late to be included either in the graph or in Table IJ, but they would not have changed greatly the percentages there indicated except for the larger proportion of females with embryos. Since the sexual phases of most individuals change progressively with their advance in age, it is obviously impossible to classify them with great accuracy. Furthermore, there are so many individual dif- ferences in the expression of the male and female characteristics that the difficulty is greatly increased. Even if all individuals with ripe spermatozoa are classed as functional males, there are always found in a large collection some with a few sperm associated with many ova. Still more difficult and uncertain is the separation of functional males into the two groups of true males and bisexual (hermaphroditic) males, since all grades occur in the relative size and abundance of the ovocytes which even in many of the true males can be detected along the walls of some of the follicles. Nor can the precise age of any individual be determined with ac- curacy because the teredo will not live and grow normally outside the wood in which it feeds. The maximum age is determined by reckoning from the time when the wooden block is placed in the water during the breeding season. The size is not strictly a reliable criterion of age, since rapidity of growth depends so largely on environmental conditions. The characteristics of the shell and of the pallets are sometimes helpful. With these limitations in mind the data in Tables I and II will show the seasonal trend of the successive broods and the sequence of sexual changes which they undergo from immaturity to advanced age. Attentiori may first be directed to the group of 43 individuals shown in Table I to have been collected on July 19 from a block which had been in the water for only 20 days. In this brief period 24, or more than half the total number, had already reached the functional male phase, while one only was producing ovocytes exclusively. The 18 indi- viduals which were sexually immature were evidently still younger and later arrivals. SEQUENCE OF FUNCTIONAL SEXUAL PHASES IN TEREDO 125 It thus appears that there were in this lot 24 protandric individuals as compared with one which was proterogynous. The evidence from the entire series of 911 small teredos listed in Table I is consistent with the conclusion that the first sexual phase is, with few exceptions, mascu- line. The gonads of these young individuals, however, form an inter- grading series, from those which have only a few small ovogonia or ovocytes on the walls of the follicles (true males) to those which are so strongly feminine that they consist mainly of large ovocytes, with rela- tively few spermatogenic cells in the lumens. All except the true males are grouped in Tables I and II as bisexual males. It will be noted that more than 10 per cent of the small teredos listed were in the female phase. Presumably most of these were dwarfed animals that had already completed a brief initial male phase rather than strictly proterogynous individuals. The second sexual phase, which immediately follows the initial male phase except in the autumn, is usually feminine. True males, however, may resume and continue spermatogenesis for an indeterminate period. No proof is at present available that a third sexual phase occurs in these localities during the same year. More frequently a long recupera- tion period follows the completion of the female phase in late summer and autumn. During the winter the gonad of such individuals is re- organized and the third sexual phase, which may be of either type of sexuality, develops during the late winter and early spring. The breed- ing period lasts from about the middle of May to the middle of October. The over-wintering teredos will thus include individuals of all ages, from those young ones which settled on the blocks in the autumn and are still sexually immature to those which completed at least two sexual phases the preceding breeding season. The most numerous group, however, will be of the younger ages. In the pre-spawning season (Table II) the two alternate sexual phases are found in about equal numbers, or with a small majority in the female phase. Some of the individuals classed as being in the bi- sexual male phase evidently have but a brief period of sperm produc- tion at the beginning of the second spawning season, for the ratio of animals in the female phase appears to increase with advancing age (Table Il). The percentages in each sexual phase are shown graph- ically in Fig. 1. It is evident either that the number of animals chang- ing from male to female phase is greater than the number going in the opposite direction or that the female phase is of longer duration than the alternative aspect of sexuality. The data compiled by Grave and Smith (1936), on the other hand, 126 W. RCOE would indicate that the ratio of animals in the female phase decreases during the summer from approximately 3 to 1 (150-58) in June to 46-55 (65?) in August, or about 1 to 1 (58-55) if the hermaphrodites are included among the females. The differences between these data and those here recorded may be due in part to chance selection and in part to differences in classification. In many cases the bisexual gonad shows such an intermediate condition between male and female phases as to make any distinction more or less arbitrary. Moreover, shortly after the middle of August the first young of the year may attain a length of 100 mm. or more—a size sufficient to be classed as “large specimens.” If some of these were to be included with the older indi- viduals there would appear to be a rapid increase in the male-phase ratio. (ABrE ol Sexual phases of small individuals (20-60 mm. in length) of approximately known ages Maxi- Sere True | Bisexual) Young pies need Date mum are male male female fea with | Total age eae phase phase phase Gineaie ieee days Mia 20 ie ba, ke astsyeos, oie — 0 4 18 0 0 0 Ds Vitale Snare Saunas 2 — 0 5 21 20 0 0 46 Tune s2 eta he ickeaveceos ae — 0 5 29 13 7 0 54 Ait ya rian ccciaonaee ys 20*| 18 6 18 1 0 0 43 ANU ISVs das5o0e0 68 71 8 71 415 6 4 3 507 Sept. 19-24........... 112 0 5 80 16 12 6 119 SeptasOee Rees es 42 | 41 0 12 1 0 0 54 INO ae 1s cea tees Renae 73 | 16 11 327 5 2 0 66 Totalieewa see eens 83 107 625 62 25 9 911 Percent: ea ee ke 9.11} 11.75 | 68.61 | 6.80] 2.74) 0.99 * From Barnegat Bay, N. J. Courtesy Dr. T. C. Nelson. {+ Winter condition. It is, however, in the late spawning season that the greatest pre- ponderance of large individuals in the female phase is found; for in- stead of the former ratio of about 2 to 1, a collection of 164 large indi- viduals in September gave a female ratio of approximately 10 to 1 (not counting 7 hermaphrodites), as shown in Table Il and Fig. 1. These animals, although only 3 to 4 months of age, had already reached a length of more than 100 mm. and had, with relatively few exceptions, already completed the initial male phase. By selecting a group meas- uring 70-95 mm. in length, some of the younger, male phases were in- cluded and the ratio approximated equality (Fig. 1). Selection of 119 still smaller and younger animals on the same dates showed more SEQUENCE OF FUNCTIONAL SEXUAL PHASES IN TEREDO 127 ARs JUL Sexual phases of large individuals (90-275 mm. in length) Date phe 20-25... see ee we ee ey July 23 ec ee es ewe Sept. 19-24.... Per cent Total all seasons...... Per cent eee ee ae Maxi- T: Her- Unri it ace fale seal ApH feriale eae (ae Spawned Total aa eal g phase phase dites phase | phase |embryos Pre-spawning Season 7-10 | 11 66 0 48 81 0 0 206 7-10] 0 33 0 16 0 0 0 49 7-10 | 3 21 0 30 0 0 0 54 8-11] 8 32 0 5 9 1 0 55 22 152 0 99 90 1 0 364 6.05| 41.76] 0 27.16} 24.73) 0.3 0 Early Spawning Season 8-11 | 7 35 0 10 11 11 0 74 8-11} 4 35 0 11 14 12 0 76 9-12 | 11 12 0 32 Sd 17 0 129 9-12 | 3 18 0 7 24 6 0 58 9-12] 7 39 0 18 51 it 0 126 32 139 ~ 0 78 157 57 0 463 6.91 | 30.03) 0 16.85! 33.90! 12.311 0 Middle Spawning Season | 12-14} 5 1 0 0 5 22 0 33 2-3 | 18 60 3 0 82 Sil 0 220 23 61 3 0 87 79 0 253 9.09 | 24.1 1.19 0 34.40] 31.20) 0 Late Spawning Season 3-4 9 5 7 0 52 84 7 164 5.5 3.0 4.3 0 Silod || Sil 4.3 End of Spawning Season 4 2 5 0 4 13 0 94 118 4 8 11 0 0 23 9 41 92 2-3 0 0 0 0 6 0 de 79 3 0 0 0 0 0 0 65* 65 5-6 3 if 0 0 0 0 70* 90 13 33 0 4 42 9 343 444 2.92 7.42} 0 0.90} 9.45} 2.03} 77.25 99 390 10 181 428 230 350 1688 Pes EES6 2310) (OL59 sO 225235 SrO2 ig 2 O57: * Winter condition. 128 ‘ W. R. COE than twice as many functioning as males as were in the alternate phase of sexuality, and the proportion of sperm-bearing individuals in the total population was still greater, since the number of young animals greatly exceeded that of the older individuals in the female phases. Toward the end of the breeding season, when most of the older indi- viduals have spawned out there is again an approximately equal number of the large animals in each sexual phase (Table II; Fig. 1). Incom- CEN vrs z 5. a5, a Wy WwW go fi g 0) 7/8 8 ; 7, < rales : 2 qd |Z < 5 -2 ae Oe : OSI ence re R : uh vi ui | ; , i Apa U) o- wrag TH 2 ee sea 2 Wee eae $ Foe 0 ix Ye) “Dy, —\=S~< . |e ree eo ® e—e —e—i-e oe? —— hose ooctin sadn APRIL MAY-JUNE SULY-AUG. SEPT OCT-NOV. Fic. 1. Graphs showing the proportions of large individuals (90-275 mm. in length) in each phase of sexuality before, during and at the end of the breeding season. Based on a total of 1,688 individuals. It is to be noted that the chart includes two generations of animals, the over-wintering individuals being sup- planted by the new broods which attain the “large” size after the middle of the spawning season. The graphs indicating total male and female phases in the late spawning season differentiate two classes of individuals, namely, those 70-95 mm. in length and those of still larger sizes. SEQUENCE OF FUNCTIONAL SEXUAL PHASES IN TEREDO 129 pleted sexual phases may be retained during the winter, but most of the partially developed sexual cells will be absorbed by phagocytosis. THe Norma LIFE CYCLE Turning now from the statistical study of selected groups of animals to a consideration of the sequence of sexual phases realized during the normal lifetime of the single individual, it becomes evident that several different influences are involved. Examination of the gonads of ani- mals of approximately known ages shows clearly the protandric nature of the vast majority of all individuals. Ripe spermatozoa may be produced within three to five weeks after metamorphosis. The ensuing phase is predominantly female, the percentage of fe- males being influenced by the season when hatched and presumably by hereditary factors as well as by the environment, since all grades of bisexuality appear simultaneously. Eggs may have been liberated and embryos found in the gill chambers within six weeks after metamor- phosis. There is no doubt of the strong tendency toward a series of alternating phases, the number of such phases that may be realized by any individual depending largely upon its length of life. At least four alternating phases may occur—two in each of two breeding seasons, but it is not improbable that the number may occasionally exceed four. The cooler water of early autumn checks the reproductive activity of such individuals as have not already ceased gametogenesis for the season and entered upon the recuperative phase. Most of the oldest individuals are killed by parasitic Protozoa about the middle of the breeding season, as noted by Grave and Smith (1936). The epidemic later affects the earlier broods of the same year, causing the death of many individuals not more than three to five months of age. The gonads of the survivors, including the spawned-out individuals as well as the others, are reorganized during the winter, in many cases with a change in the sexual phase. Disintegrating residual cells from the previous phase of sexual activity are removed by phagocytosis before the proliferation of the definitive germinal cells. TRUE MALE PHASE This phase represents the extreme aspect of masculinity in that the gonads have only a very few ovogonia or minute ovocytes along the walls of the follicles. Such individuals form but a small proportion of the youngest sexually mature animals (Table I). They occur also as the third sexual phase, following the discharge of the ovocytes of the female phase, but appear to be indistinguishable from those which retain the initial male phase after a long period of growth (Table IL). 130 W., RECOE Individuals of this type were observed by Grave and Smith (1936), who conclude that a “careful study of adult individuals leads us to doubt if there is such a thing as a true male or a true female Teredo in the sense that it remains permanently one sex, as postulated by Coe, who agrees that sex inversion in true males may ultimately take place.” The two principal parts of this sentence are obviously contradictory, and the word “ permanently ” has been substituted for “ indefinitely ” which was used in the papers under discussion. Otherwise there would have been no ground for disagreement, for “true males” are defined in those papers (Coe, 19330) as individuals in which “ the preliminary male phase is of longer duration and the crop of spermatozoa is much larger.” The fact that such individuals are found at earliest sexual maturity as well as later is indicative of the operation of genetic factors which differentiate them from others of the same brood. The latter reveal by their gonads a more feminine tendency and are classed as bisexual males or protandric females. This appears to be equally true of other mollusks in which a similar type of sexuality has been reported (Coe, 19345, 1936). BISEXUAL MALE PHASE The majority of young individuals on reaching sexual maturity are obviously protandric, with gonads showing a highly variable propor- tion of spermatogenic and ovogenic cells. During the recuperation period there is frequently a return to a similar condition (Tables I and I, Jere, Jk))e FUNCTIONAL BISEXUALITY Most of the sexually mature teredos examined during the breeding season were actively producing either sperm or ova, but a few indi- viduals occur, as Grave and Smith (1936) have reported, which con- tain functional gametes of both types. Only 10 of these hermaphro- dites were recognized among the 2,599 animals examined (Table II). In other cases spermatogenesis is well advanced while embryos are present in the gill chambers. Moreover, in an occasional individual some follicles of the gonad may be producing spermatozoa while ad- jacent follicles contain nearly ripe ova. Such hermaphrodites would not of themselves be indicative of a change of sex in either direction. Of the 618 sexually active individuals reported by Grave and Smith, 27 were classed as hermaphrodites. A collection of 388 adults from the Eel Pond at Woods Hole examined during the third week in June, 1936, contained 17 individuals in this phase of sexuality. SEQUENCE OF FUNCTIONAL SEXUAL PHASES IN TEREDO 131 PROTEROGYNY The finding of an occasional individual in the female phase at earliest sexual maturity indicates that the initial male phase may some- times be aborted or omitted, as it is known to be in some other species of normally protandric mollusks (Coe, 1936). PostT-SPAWNING PHASE The “ spawned-out ” condition which may lead again to the bisexual phase or to either of the unisexual phases during the period of re- cuperation appears to follow the female phase more often than other- wise, although a similar condition results when spermatogenesis ceases and the sperm are discharged without a reciprocal growth of ovocytes. This phase occurs principally toward the end of the breeding season (Table II, Fig. 1), but is also found whenever gametogenesis ceases from any cause. SUMMARY Examination of the gonads of 2,987 specimens of Teredo navalis of all ages, at all seasons, and during three successive years supplements the conclusions previously published regarding the sequence of sexual phases in this species. The functional phase, if any, in which any given individual is found will, in general, depend upon: a, age; J, size; c, season; d, season when hatched; e, its particular combination of genetic factors. Each normal individual shows a tendency toward an alternating series of functional sexual phases. There is good evidence that at least four of these phases may sometimes be completed in a lifetime of less than two years, although the average individual experiences hardly more than one or two. Among the variable sequences of phases which different individuals may experience the following may be considered typical: Pe Walewen . kemale: Z. Walle o's), ieaale oy py imelle 5 bg TSE: Se Valen ee teniale even tilales AL Malle 4 5 Gaels Ys. ieneile. 5. Male (« ce male my Hf iblalers poe teniadles Gy Miaten (= stauermales )\ ee eesimlale: Jepiicinialem (exceptional) es) a ilialejay 1.) female: Because the population consists of a vastly greater number of indi- viduals of the younger ages than of the older age groups, it is obvious 132 W. COE that death by suffocation, starvation, parasitization or some other cause will usually terminate the individual’s existence before the normal cycle is completed. Protandric bisexuality 1s generally characteristic of the early sexual phases, with a tendency toward alternating unisexuality in later life. Of the latter, the female phase is usually of longer duration than the alternate male phase. The proportion of individuals in each sexual phase in the general population or in selected groups will obviously be correlated with a combination of the conditions mentioned above, but limited by associ- ated environmental influences. The total population of teredos living in a particular block of wood may thus consist of upwards of 90 per cent of individuals in either functional sexual phase or of about equal percentages of each, according to the age group or groups represented at the time of the examination. Attention has been previously (Coe, 1934a) called to the similarity of the sexual rhythm in Teredo to that of the larviparous oysters, “ al- though the usually shorter life of the former limits the number of the alternating sexual phases.” There is some resemblance also to the sexual changes in the oviparous oysters (Coe, 1934b) and in the hard- shell clam Venus (Loosanoff, 1936), which are predominantly, but not exclusively, protandric, with a later change to the seasonally unisexual condition of the older individuals, in which there is an approximate equality in the sex ratio. LITERATURE CITED Coe, W. R., 1933a. Destruction of mooring ropes by Teredo; growth and habits in an unusual environment. Science, 77: 447. Cor, W. R., 1933b. Sexual phases in Teredo. Biol. Bull., 65: 283. Cor, W. R., 1934a. Sexual rhythm in the pelecypod mollusk Teredo. Science, 80: 192. Cor, W. R., 1934b. Alternation of sexuality in oysters. Am. Nat., 68: 236. Coz, W. R., 1935. Sequence of sexual phases in Teredo, Ostrea and Crepidula. Anat. Rec., 64, suppl.: 81. Cor, W. R., 1936. Sex ratios and sex changes in mollusks. Manifestation Paul Pelseneer. Bruxelles, 1936: 69. Grave, B. H., 1928. Natural History of shipworm, Teredo navalis, at Woods Hole Massachusetts: Biol. Bull., 55: 260. Grave, B. H. anp JAY SMITH, 1936. Sex inversion in Teredo navalis and its relation to sex ratios. Buol. Bull., 70: 332. Loosanorr, Victor, 1936. Sexual phases in the quohog (Venus mercenaria). Science, 83: 287. Sicerroos, C. P., 1908. Natural history, organization and late development of the Teredinide, or ship-worms. Bull. U. S. Bur. Fish., 27: 193. Yonce, C. M., 1926. Protandry in Teredo norvegica. Quart. Jour. Micr. Sci., 70: 391. ii OCCURRENCE AND SIGNIFICANCE OF NITRITE IN (sha, SBA NORRIS W. RAKESTRAW (From the Woods Hole Oceanographic Institution and the Metcalf Chemical Laboratory of Brown University) In studying the nitrogen cycle in the sea a consideration of the nitrite concentration in the water is especially useful, because of its intermedi- ate position between ammonia on the one hand and nitrate on the other. Although it is not a stable end-product, its very instability makes it a possible indicator of the state of equilibrium between the oxidation and reduction processes making up the cycle. Its order of magnitude in the water is much smaller than that of nitrate, or even ammonia, but the methods for its analysis are correspondingly more accurate, so that its normal variations may be followed as closely as those of the other nitrogen compounds. In view of the ease with which nitrite determinations may be made, it is surprising that there are not more data on its occurrence in sea water. Raben (1914, 1918) and Brandt (1927) were among the first to report such figures, but their data were only scattered and frag- mentary. Orr (1926) reported some analyses of samples from the vicinity of the Clyde Sea, made observations on diatom cultures, and discussed the factors concerned in the appearance and disappearance of nitrite. Harvey (1926) failed to find significant amounts of ni- trite in the English Channel or in deeper waters of the Atlantic further south, but noticed their occasional appearance in stored samples. Atkins (1930), however, was able to make a fairly comprehensive study of the seasonal changes in nitrite in the English Channel, finding high concentrations at certain levels and at certain times. Orton (1924), investigating various influences upon the growth of oysters, found mod- erately high concentrations in samples from Plymouth Sound, Thames Estuary and Helford River. He also followed nitrite changes in the water of storage tanks. In a few cases, figures for nitrite have been reported incidentally, or in connection with other chemical studies of sea water, as by Braarud and Klem (1931), Robinson and Wirth (1934), Rakestraw (1932, 1933), Matsudaira and Yasui (1932, 1933), and in connection with 1 Contribution No. 110. 133 134 NORRIS W. RAKESTRAW the Antarctic cruises of the “ Discovery”’ (1932). Three good sea- sonal studies have appeared more recently, by Soot-Ryen (1932), Verjbinskaya (1932), and Cooper (1933). The last of these is most comprehensive, containing very complete data for the relationship be- tween various chemical factors in the water. Unfortunately, its limi- tation to one or two shallow stations in the English Channel raises a question as to the validity of its conclusions when applied to other, more general conditions. The present paper includes results collected over a period of five years at the Woods Hole Oceanographic Institution in the effort to explain the production and utilization of nitrite and its place in the nitrogen cycle of the sea and the general process of marine metabolism. The data include analyses from: 1. A small group of inshore stations in shallow bays. 2. Stations on the continental shelf to the south of Cape Cod, including two serial sections to the edge of the shelf. 3. A number of stations in the Gulf of Maine, so grouped as to indi- cate seasonal variations for more than a year, at seven different locations. 4. A group of deep sea stations to the south of Cape Cod, from the edge of the continental shelf to Bermuda, at different times of the year. 5. Stations to the south of Bermuda, across the edge of the Sargasso Sea to the Bahamas, and fifteen stations in the Caribbean. 6. A large number of scattered surface samples, at various places and times. 7. Bottom muds and waters contained in them. 8. Samples stored under a variety of conditions. MeEtTHODS The Griess-Ilosvay method, used in this work as in that of prac- tically all previous investigators, has been repeatedly described, with minor variations. It depends upon the color developed in the sample by means of sulphanilic acid and alpha-naphthylamine (or its dimethyl or diethyl derivative). Buch (1923) established its suitability for sea water by a careful spectrophotometric study. It is simple, reliable and not easily susceptible to temperature changes and other fortuitous fac- tors. ‘There is no salt effect upon the color development, so that stand- ards may be made up in distilled water. It was easily established that the color developed was constant between 2 and 48 hours, an adequate range for routine purposes. Two milliliters of sulphanilic acid solution (8 grams in 1 liter of OCCURRENCE OF NITRITE IN THE SEA 195 30 per cent acetic acid) are added to 100 ml. of the sample, followed by 2 ml. of a solution of alpha-naphthylamine, or diethyl-alpha-naph- thylamine (8 grams in 1 liter of 30 per cent acetic acid, with 10 ml. of conc. HCl added). After two hours or more the samples are compared with appropriate standards. This was done in a colorimeter of the Hehner tube type, which has long been used in this laboratory for phos- phate, nitrite and other determinations. We have seldom had trouble in obtaining nitrite-free distilled water (as Cooper seems to have had), and have very frequently found sea water samples free of nitrite to the limit of sensitivity of the method, which is well under .005 microgram atom per liter. Stock standards have been made either from sodium nitrite or from silver nitrite, pre- cipitated by NaCl. In either case, such solutions may be preserved for many weeks and their concentrations checked by titration with KMnO,, or iodimetrically. Dilute standards for use are prepared from these and kept for not longer than one day. TABLE [| Accuracy of the analytical method for nitrite. Triplicate analyses Depth in meters A B Cc 30 4.3 4.3 4.4 mg./cu. m. 307 307 Bs) ug. at./liter 40 1.9 1.9 1.9 mg./cu. m. 135 -135 135 ug. at./liter 50 1.6 1.6 1.65 mg./cu. m. 115 115 118 ug. at./liter Throughout this work results will be expressed in microgram atoms per liter, according to the recent recommendations of the International Council. These may be translated into the older units of milligrams of nitrite-nitrogen per cubic meter by multiplying by 14. In order to indicate the accuracy and reproducibility of results by — this method there are given in Table I the results of triplicate analyses (A, B, C) on three different water samples, expressed in both the old and new systems of units. Later results are given to the nearest 005 microgram atom. 136 NORRIS W. RAKESTRAW VARIABILITY OF NITRITE As far as our experience goes, the nitrite concentration in sea water varies normally from 0 to .35 microgram atom per liter. Occasional values up to double this amount have been encountered, and the highest found (in Cape Cod Bay) was .91. No values like those reported by Atkins (2.78, or 38.9 mgm./m.*), or by Matsudaira and Yasui, in the Yellow Sea, have ever been found in normal waters on this side of the Atlantic. (0) 10 iS CES) fe Ae A 8:29 9:31 10.33 35 12:37 1:39 2:41 343 445 5:47 6:49 AM AM AM AM PM PM PM PM. PM PM PM SCARE oe eee 27 M Fic. 1. Variation of nitrite during a tidal cycle. Station K-2, Buzzards Bay. Ordinates, depths in meters. STATION 1722 Wri4 Weis W-i5 WI-I5 WI-16 VWI-I6 7PM l2MN 7AM 6 PM 9AM 1AM Fic. 2. Variation of nitrite during a tidal cycle. Station 1722, Gulf of Maine. Ordinates, depths in meters. Nitrite is one of the most variable factors in the sea. Analyses from two stations close together, or from the same station at different times not far apart, sometimes yield quite different results. In fact, the horizontal distribution of nitrite is highly dependent upon local con- ditions. This would be better appreciated if previous data had not been so scattered and incomplete. Cooper seems to have been the only one sufficiently well aware of the fact. OCCURRENCE OF NITRITE IN THE SEA 137 In view of this, it is rather surprising that any general trends can be seen and conclusions drawn. Absolute values for nitrite concen- tration, especially isolated ones, are of little use; only by a comparison of relative amounts, with fairly complete data, can one obtain a clear picture of conditions. To show how general trends do appear, however, the nitrite con- centration was followed at two stations (K—2 and 1722) throughout a complete tidal cycle. The results are shown in Figs. 1 and 2. Here, as in all the subsequent figures, the vertical distribution of nitrite is represented by a solid black area. The vertical variation in nitrite concentration is one of the most striking and interesting facts concerning it. As will later be shown, both the production and utilization of nitrite are processes taking place TABLE II Stations in shallow bays. Nitrite concentrations in microgram atoms per liter. Station:| K-1 | K-2 | K-3 L-1 L-2 R-3 | R-4 | S-1 S-2 Depthin| 8-30 | 8-30 | 8-30 | 7-18 | 8-28 | 7-18 | 8-28 | 8-26 | 8-26 | 7-14 | 7-14 meters | 1931 | 1931 | 1931 | 1931 | 1931 | 1931 | 1931 | 1932 | 1932 | 1932 | 1932 0 0 0 0 0 0 0 07 0 0 0 0 5 0 0 01 0 Al 10 01 0 0 0 01 0 10 085 61 86 15 01 20 01 0 O01 0 035 43 .20 40 91 in particular, but different, layers. This often gives rise to large ver- tical differences, and usually, in the summer, a maximum zone, of vari- able depth, sharpness and intensity. Good examples of this are found in Atkins’ results from the English Channel. Strangely, Cooper’s data from the same location, a few years later, fail to show the phenomenon —probably, as he suggests, due to different hydrographic conditions that year. The nitrite concentration in bottom muds is most highly variable and will receive special attention later. All this variability in nitrite concentration is only to be expected from the precarious position which it occupies in relation to other sub- stances in the water. Conceivably, it may be produced by oxidation of ammonia or reduction of nitrate, by means of bacteria, catalytically or photochemically. It is not impossible that the phytoplankton may play a part in the oxidation or reduction. In any event, nitrite is one step 138 NORRIS W. RAKESTRAW in the process of decay and “ mineralization” of organic matter, either formed or particulate, or dissolved in the water. On the other hand, nitrite may be destroyed by: (1) oxidation to nitrate or reduction to ammonia—atmospherically, photochemically, or TaBLe III Stations on the continental shelf, near Cape Cod. Nitrite concentrations in microgram atoms per liter. Station: 1270!) 1295 1274) 12060 12.72% ial 07 meee sen| e298) 0| (2 o0ls 29 . Depth in 6-29 | 7-20 | 6-29 | 7-20 | 6-29 | 7-20 | 6-29 | 7-21 | 6-28 | 7-21 meters 1932 | 1932 | 1932 | 1932 | 1932 | 1932 | 1932 | 1932 | 1932 | 1932 0 0 == L0G |) == OL OS .07 = = = 10 0 0 0 0 0 0 01 0 0 01 20 01 | O 0 0 0 01 01 0 0 01 30 01 plitts LS) 01 0 40 085 | .07 14 0 50 Al 60 .09 23 ald) 75 0 26 | .36 100 125 0 01 150 175 01 Station: 1261 | 1283 | 1282 | 1281 | 1280 | 1725 | 1739 | 1684 | 2143 | 2156 Depth in 6-28 | 7-1 7-1 7-1 7-1 | 7-25 | 7-31 | 7-27 | 5-8 | S211 meters 1932 | 1932 | 1932 | 1932 | 1932 | 1933 | 1933 | 1933 | 1934 | 1934 0 0 0 0 0 0 0 0 0 04 0 10 0 0 0 0 0 0 0 .05 0 20 0 0 0 0 0 02 0 .06 0 30 aif 0 0 44 | 01 07 0 40 .065 44 | 39 | .O1 50 (NO 39 0 0 yal 08 | .085 60 20M" -245a.03 75 36 0 all alll 100 O01 AD | OF | 2 0 Sie ert 125 01 26 | .26 150 .03 .05 7s Oi 175 .05 .05 .08 200 0 O01 by bacteria; or, (2), consumption by phytoplankton. We cannot be altogether certain whether or not free nitrogen enters into the process in any way. However, at any moment, the concentration of nitrite is an expression of the equilibrium between production and destruction. The distribution of nitrite may therefore be related to a large number OCCURRENCE OF NITRITE IN THE SEA 139 of factors: the distribution of bacteria, phytoplankton and zooplankton ; temperature; light intensity; oxygen tension in the water; density and circulation of the water; and perhaps other seasonal influences. INA IW Serial sections on the continental shelf. Nitrite concentrations in microgram atoms per liter. Line BONY? July—August, 1931 Miles off: Depth in meters 1 2 5 7 10 13 39 61 82 90 0 0 0 0 0 0 0 0 0 0 0 10 07 0 0 0 0 .09 0 25 13 .06 .04 10 0 0 35 03 | .115 50 0) 0 65 40 100 135 0 565 .035 885 0 Line ‘‘D” August 5-6, 1931 Miles off: Depth in meters Me D 5 10 20 30 40 50 66 0 0 0 0 0 0 0 0 0 0 10 0 20 -05 01 0 01 0 0 0 0 40 .09 .215 60 .66 0 75 05 AS .165 100 0 200 .035 DESCRIPTION OF DATA The results from shallow, inshore stations, shown in Table II, were mostly obtained in the early stages of the work and are useful in dem- onstrating that the occurrence of nitrite is not closely associated with such influences as land drainage. Although concentrations are usually low, there are exceptions, as at Stations S—1 and S—2, where the highest concentrations of all were found. Great differences in local conditions evidently exist. 140 NORRIS W. RAKESTRAW The stations on the continental shelf, shown in Tables III and IV, and in Fig. 3, present a fairly regular picture. Since these were all sampled during the summer months, nitrite was seldom found at the surface in more than small traces, while significant amounts were en- countered below 30 meters, or near. the bottom, if shallower than that. Surface conditions off New York Harbor were somewhat exceptional. Lines A and D, to the edge of the shelf, bring out the fact that the rate of nitrification on the bottom, as indicated by nitrite concentration there, is apparently greatest at depths of 50 to 75 meters. The number- 01270 1285 Pall 129 3772 1273 1260.” 1297@ 273 260 ,” 2 .@f261 1298@ “5 - S299 lyn y A O\ - Fic. 3. Location of stations in the Gulf of Maine and vicinity. ing of stations on these lines (4-5, A-39, D—20, etc.) indicates the number of miles off-shore, in each case. An unfortunate spacing of the sampling levels at the deeper. stations on these lines, as well as else- where on the shelf, failed to reveal the zone of maximum nitrite gener- ally found at such stations between 30 and 50 meters, and which prob- ably existed here also. At the deepest stations, along the edge of the shelf (1725, 1684, 2143 and 2156), this maximum zone is clearly seen. Its depth varies considerably, showing the influence of the adjacent oceanic water, in which the maximum zone is invariably deeper than at stations on the shelf. 141 j=) © S Saal oqoQooooqcooocoo OCCURRENCE OF NITRITE IN THE SEA 0 0 000% 40° 00s¢ sco’ 0 000€ Seo’ 0 0 00SZ O | cor) Sar} © 0 0 0 0002 | way 0 008T 0 wy | O 0 0 0 0 009T cor} ©} © 0 0 0 0 OOFT 0 Lot Oo} O 10° ® | © | @ 10° 0 0 00ZT 0 0 OOTT oO} © | to] Mor} © | Oo 0 © | @ | 0 0 0 0 0 0 OO0T 0} @ 0 | OOF | © F 0 Oo} © | @ 0} 0 | 10 0 | 0 0 008 0 | zo 0} 1 | © | Oo oO} Oe} oO | o 0 | zo | 10° @ | io | © 0 009 Oj Ol OP @ | Oo} © | a | © | © 10° | O | O 0 00S © | wor | oy} wer | Mer | Woe Ol} O OF O | Oo 0] O |s9t'| O O | O | wy} o 10° 00F 0| 0 0 ZO’ | 40° O)}o% | © | © | © Iso | O | Wo} © oO | © eM) © | © OOF 0 | 10° 0 £0" 0 $90 | O 80° 0 0SZ 0 or | Wer) SOF | Or |) HOP ISSO | GOP |) O | © | OO er] O |} Oar | GO} ©} © | 10 002 0 | 6c | zo’! 10° | zo’ | 40° | S€0' |$90° | O | O | O |S90° | ZO" {Sst FI | SO @ | Woy OST @ | ie aor aor) Tor |} Co} Sor] oO O | © | OO | SO] ie tl Co} we | Ze | © fF © | iz | Setoy OOT 09° | So'| zo’ | 10° ¢0°| O © | © | SO | to QP | 90° | 07 cL ge | Sot’ | £0" | 90° @ || Se 0 0 09 0 el Oo | so | O | tO] Sor h O | O | © | OO | © IGSie | Os? Sie) © | © 1 Se 1 Oe | o os Gls 0|/ 0} O £0° Sir 0 OF O} OF O | Oe 0 0 0 o¢ Ola |l oO} @} O | O | wl or | ©} O | © Ise) O fw} OF OF Se | oO} O | 20} Oo 02 0 @} O | @ Seo" 10° tT 0 0 OT Oo} © o| O | @ lore} = ler | or, oO |} © IS! © |i | ©) ee} O1 © 1a} 0 SS6L | FE6T | SS6T | S61 | EL6T | SE6T | EF6T | SE6T | ELOT | SEOT | SEOT | SEGT | EEOT | FE6T | SEOT | SOT | SEoT EF6T | SE6T | SE6T | SL6T S19}oUL Tc-L L-S O1-6 | LZd-L | Le-L | LI-S LEG €Z-1 | Ol-@ | PC-L | EC-L | EC-T | 8C-L HAG (SW | (6E=It, | WEL WW eae reg | et |) ele, ul yidoq OStZ | ZFIZ| TSLT | O€ZT | E89T | OFOT | E9FT | LIEZ | SOFT | O9FZ | SSPZ | OTET | FELT | ZOFT | LSPS | OSHC | STET | CELT | SSHC eTec | TELt sTOTJEIS SNOUT IZISITL 7 dno15 ¢ dnory q dnoin J dnoiy “Jo}] Jod sul0je WeIZOAIUM UT SUOTZLI}UBIUOD 9}IVIN “DpNUdag 0} pod advgQ ‘suoynis vas daaq A ai1avy, 142 NORRIS W. RAKESTRAW TABLE VI Seasonal variation of nitrite in the Gulf of Maine. atoms per liter. Nitrite concentrations in microgram Location 3 Station: | 1284 | XX | 1699 | 1748 | 1857 | 1864 | 1909 | 2022 | 2074. 2168 | 2220 | 2274 Depthin | 7-12 | 5-6 | 7-3 | 9-4 |10-19| 12-2 | 1-9 | 3-22 | 4-18 | 5-22 | 6-26 | 9-19 meters 1932 | 1933 | 1933 | 1933 | 1933 | 1933 | 1934 | 1934 | 1934 | 1934 | 1934 | 1934 0 0 0 0 O |.215).09 | .08 | .085 | .10 | .01 | .035 | O 10 0 0 0 0 OSI 2027 OSSa elms 0 0 20 0 .06 | 0 0 .08 115] 0 02 30 0 se Oil) @ Pky Us} | 40 Pollil Wels |) O 0 11 40 165 14 | .04 oy || al 50 os |) AKO) |} itil all || De O | .085 | .135 ilsiey |) 0) 60 32 SO me O2, .065 .39 | .09 0 80 01 | O 0 0 0 |.09 | .06 |.14 | .04 0 100 .02 | .02 | O .03 | .01 | 0 Oy ers) | OB} iS yy SO 0 125 0 0 .065 | 0 10 | .02 0 150 0 01 | O 01 | 0 .01 | .04 | O 01 | .03 0 175 03 | O 0 01 0 200 035 0 0 0 0 0 0 0 02 0 225 0 03 | 0 0 250 0 0 0 0 0 Location 5 Station: 1287 | 1705 | 1761 | 1871 | 1911 | 2031 | 2083 | 2176 | 2224 | 2282 Dea | we |) rer ees | S| Te Sa | a || See | eae || on meters 1932 | 1933 | 1933 | 1933 | 1934 | 1934 | 1934 | 1934 | 1934 | 1934 0 .035 | 0 0 .035 0 .09 11 .08 0 0 10 0 0 035 0 ADS) sf oll | W 0 0 20 0 0 08 | .02 0 12 18 0 30 .10 07 0 02 0 JOSH 4) gl] ol 28 035 40 25 085 0 02 0 21 230 50 .09 .09 0 02 0 08 12 ANS | Ald | 0 60 .03 04 0 .03 0 alli) |} oil 80 01 01 0 02 0 .05 05 milion eales 0 100 01 01 0 01 0 Sh |) 4085 | S03 || O 0 125 01 01 0 01 0 02 .03 .10 0 01 150 .03 01 0 03 0 030 aie Ol OS 0 0 175 01 0 0 .035 | 0 0 0 0 200 0 0 Table V contains the deep stations from Cape Cod to Bermuda. A number of these are located in roughly corresponding positions and are arranged in four groups accordingly. Of these, No. 3 is most complete, with four stations at different times of the year. plotted in Fig. 8 to show the seasonal variations. These have been At all these stations, the maximum zone is found between 50 and 100 meters, with only traces OCCURRENCE OF NITRITE IN THE SEA 143 TABLE VII Seasonal variation in the Gulf of Maine (continued). Nitrite concentrations in micro- gram atoms per liter. Location 7 Station: 1712 1765 1888 1916 2037 2089 2182 2286 Depth in 7-8 9-7 12-7 1-10 3-24 4-20 5-25 9-21 meters 1933 1933 1933 1934 1934 1934 1934 1934 0 0 0 03 .02 .065 .09 12 19 10 0 01 .02 01 07 10 12 25) 20 165 16 .03 01 13 30 Pe, PS) 05 01 .065 10 11 33 40 21 16 .035 allil 50 18 .04 .05 0 .06 12 .06 31 60 lui 035 .035 .06 80 ely .035 .03 0 .06 13 .09 .30 100 .065 .035 .02 0 .065 13 10 .06 125 .02 0 .02 .06 07 085 0 150 03 01 .02 02 04 08 0 175 04 02 05 0 200 0 0 Location 8 Station: 1716 1764 | 1853 | 1886 | 1914 | 2035 | 2087 | 2180 | 2227 | 2288 Depth in 7-9 9-7 10-19 | 12-6 | 1-10 | 3-24 | 4-20 | 5-25 | 6-29 | 9-21 meters 1933 1933 | 1933 | 1933 | 1934 | 1934 | 1934 | 1934 | 1934 | 1934 0 0 0 32 .085 .06 .065 | .04 .065 0 0 10 0 0 .08 .06 .05 05 01 0 0 20 0 0 28 O85" 11 .03 24 30 115 .07 07 05 .06 SiS | cilil .085 | .10 40 21 01 40 .08 .08 07 50 215 0 .085 .03 .05 115 | .165 | .05 04 60 al 01 15 .085 17 05 80 02 0 .16 .07 01 06 08 17 0 0 100 0 .02 .08 085 | .01 01 035 | .15 0 0 125 0 .02 O01 .03 135 0 .035 150 .04 01 01 0 0 0 0 13 0 01 175 .02 0 0 0 0 .09 0 0 200 0 08 0) 0 0 .05 0 01 225 0 0 0 0 04 250 0 0 0 of nitrite below this. Only during the winter months is nitrite found at the surface. The extremely sharp maximum zone at Station 1781 may be noted, at which nearly all the nitrite in the water column occurs in a twenty-meter layer, somewhat nearer the surface than at stations further to the south. Stations in the Gulf of Maine, shown in Tables VI, VII and VIII 144 NORRIS W. RAKESTRAW are grouped into seven “locations.” These include Location 22 on Georges Bank, which may be considered an adjunct to the Gulf of Maine. The separate stations in each of the locations are close enough 245! @1683 24509 @2312 17308 2313 1731@ 2455@ 81732 24 50082457 2460 1734@ @$ 2316 2458®@ e 462g ©2317 1463 @ 1465 4 @1467 1606 + ® 1596 3s! @(572 1588 ng ide ISB9@@ gpl542 1541 i524 See Fic. 4. Location of stations from Cape Cod to the Caribbean. together to be comparable, and were visited on different cruises at various times of the year, covering a total period of thirteen months or more. The data from six of these locations have been plotted in Figs. 5, 6, and 7 to show the seasonal variations of both nitrite and nitrate. The @@CCURRENCE OF NITRITE IN THE SERA 145 data from Location 9 (not tabulated) are shown in Fig. 12, to indi- cate the relation between nitrite concentration, temperature and density. These figures from the Gulf of Maine are but a small portion of the TABLE VIII Seasonal variation in the Gulf of Maine (continued). Nitrite concentration in micro- gram atoms per liter. Location 15 _ Station: 1659 | 1781 | 1846 | 1899 | 1924 | 2052 | 2118 | 2137 | 2197 | 2296 Depth in 6-22 | 9-10 | 10-19 | 12-7 | 1-11 | 3-26 | 4-30 | 5-6 | 5-28 | 9-23 meters 1933 | 1933 | 1933 | 1933 | 1934 | 1934 | 1934 | 1934 | 1934 | 1934 0 0 0 SUL) Nase a 085 | 0 0 01 0 10 0 0 LOS) 0 0 0 0 20 0 0 14 | 15 065 | .085 | .07 0 0 30 Onl mili 11 0 0 0 40 O20 I) |) Oss) 01S 10 0 50 06 | .18 .08 SS 11 065 | .01 60 09 165 eS oma 0G: .10 LS) atl 80 05 | .02 01 0 01 .10 08 16 1859-205 100 Smee 02 02 0 0 07 16 085 | .07 125 O01 | .02 01 135 04 01 150 01 02 0 0 0 BLS) 05 03 035 175 SOF Aah 0 .03 203) 9) vatslioy |e. 02 _ 200 01 | .02 02 0 0 07 (OL yy aL) KO 225 .06 0 250 LO Or: .02 0 0 0 0 0 Location 22 Station: 1686 1742 1934 2069 2154 2214 2269 Depth in 7-27 9-3 1-13 3-29 5-10 6-2 9-18 meters 1933 1933 1934 1934 1934 1934 1934 0 0 0 .03 = .05 05 .04 10 0 0 035 .04 .06 .06 20 .03 085 035 -035 04 30 035 5 .03 035 .05 08 40 04 165 03 .05 50 04 18 07 .03 .05 aks 60 04 1385 .03 035 .05 80 185 0 21 analyses made during a survey of that region, the results of which will appear at a later time. The stations to the south of Bermuda, including those in the Carib- bean (Table IX), present a rather. different picture from that in the north. Generally, smaller amounts of nitrite are found, but frequently 146 NORRIS W. RAKESTRAW relatively large amounts at the surface. The maximum zone is near 100 meters, sometimes sharp, sometimes poorly defined. ie LOCATION 22 -LEGEND- MICROGRAM- ATOMS PER LITER >18 10-18 <02 02-07 O7-10 % ‘) sR Mia unniiines, RSS cau NITRATE RS XHRYXANAN NS RN) lente OR 7 p POSER one ree SR Ry \ Y/ ARAN) RR RXEXYY /)} RK RIN RYEXX RXR \ 1) He <3 3-10 10-15 LOCATION 3 Fic. 5. Seasonal variation of nitrite and nitrate. Locations 3 and 22, Gulf of Maine. Ordinates, depths in meters. RELATION OF NITRITE TO OTHER FACTORS Nitrate and Phosphate Since nitrite participates in the metabolic cycle with the other “ nutri- ent salts,’ it would seem important to compare its variations and be- havior with that of the other nutrients, nitrate and phosphate. Figures 9 and 10 show these relations at a number of stations in the Gulf of Maine at various times of the year. Certain conclusions are evident : 1. High nitrite concentrations in the upper water levels are associated with high concentrations of nitrate and phosphate. The converse is not always true, especially at certain times of the year. During the winter, when the surface is being replenished with nutrients by overturn of the water, mass, nitrite is also in evidence, particularly as the season ad- vances. In January, the production of nitrite is at a minimum, and since this corresponds to the time of greatest instability of the water mass it occasionally happens that little or no nitrite is found at any level, OCCURRENCE OF NITRITE IN THE SEA 147 oe % % ie RKO y/ \) 50(\) / NITRATE 100- y LOGATION 5 LOCATION 7 Fic. 6. Seasonal variation of nitrite and nitrate. Locations 5 and 7, Gulf of Maine. (See Fig. 5 for legend.) Ordinates, depths in meters. LOCATION 8 LOCATION [5 Fic. 7. Seasonal variation of nitrite and nitrate. Locations 8 and 15, Gulf of Maine. (See Fig. 5 for legend.) Ordinates, depths in meters. 148 * NORRIS W. RAKESTRAW even though there is abundant nitrate. This condition obtains at Sta- tions 1908 and 1910. 2. High nitrate and phosphate in the deeper water is not associated with nitrite. The latter is seldom found in large amounts below 100 meters. It is evidently produced near the surface and works down through the unstable layer during the winter until stratification again sets in. 3. When diatom growth begins in the spring, with the utilization of nitrate and phosphate, nitrite also disappears in the surface water. At Station 2099 this process has apparently just begun, while at 2098 the season is further advanced, nitrate and phosphate are down almost to their summer values at the surface, and nitrite is absent. Figure 10 is compiled from a number of stations during the summer months te r of 0) 200 400 600 800 1000 1462 1734 2316 2460 1-7-33 W28-33 21-23-35 VE24-35 Fic. 8. Seasonal variation of nitrite at an oceanic station. (Group 3, Table V.) Ordinates, depths in meters. show the close correlation between the gradients of nitrite, nitrate and phosphate near the surface. The origin points for nitrite and nitrate coincide in these diagrams, but the phosphate curves are arbitrarily displaced to the left to facilitate comparison. The shape, slope and position of the curves indicate that the same conditions govern all three. As a consequence of this correlation, it is proposed that nitrite, which is normally produced near. the surface, is utilized by the phyto- plankton during the period of its growth, in the same manner as nitrate and phosphate, so that the upper levels become depleted. If the upper right-hand corner of the nitrite area shown in Fig. 9 for Station 1896 (or 1884) were cut off, and the curves for nitrate and phosphate deflected towards the origin, a condition would be represented exactly similar to those in Fig. 10. This is what would be expected during July or Sep- tember. Abundant evidence of the utilization of nitrite by diatoms is 149 OCCURRENCE OF NITRITE IN THE SEA 10° 10° 0 0 000P 0 OOSE 10° 10° 0 0 0 0 0 000€ 0 0 0 0 0 00SZ 0 0 0 0 0 0 0 0 0002 v0" 0 0 0081 CO" 0 LO} a | 0 0 0091 10 0 cO 0 0 0 0 0 0 Oort 0 CO” WO 0) 0 0 0 00cT 10° | £0 OOTT 10° | cO" 0 10° 0 10° 0 0 0 0 0 0 OO0T 0 SOM alOr 10° 0 006 LO} 0 | co} Vo) 10° 0 0 10° DOS a0. 0 0 0 008 0 cO'| vO | cO O0L 10°; O 0 0 |} co) co | 10 A) 0 0 | 10°} vO 0 0 0 0 0 | 10 009 10° | 10° 0 O | wor) Or |} clo) 10° 0 | 10° @ | HO aor wor | 0 0 | 10° 00S 10’ | O 0 0 | 20 | 20 | 10 £0" |) War | aor) oor) a) Wo) 0 0 0 0 00P cO | O 0 0} £0'| £0 | cO 10° On LOM LOM ROOMS cO se iOn ev 0 0 0 0 00€ £0 | 10 0 0 | S€0'| ZO’ | 720’ | cO €0° | ZO’ | zo’ | 10° | 10° | SEO" |SE0" | TO" £0 00¢ s£0' | 20” 0 Se0' | 91 | F0' | $€0° | CO” ¢0° | $0" | 90° | SEO" | ZO" | £0°| vO} TE | O On e0n = 0 OST EE WO) op hie ¢8T° | S€0°| cO’ | €0°| ce 90° 0 0 | $0O°| 10 0 \se0'| O 0 Om e20e 0 OOT 10°} O 0 0 O SO) ) 10) FO") 0 10 OF On esl a ell 0: 0 OS cO | O 0 O | 0 OV 10°; O 0 0 0 CO” 10 | O | co Oi ao y vl | CO Wo 0 0 0 0 0 Sc COTa One a PO 0) s(S9h: Olea 70: | SG0G5) 70) 0 Of ory | SE ae AO ISO ony) 0 0 0 |90' | O 0 esol | soot | Se6r | Seor | Seer | CLOT | SE6T | EL6T | EE6T | CGT | SE6T | SCOT | CLOT eset | €f61 | Se6t | C61 | EL6T | E61 | EET | E€6T | ELT SIo}yoUr t-S 1-S | O€-— | 6c-F | L7-p | St-P | SI-F| SP | VP tT G-¥ | 67-€ | €C-E | ST-E€ | BE €-€ | O@-7Z | 61-7 | 61-2} LT-c | ST-@ | VI? ur y3deq 909T | 86ST | 68ST | 88ST | TST | ZLST | 8SST | 6ST | CPST | THST | VEST | Hest 60ST | OST | LOPE | Z8PT | SZFT | LLPT | SLOT | LPT | 691 | LOFT sUOI}24$ “1op] Jed su0je UIeIZOIOW Ul SUOTZEI]UGIUOS SYIVN “WDIqqisD) ay1 surpnjour ‘ppnusag fo yjnos suouDis X] Fav iL | — PHOSPHATE -0 | -PHOSPHATE-0 | 10 I5-NITRATE-0 5 10 I5-NITRATE-o - NITRITE- 0 | - NITRITE - 0 a oe Uv -—-- - —_—-——-—--- =F pas 2098 ae tres 50- Ir22 \P 100 - { 100- \ \ \ \ \ \ P! | 1 1 H 1 NITRITE i \ NITRATE (N) y PHOSPHATE (P) ---- / Fic. 9. Relation of nitrite to nitrate and phosphate. Gulf of Maine at different times of year. of stations as follows: Various stations in the Ordinates, depths in meters. Positions 150 OCCURRENCE OF NITRITE IN THE SEA 151 Sta- Sta- ae Position Hon Position ee Position 1896 | 42°48’N 67°15’W 1910} 42°43’N 69°15’W | 1918! 43°39’N 66°39’W 1884 | 43°00’N 68°37’W 1908) 42°10’N 69°53’W | 1933} 41°25’N 69°02’W 2037 | 43°58’N 68°03’W 2062| 42°44’N 68°44’W | 2099] 42°22’N 68°32’W 2031 43°18’N 69°23’W 2063) 42°31’N 68°31’W | 2098] 42°50’N 67°19’W — SCALE — Fic. 10. Relation of nitrite to nitrate and phosphate. Gulf of Maine in the summer. NITRITE NITRATE (N) PHOSPHATE (P). Various stations in the (The origins for nitrite and nitrate coincide, but the phosphate curves are arbitrarily displaced to the left, to facilitate comparison of all three.) Ordinates, depths in meters. 2 NORRIS W. RAKESTRAW afforded by Schreiber (1927), Cooper (1933), ZoBell (1935) and others. But the Gulf of Maine is a rather specialized area; let us turn to the open ocean. Similar relations between nitrite, nitrate and phos- phate are shown in Fig. 11 for a number of deep sea stations from Latitude 12 to 42 N. Most of these show a clearly-defined nitrite maximum, at about 100 meters, but all show nitrite at the surface, except the three northern stations, which were visited during the summer. At most of them there is a small but distinct amount of nitrite below 150- 200 meters. Two stations, 1482 and 1491, are somewhat exceptional. At these, nitrite concentrations are rather low, but extend further down, with no well-defined maximum. The fertility of a region is obviously concerned in nitrite production and these stations are on the edge of the barren Sargasso Sea. Other stations in the Sargasso, as shown in Table IX, are largely devoid of nitrite. Nearly all the nitrite formation takes place well above the level of maximum nitrate and phosphate concentration, which is found at about 800 meters south of the Gulf Stream and 400 meters north of it. Density and Temperature Figure 12 shows the relation between nitrite, density and temperature at stations in Location 9 at various times of year. It can readily be seen that : 1. As the inflection point in the density curve moves further down nitrite moves down also. This is another clear evidence that nitrite is formed near the surface and diffuses or mixes downward, more com- pletely the more unstable the water column. 2. The production of nitrite goes on even at low winter temperatures, but with the rise in temperature nitrite is utilized as diatoms begin to grow. 3. The occurrence of a sharp nitrite maximum coincides with a high density gradient. If the depletion of nitrite above this is the result of diatom activity, this is localized and intensified by strong stratification. Ammoma Unfortunately, it was not, possible to include simultaneous determina- tions of ammonia with those of nitrite and nitrate throughout this study. A few such determinations were made, however, using the method de- scribed by Krogh (1934). The results are shown in Table X, cor- related with nitrite and nitrate. The low values for all three at. the surface of Station 1739, and their increase toward the bottom, lend support to the view that they are all utilized at the surface and that OCCURRENCE OF NITRITE IN THE SEA 153 ammonia is oxidized to nitrite and nitrate below. The increase of am- monia at the surface of Station 1742 (in Location 22) in September would seem to indicate that organic decomposition is going on more rapidly at this time, and ammonia being liberated, but that the second TABLE X Relation of nitrite to ammonia and nitrate. Concentrations in microgram atoms per liter. Station Depth in meters Nitrite Ammonia Nitrate 77 01 A 1.1 1734 340 0 3 4.6 855 0 5) 19.5 7/28/33 2650 (0) 6 20 3280 (0) 4 S725) (0) +8) 20.5 4180 (0) 6 23 0 0 Me? ileal 10 0 1.1 1739 20 02 1.4 DS 30 44 4.0 7/31/33 40 39 4.3 Si 50 36 Soll 60 24 4.1 Hell 75 36 4.3 6.4 0 0 — 1.1 10 0 2.8 0 1742 20 085 3.0 Boll 30 15 Soe 4.3 9/3/33 40 165 oe 6.1 50 18 3.8 eS 60 185 4.2 Tail 75 185 — fle) Dai 15 .04 8 Sg 200 05 0 10 2313 15 02 8 6.4 2314 15 05 if 2.8 2315 5 .08 7 1.4 2316 5 .065 0 1.8 2317 5 .065 3 od stage—appearance of nitrite—has not yet been reached. Or it may be that late in the summer, although nitrite is still being utilized, the am- monia is not drawn upon as a nutrient source. This was Cooper’s assumption on finding the same condition in the English Channel. 154 NORRIS W. RAKESTRAW At the various depths at Station 1734 the ammonia and nitrite are practically constant throughout, but the nitrite varies widely. From Sta- tion 2312 to 2317 the nitrate progressively diminishes, but.both nitrite and ammonia are approximately constant. In fact, throughout all the data shown the nitrite is more closely related to ammonia than to nitrate. Unpublished data in the hands of Dr. A. C. Redfield, on the dis- tribution of ammonia in the Gulf of Maine in May and June, show a distinct tendency for ammonia to concentrate between 20 and 50 meters in a maximum zone similar to that for nitrite. NITRITE CHANGES IN STORED SAMPLES A number of previous investigators have tried to determine what happens to the nitrite in samples on standing. The results are alto- gether irregular and confusing. Buch (1923) reported the formation of as much as 28 mg./cu.m. of nitrite-N in two days, in samples standing in the dark. Harvey noted the appearance of nitrite in a few cases. Orr found increases in nitrite in diatom cultures, even after the diatoms had stopped growing. Braarud and Klem cite some experiments on nitrate reduction, but are not specific as to the relation of nitrite. Cooper followed 44 samples for 10-13 days, to some of which nitrite and am- monia had been added. He reports: “ In all cases the changes in nitrite were small, little more than the experimental error.” Keys, Christensen and Krogh (1935) followed the ammonia (and in a few cases nitrite, nitrate and oxygen) in samples stored for several days. Results were so variable that they finally conclude: “It appears to us that the only conclusion that can be drawn with regard to nitrogen metabolism in stored sea water is that there are usually a number of forces at work, some operating to reduce the free ammonia, others tending to release ammonia; the observed ammonia will always be a resultant of these forces enue . Although with little hope of clearing up this confusion, a number of similar determinations on stored samples were made. The first of these were simply repeated analyses of the ordinary routine station samples made after a number of days standing. No particular precau- tions were taken to keep the samples in the dark, but they were exposed to probably less than normal illumination. Nearly 300 samples were followed in this way, with the results shown in Table XI. One section of the table concerns samples which were originally “ nitrite-free,’ or in which the initial concentration was not greater than 01. These are separated into three periods of standing, approximately 1, 2 and 3 weeks. Those which showed a significant increase in nitrite OCCURRENCE OF NITRITE IN THE SEA SS) in that time (greater than .035 microgram atom per liter) amounted to ie peenccnmel2nner cenmeaide ss per centiyoretme totals. Whese) 162 samples include many from the surface, with low nitrate, and many from deep levels, with high nitrate. There was no difference in the behavior of these two classes, a fact which has a bearing on the question of reduction of nitrate to nitrite. The other group of 178 samples contained initial concentrations of nitrite from .O1 to .91. Changes took place in both directions. But relatively fewer cases of increase, and more of decrease, were found AAR EE Ol Nitrate changes in stored samples A. Samples with initial concentration = 0 (<.01 microgram atoms per liter) 5-7 12-15 | 20-24 days days days Motallemuinmberteuee tei eer re ans eyed cieestayaissrn yah e cis isiesntdloners 109 41 12 No. increased >.035 microgram atom per liter.............. 12 5 4 ReRCembA gC ey Oley OvallUwueey eats Herve lala orci vals ay idan eens cnetgetene 11% | 12% | 33% B. Samples with initial concentration >0 (.01-.91 microgram atom per liter) 5-7 12-15 | 20-24 days days days so aleriuimilne reer). Paesenea ence srtn ent: uence ke on oe wah 119 25 34 No change (Increase or decrease <10 per cent of initial concen- ERATION) eam eee eee ete at nel i prises Stuer dees ot Me Woes 49 9 9 Number increased, over initial concentration, 10-50 per cent. .| 18 1 1 Number increased, over initial concentration, >50 per cent...| 11 1 0 Number decreased, of initial concentration, 10-50 per cent... .| 31 14 17 Number decreased, of initial concentration, >50 per cent..... 10 0 7 Percentage of total number which increased more than 10 per cent of initial concentration........................... 24% | 8% | 3% Percentage of total number which decreased more than 10 per cent of initial concentration....................000--- 34% | 56% | 70% as the time of standing lengthened. The predominating tendency under these conditions is for nitrite, if initially present, to disappear on stand- ing. If initially absent, it may appear in small amounts. The results of these experiments do not lead to a conclusion much more definite than that of Keys, Christensen and Krogh. Subsequently, however, several series of more carefully controlled experiments were carried out. Samples of filtered surface water were placed in Pyrex flasks. Some of these contained added amounts of ammonia-nitrogen, others added nitrate. Half the samples were steri- 156 NORRIS W. RAKESTRAW lized in an autoclave, and of the total number, half were stored in the dark and half were placed in the sunlight as much as possible during the period of storage. Analyses for nitrite were made at intervals up to 19 days. Two series of experiments of this sort were carried out, with the following results: 1. The samples stored in the dark, whether sterile or not, showed no significant change in nitrite, even in the presence of added nitrate or ammonia. 2. Those kept in the light invariably showed a change in nitrite. If it was high at the beginning it decreased, but not so sharply in the presence of excess nitrate. If the nitrite concentration was low at the start it increased slightly, but more so in the presence of added nitrate. In fact, after 19 days the nitrite reached the same value, whether by increase or decrease. This seems to depend, however, upon the pres- ence of sufficient nitrate, and is independent of the concentration of ammonia. The conditions surrounding the attainment of this apparent “steady state’ will be the subject of further investigation. 3. The rate of disappearance of nitrite was unaffected by the pres- ence of excess ammonia, and only a very slight increase was found in samples of low original nitrite content. In view of the conditions of these experiments this result is not necessarily in conflict with those of ZoBell (1933), who reported the photochemical oxidation of ammonia to nitrite under certain conditions. 4. An effort to follow the changes in nitrate led to inconclusive results. In one series of experiments the sum of nitrite and nitrate (determined by Harvey’s “ nitrate” method, using reduced strychnine as a color reagent) remained constant, although the concentration of nitrite alone decreased considerably; but in another series the increase in nitrate did not compensate for the disappearance of nitrite. Thiele (1907) was apparently the first to show that nitrate can be reduced to nitrite by ultraviolet light. Moore (1918) found that sun- light can accomplish the same result, but that the shorter wave-lengths are the most active. More recently, Villars (1927) has studied the photochemical decomposition of KNO, in considerable detail and has shown that wave-lengths less than 3,000 Angstroms are effective in the reduction. About 1 per cent of the total energy of the solar spectrum at sea level falls within this range. 6 NITRITE IN THE SEA BoTTrom It is evident that at shallow stations, at least, a great part of the de- composition of organic matter takes place on the bottom. A study of the nitrite content of bottom sediments might therefore be significant OCCURRENCE OF NITRITE IN THE SEA SV Brandt gives figures for the ammonia and albuminoid-nitrogen con- tent of “bottom water’ (that in contact with the bottom surface) and of “interstitial water’’ (that contained in the bottom cores brought up) showing from three to five times as much in the latter as in the former. This relation was apparently independent of depth, for the bottom samples varied from 45 to 3,000 meters. He also cites de- terminations on muds from the North Sea showing 2-4 mg. of am- monia-N and 5-10 mg. of nitrate-N per kilogram of moist mud. A number of bottom samples were obtained from various sources and determinations made of the nitrite and nitrate in the interstitial and adjacent water. The results are contained in Table XII. Figures are given for the nitrite and nitrate not only in the bottom samples, but also in the lowest water bottle, which will vary from 2 to 20 meters from the bottom. Care must be taken in comparing the results, for different methods were used in obtaining and treating the samples. In some cases, the interstitial water was removed from the whole mud core by suction and the nitrite concentration expressed in terms of the volume of water thus recovered. In other cases, samples of mud and the water in con- tact with it were collected, either in the coring device, the Pettersson grab, or a special mechanism which sucks up a sample of mud and water off the bottom. Nitrite was determined in this water, after filtration. Although the data from deep stations are fragmentary they are sufficient to show that conditions in deep bottoms are different from those in shallow ones. The single determination at Station 1731 is not quantitative, but it at least shows that there is very little nitrite in the mud, compared with Stations 1331, 2220 and 2227, where similar technic was used. The condition of the sample from Station 1734 is doubtful, but it should approximate those from A-8 and A-25. Nevertheless, the content of nitrite is very low, and the nitrate is very little greater than that in the lowest water-bottle. The results from Stations 2220, 2224 and 2227 are interesting in showing the enormous quantities of nitrite in the interstitial water of shallow bottoms. A comparison of these quantities with those in the supernatant “bottom water ” shows that the rate of nitrite formation in the mud is very much higher than its rate of diffusion into the surrounding water. _ At Station A-8 and again at A-25 two samples were obtained as close together as possible. The variation in their nitrite and nitrate content is partly due to the method of sampling, which affords a mixed sample of mud and water sucked off the bottom. Nitrite was deter- ee mined in such a sample after filtration, and although this contains “ in- 158 NORRIS W. RAKESTRAW TABLE XII Nitrite and nitrate in the sea bottom. Concentration in microgram atoms per liter. In lowest In bottom sa an ESA pe Bemis Method of obtaining sample meter for analysis Nitrate | Nitrite |Nitrate| Nitrite 1734 | 4630 20.52 | O? 23.01 .05 | Water-bottle closed on bottom; brought up mud and water. 1731 | 3840 .04 | Portion of core extracted with sur- face water. 1336 65 5.0 | .28 25.0 1.8 Water brought up with sand in Pettersson grab. 1331 200 115 .63 | Portion of core extracted with sur- face water. 2220 218 .04 .065 | Supernatant ‘‘bottom water’’ on top of core. 2220 218 04 28.38 Interstitial water removed from core by suction. 2224 195 0 .035 | Supernatant ‘‘bottem water’ on top of core. 2224 195 0 3.3 | Interstitial water removed from core by suction. 2227 210 01 .035 | Supernatant ‘‘bottom water’’ on top of core. 2227 210 01 10.84 Interstitial water removed from core by suction. A-8* 32 |A 235 4.6 .26 | The sampling device sucks up a B 235 | 10.3 7A sample of mud from the surface eT —= of the bottom, along with water in A-25* 45 |A 10.3 | .13 7D 20 and around it. Contains inter- B 10.3 | .13 24.6 WS) stitial and bottom water in vary- ing proportions. Near I5|A .065 K-3+ B alee. C 285 Near IS|/A 715 K-3* B 17 * Two samples, A and B, taken as close together as possible. + Three samples, A, B, C, taken as close together as possible. 1 Ammonia-nitrogen concentration, 0.65. 2 Ammonia-nitrogen concentration, 0.35. 3 Total extractable nitrite-N, 1.5 mgm. per kgm. of dry mud. 4 Total extractable nitrite-N, .25 mgm. per kgm. of dry mud. ta OCCURRENCE OF NITRIDE INTHE, Si 159 terstitial”” and “ bottom” water in varying proportions the irregularity of results would nevertheless seem to indicate great local differences in bottom conditions. Similarly, near Station K-3, three samples were taken simultaneously at one place and two at another. The first three were quite different in nitrite content, while the latter two were practically identical. The general conclusion to be drawn is that the decomposition of organic matter on the sea bottom involves oxidation of nitrogen com- pounds to nitrite and-nitrate, and that this process is much more intense on shallow bottoms than deep. Nearly all of the nitrification must have taken place before organic matter reached the bottom in the latter cases. The samples from near Station K-3 were analyzed again after hav- ing stood for one day. The nitrite in three had approximately doubled ; in the fourth it had increased 50 per cent. The water from the fifth mud sample had been separated from the mud at the time of the first analysis, and the nitrite in this water had decreased 50 per cent after one day. Evidently nitrite is formed rapidly and continuously in the mud, diffusing into the water in contact with it. In the water, how- ever, the nitrite is rapidly destroyed, presumably by oxidation to nitrate. Samples of white coral mud from Castle Harbor, Bermuda, at a depth of about 5 meters, were extracted by distilled water and surface sea water, respectively. The distilled water extract contained: nitrate, 85; nitrite, 0. The surface water extract contained: nitrate, 15; nitrite, .085. The surface water itself was nitrite-free and contained . microgram atom per liter of nitrate-N. The more efficient extrac- tion by sea water shows the same phenomenon of “ base-exchange’”’ so common in the extraction of soils. NITRITE AT THE SURFACE The question of the occurrence of nitrite in the immediate surface layer seems to merit particular consideration. The first intimation of irregularity occurred on the section from Cape Hatteras to Bermuda (see Fig. 4). At Stations 1462 and 1463 suspiciously high surface values were obtained. These were traced, apparently, to contamination, and were finally rejected. But later, at Station 1465, there could be no doubt of a fairly high concentration at the surface, with very little immediately below. However, no sur- face nitrite could be found in the vicinity of Bermuda or at the im- mediately following stations, except 1469. Between this station and 1472 surface samples were picked up every hour or two, while the ship was in motion, without finding more than a trace of nitrite. ‘The same was true of a number of samples taken up between Stations 1473 and 160 NORRIS W. RAKESTRAW 1475. Much later, six samples taken in the same way between Stations 1730 and 1731 were nitrite-free. Throughout the Caribbean cruise nitrite was found at the surface at the majority of the regular stations, but at four places where surface samples alone were taken none was found. In all, out of thirty sur- face samples taken with the ship in motion, detectable traces of nitrite -SCALE- MICROGRAM-ATOMS PER LITER NITRITE NI ATES (Nee Sue Oats PHOSPHATE (P) ———-—___+-__1 Fic. 11. Relation of nitrite to nitrate and phosphate, at deep stations. Ordi- nates, depths in meters. were present in only six, and only once a concentration of more than 01. Whether the occurrence of considerable amounts of surface nitrite in the Caribbean is constant or limited to a certain time of the year (as in the north) cannot be said from the results of this one cruise, but the strange irregularity suggests that special factors—perhaps photo- chemical—may operate in the immediate surface, a matter which will later be investigated more fully. Highly irregular variations in sur- OCCURRENCE OF NITRITE IN THE SEA 161 face nitrite have also been reported by Matsudaira and Yasui between Nagasaki and Shanghai. The vicinity of Bermuda is an area especially devoid of nitrite and other “nutrients.” A number of surface samples were collected from the harbors and bays of the Bermuda Islands, and even from their fresh water streams and ponds, with no more than traces of nitrite, and mini- mal amounts of nitrate. SEASONAL CHANGES IN NITRITE Figures 5, 6, and 7 contain the data from Locations 3, 5, 7, 8, 15 and 22 in the Gulf of Maine, arranged to show seasonal variations of nitrite (0) : 50 ! ; 100 ' i ! t] ! B H ! ' t 150 H ! H i 1 i] 200 (715 1917 4 I-10 8 VI-9 IX XI- NITRITE 25 -DENSITY— 27 (03) SCALE - So. 096 2 2189 2289 IX 22 W-25 W-22 Y-27 Location 9, Fic. 12. Relation of nitrite to temperature and density (¢,) Gulf of Maine, at various times of year. Ordinates, depths in meters. my) ro) rs is and nitrate. Figure 12 shows the progressive changes at Location 9 in a different way. The following points are revealed : 1. A minimum of nitrite occurs in January and February, concen- trated at the surface. This corresponds to the period of maximum ni- trate and phosphate, when the water mass is nearly homogeneous. Nitrification is going on at the surface, but only very slowly. 162 NORRIS W. RAKESTRAW 2. There follows a slow increase in nitrite until April, when more | rapid production accompanies the diatom outburst and continues during the whole period of growth. 3. Simultaneously with the growth of diatoms, however, the removal of nitrate from the surface begins, and nitrite also starts to disappear in the same regions. At Locations 3 and 5 the removal of nitrite lags somewhat behind the nitrate, possibly because the latter is utilized more readily. At Location 7, during July, August and September of the second year, considerable nitrite is still present at the surface, but since the nitrate has not yet been completely used up it would seem that the nitrite store has not been heavily drawn upon. 4. The maximum of nitrite occurs between June and September, varying with local conditions at different stations. There is evidence that the seasons are somewhat displaced in different parts of the Gulf of Maine and that the growth of organisms does not take place uni- formly throughout. 5. The utilization of nitrite in the surface layers, simultaneously with its production throughout a much deeper water mass, results in the appearance of a maximum zone, generally between 30 and 50 meters. 6. The occurrence of considerable nitrite in the surface layers at Locations 3, 8 and 15 late in October is noteworthy. This did not occur in all parts of the Gulf but does not seem to be fortuitous. It may well be associated with the death and disintegration of organisms after the close of the growth period. Soot-Ryen reported the same thing in his study of one of the Norwegian fjords. It is possible that this phenomenon was missed at Locations 5 and 7 from lack of data. 7. Location 22, being shallow, shows only the upper part of the picture. Maximum nitrite production has taken place here a little later. The results of Soot-Ryen are altogether similar to these, with a slight displacement of the seasons. He found a minimum of nitrite in January, increasing thrcugh the spring and early summer to a max- imum in August, but with depletion of the surface during the summer. This was followed by another short period of nitrite production in the upper layers in November. ’ Verjbinskaya’s data on seasonal variation of nitrite in the Barents Sea.are somewhat incomplete, but as far as they go they indicate the same general sequence: nitrite-free water in December and January, building up to a maximum in August, but with surface depletion during the summer. Cooper’s very complete study in the English Channel deals with a station too shallow to show vertical distribution of nitrite, but in all other respects is in complete accord with the above findings. It shows OCCURRENCE OF NITRITE IN THE SEA 163 the same simultaneous removal of nitrite and nitrate from the surface in the summer, followed by rapid production of nitrite and ammonia in October, and a minimum during December and January. No study of seasonal variations at a truly oceanic station seems ever to have been made. Figure 8 shows the nitrite concentration in the upper 1,000 meters of the four corresponding stations in Group 3 (Table V). The only feature in common with the data from shallower stations is the surface depletion during the summer. Nitrite production evidently takes place in the upper levels during the winter. The hydro- graphic and biological conditions are so different here that a much more complete investigation is necessary. THE Position oF NITRITE IN THE NITROGEN CYCLE IN THE SEA Brandt, in his monumental work on nitrogen metabolism in the sea, summarized all the earlier results and finally came to the conclusion that ammonium salts are oxidized to nitrite and nitrate by nitrifying bacteria in the sea bottom. Impressed by the fact that such organisms had never been found in the water itself, but that nitrite was frequently present at the surface, he proposed that bacteria might oxidize dissolved organic matter directly to nitrite and nitrate without ammonia as a necessary intermediate. One view of the significance of nitrite was that expressed by Atkins: that its presence in sea water indicates the production of nitrate. Its absence from deep oceanic samples, on the other hand, shows that nitri- fication is complete. This assumes, of course, that nitrite is solely pro- duced by oxidation of ammonia, or perhaps from dissolved organic matter directly. Others, such as Brujewicz (1931) and Braarud and Klem (1931), have rejected this view and incline to that which explains nitrite forma- tion as a result of the reduction of nitrate. Either theory will explain many of the results. Organic matter in the zone of nitrite production, present in considerable amounts, may either be the direct source of nitrite, with ammonia as a likely intermediate, or it may serve as the necessary source of energy for the nitrate-reducing bacteria. These have, in fact, been found there in small numbers. The presence or absence of the necessary organisms is a bacteriologi- cal problem which this present paper cannot go into, but there are a number of considerations from the observed chemical data which favor oxidation as a source of nitrite, rather than reduction. First, it is unlikely that the upper layers of the water, where oxygen tension is high, would be a suitable environment for reduction processes 164 NORRIS W. RAKESTRAW to take place. The bacterial reduction of nitrate to nitrite is known to be possible under nearly anzerobic conditions, and in the presence of enormous quantities of organic matter, but it is not certain that the same thing would take place in surface sea water. The relation of nitrite and nitrate in bottom muds to each other and to the surrounding water, mentioned above, makes it clear that both are being formed in the bottom, obviously by oxidation of organic matter. Since nitrite is an intermediate in this process, it is scarcely possible that the large amounts found here could have been produced by a subsequent reduction of the nitrate; and if the nitrite in shallow muds is the result of oxidation it is very likely that all in the water above is of the same origin, for there is no evidence of any discontinuity of chemical mech- anism from top to bottom. A comparison of the profiles of stations of varying depths suggests that the concentration of nitrite depends on a balance between the rate of chemical transformation and the rate of descent of the matter being transformed. The relation between ammonia and nitrite which Cooper’s results bring out, as well as the data presented above, especially the increase of both immediately after the summer growth period, are best explained on the assumption of oxidation of the former. The correspondence of the intermediate maximum zone of both nitrite and ammonia, already referred to, supports this view. The general correlation of low concentrations of both ammonia and nitrite in the deep sea, and of higher concentrations of both, frequently, in other regions, is of some significance. Furthermore, dissolved organic matter is nearly as abundant at great depths as near the surface, and nitrate is present in large quantity. This would seem to be an optimum condition for reduction of nitrate; but nitrite is seldom found here. It may, of course, be true that dissolved forms of organic matter are unsuitable for nitrate-reducing bacteria. Finally, the following theory is presented to explain the observed facts: During the winter, when biological activity is at a minimum, the decomposition of organic matter is going on very slowly. This takes place largely near the surface, either as the result of oxidation of ex- cretory products liberated here, or the death of organisms, with dis- integration and decay beginning before they ave sunk far. Quanti- tatively, this same process goes on with greater rapidity as the season progresses. Nitrite is an early stage in the oxidation of this organic matter, presumably after the liberation of ammonia as a previous step. This is in keeping with Wattenberg’s (1930) conclusion that ammonia formation is mainly a surface phenomenon. Later in the winter and OCCURRENCE OF ENTREE IN VEE SrA 165 early spring, as the unstable column becomes deeper, these processes take place through a deeper vertical range. Still later, when diatoms begin to grow actively they utilize the nitrite in the surface water, along with nitrate. Occasionally, when there are plenty of other nutrients, nitrite is not exhausted. Such is the case in the Antarctic, where determinations made on the “ Dis- covery’ show high concentrations of nitrite above 100 meters during the summer, but associated with very large quantities of nitrate and phosphate. Organic matter works its way downward, by settling, if it is par- ticulate, or by diffusion or mixing, if dissolved. The appearance of nitrite by oxidation therefore extends below the zone of its utilization. At greater depths it disappears again by further oxidation to nitrate. Nitrate concentration continues to increase to still greater depths, “reaching a maximum at 800-1,000 meters, generally. Nitrification ex- tends further than the occurrence of nitrite would indicate. The decomposition of organic matter takes place in several steps, and we may suppose that the products of the first stages are more rapidly oxidizable than those of the later ones. This may be the result of either the chemical nature of the material or its concentration, which would be greater at the start of the process, before it has become diluted. The formation of nitrite in significant amount is associated with the initial stages of the decomposition of organic matter, probably becatise of their greater rapidity, but the whole process of nitrification does not complete itself until greater depths are reached, and by this time the nitrite, formed in the first part of the process, has been further oxidized to nitrate. Vertical differences in temperature may also play a part in deter- mining the rates of decomposition. Recently, ZoBell (1935) has suggested that diatoms may reduce nitrate to nitrite extracellularly and then absorb it as such. Such a theory would explain many of the observed phenomena with respect to nitrite distribution, to be sure, particularly its close association with phytoplankton. Its acceptance, however, will depend upon its satis- factory explanation of the relation of nitrite to ammonia. Such data as we now have covering this point seem difficult to reconcile with this view, but until we have had an opportunity to extend them we prefer to withold judgment. ACKNOWLEDGMENT I should like to express my thanks to Doctors H. P. Smith and H. E. Mahncke and Mr. E. F. Beach, who carried out some of the 166 NORRIS W. RAKESTRAW analytical work, as well as to Doctors August Krogh and Ancel Keys, who furnished some of the ammonia results. LITERATURE CHED Arxtns, W. R. G., 1930. Seasonal changes in the nitrite content of sea water. Jour. Mar. Biol. Ass., 16: 515. Bounecke, G., E. HentscHer anp H. Warrenserc, 1930. Uber die hydro- graphischen, chemischen und biologischen Verhaltnisse an der Meeresober- flache zwischen Island und Gronland. Ann. d. Hydrog. und Marit. Meterrol., 58: 233. Braarup, T., AND A. Kiem, 1931. Hydrographical and chemical investigations in the coastal waters off More and in the Romsdalsfjord. Hvalradets Skrifter, No. 1, p. 48. Branopt, K., 1916-20. Uber den Stoffwechsel im Meere. Wiss. Meeresuntersuch., Kiel, 18: 185. Branpt, K., 1927. Stickstoffverbindungen im Meere. Wiss. Meeresuntersuch., Kiel, 20: 201. Brujewicz, S. W., 1931. Hydrographic Work of the State’s Oceanographic Insti- ° tute of U. S. S. R. in the Barents Sea in 1927-30. Rep. First Session, State Ocean. Inst., No. 1. Bucu, K., 1923. Methodisches tiber die Bestimmung von Stickstoffverbindungen im Wasser. Merentutkimuslaitoksen Julkaisu, No. 18:.21. Coorer, L. H. N., 1933. Chemical constituents of biological importance in the English Channel, November, 1930, to January, 1932. Pt. 1. Jour. Mar. Biol. Ass., 18: 677. Discovery Committee, 1932. Station List. Discovery Reports, 5: No. 1. Harvey, H. W., 1926. Nitrate in the sea. Jour. Mar. Biol. Ass., 14: 71. Keys, A., E. H. Curistensen anv A. Krocu, 1935. The organic metabolism of sea water, with special reference to the ultimate food cycle in the sea. Jour. Mar. Biol. Ass., 20: 181. Krocu, A., 1934. A method for the determination of ammonia in water and air. Biol. Bull., 67: 126. Matsupaira, Y., AND Z. Yasut, 1932-33. Surface observations between Nagasaki and Shanghai. Jour. of Oceanog. (Kobe), 4: 473; 5: 439. Moors, B., 1918. The formation of nitrites from nitrates in aqueous solution by the action of sunlight, and the assimilation of the nitrites by green leaves in sunlight. Proc. Roy. Soc., B, 90: 158. Orr, A. P., 1926. The nitrite content of sea-water. Jour. Mar. Biol. Ass., 14: 55. Orton, J. H., 1924. An account of investigations into the cause or causes of the unusual mortality among oysters in English oyster beds during 1920 and 1921. Fisheries Investigations (Ministry of Agric. and Fisheries), Ser. 2, 6: No. 3, 130: Rasen, E., 1914, 1918. Uber quantitative Bestimmungen von Stickstoffverbind- ungen im Meerwasser und Boden, etc. Wiss. Meeresuntersuch., Kiel, 16: 2071S Ie Raxestraw, N. W., 1932. Phosphorus and nitrogen in the neritic waters of the Gulf of Maine. Int. Rev. Hydrobiol. und Hydrogr., 27: 151. Raxestraw, N. W., 1933. Studies on the biology and chemistry of the Gulf of Maine. I. Biol. Bull., 64: 149. : Rosinson, R. J., anp H. E. Wirtu, 1934. Report on the free ammonia, albuminoid nitrogen and organic nitrogen in the waters of the Puget Sound area, during the summers of 1931 and 1932. Jour. du Conseil, 9: 15. Rosinson, R. J., anp H. E. Wirt, 1934. Free ammonia, albuminoid nitrogen and organic nitrogen in the waters of the Pacific Ocean off the coasts of Wash- ington and Vancouver Island. Jour. du Conseil, 9: 187. OCCURRENCE OF NITRITE IN THE SEA 167 ScurerBer, E., 1927. Die Reinkultur von marinem Phytoplankton und deren Bedeutung fiir die Erforschung der Produktionsfahigkeit des Meerwassers. Wiss. Meeresuntersuch., Helgoland, N. F., 16: No. 10. Soot-Ryen, T., 1932. Changes in the nitrite content through the year in North- Norwegian Fjords. Jour. du Conseil, 7: 246. Tuete, H., 1907. Einige Reaktionen im ultravioletten Lichte. Ber. deutsch. chem. Ges., 40: 4914. VERJBINSKAYA, N., 1932. Observations on the nitrite changes in the Barents Sea. Jour. du Conseil, 7: 47. Vittars, D. S., 1927. The photolysis of potassium nitrate. Jour. Amer. Chem. Soc., 49: 326. ZoBELL, C. E., 1933. Photochemical nitrification in sea water. Science, 77: 27. ZoBELL, C. E., 1935. The assimilation of ammonium nitrogen by Nitzschia closte- rium and other marine phytoplankton. Proc. Nat. Acad. Sci., 21: 517. Mie BIOLOGY (OF OMEIONA Ss UMENE Sian ea rl eyes Gr Oiler MAINE AND BAY OF FUNDY CHAREST Eis (From the Woods Hole Oceanographic Institution 1) INTRODUCTION Although Oithona similis, one of the two microcalanids found breed- ing in the Gulf of Maine, is small in size, the relative abundance of larvee indicates that at times it forms a significant factor in the natural economy of the region.?_ It is also particularly favorable for quantitative study because adults, larve, and egg cases all appear to be sampled representatively with fine nets (No. 10 silk) and the pump. This paper forms the third of a series on the biology of zooplankton species in the Gulf of Maine and Bay of Fundy with special reference to production and dispersal, and is based on material obtained between July 28, 1931 and September 29, 1932, in the course of investigations carried on for the International Passamaquoddy Fisheries Commission. For a description of the area covered, location of stations, and methods, the reader is referred to a previous report (Fish, 1936c). REGIONAL DISTRIBUTION OF THE ADULT STOCK Oithona similis is almost “world wide” in distribution (Bigelow, 1926), but on the American coast’ appears to be largely centered north of Cape Cod. South from this point it is gradually replaced, in neritic waters at least, by O. brevicornis, a species not yet reported north of the Cape. At Woods Hole in 1922 and 1923 (Fish, 1925, Fig. 47) the latter species occurred with southern forms during the summer from late June to mid-October, and O. similis with the boreal population in autumn and winter. Farther south in Chesapeake Bay O. similis is decidedly a winter form limited largely to the inner Bay and is ex- ceeded by O. brevicornis, which occurs everywhere in that region throughout the year (Wilson, 1932). North of Cape Cod MecMurrich (unpublished data) found O. similis in small numbers in Passamaquoddy Bay regularly during the colder months from November until March, but very rarely during the spring 1 Contribution No. 109. 2 Microsetella norvegica, the second species, is far less important numerically. 168 BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 169 and summer. Elsewhere in Canadian waters, however, it has been reported widespread offshore during the summer (Herdman, Thompson, and Scott, 1898; T. Scott, 1905; Pinhey, 1926). In 1931-32 O. similis was taken to some extent throughout the Gulf of Maine and Bay of Fundy, the most noticeable changes in the quanti- tative distribution of the adult stock occurring with the vernal warming of the water. In April, 1932 it was most numerous (up to 5,423 per minute) in the western area of the Gulf with smaller and apparently more local centers of abundance in the Middle Channel, off Yarmouth, and off Halifax. In the outer basin of the Gulf it was almost totally absent in the hauls (Fig. 1). Fic. 1. Distribution of adults of Oithona similis in April, 1932. Number per minute of towing. The changes in distribution from April to May would suggest that Oithona tends to withdraw into deeper water in winter, and approaches the coast with rising temperatures in the spring. By the end of May the greatest number of adults (6,087 per minute) was found at the innermost station off Casco Bay where the surface temperature was 8.63° C., although they were abundant throughout the western area (6.08°-9.38° C.) and in the Bay of Fundy along the course of the entering drift. The richest hauls at this time were everywhere ob- 170 CHARLES Ji RISE tained at temperatures exceeding 6° C. O%thona was still sparse in the colder waters of the upper Bay, the Quoddy region, and thence westward along the coast of Maine to Frenchmans Bay. There was a marked increase in the adult stock between May and June but no appreciable changes in distribution. Again the richest hauls were made nearest Casco Bay, and the poorest in the New Bruns- wick and eastern Maine coastal areas. The adult stock reached its peak at this time in the western Gulf and, although the subsequent crop yielded larger numbers of larve in August (Table IV), few apparently reached maturity. East of Mt. Desert adults did not reach their peak until August and the maximum was of short duration for, in the Bay of Fundy at least, they had almost reached the April level by mid- September. PRODUCTION AND DISPERSAL OF EcGcs AND LARVz Spawning Areas Spawning is widespread but varies greatly in different parts of the region. Until July, in 1932, with the major portion of the Ozthona similis population located in the western part of the Gulf, the rapid decline to the eastward resembled that of Calanus finmarchicus (Fish, 1936a, Fig. 11). Only in the outer part of the Bay of Fundy was there a slight break in the descending curve, and here the greatest in- crease was in copepodite stages, probably transported for some distance from the western or outer Gulf (Fig. 3). This is indicated by the low values in the sections from the inner Bay (F and G) and in the path of the drift from the Bay along the coast as far west as Mt. Desert (£). The eastern region assumed greater importance in midsummer, and the curve for August (Fig. 3) more closely resembles that of Pseudo- calanus (Fish, 1936), Fig. 1) than Calanus. Extensive spawning con- tinued in the western area, but the large numbers of early and late nauplii entering the Bay of Fundy and eastern Gulf at this time are indicative of important propagation nearby. Early nauplii, up to 19,258 per cubic meter, in the Bay could hardly have originated beyond the eastern basin. This is also evident from the number of early larve at all offshore sta- tions east of Penobscot Bay in the Gulf (Fig. 9C). The Bay of Fundy itself, however, appears to be a relatively un- important spawning area. In the Quoddy region beyond the immediate influence of the drift from the Gulf, no eggs were found before July 30, and at all times both eggs and early stages were comparatively sparse, a condition also prevailing eastward along the New Brunswick coast and at the innermost stations (9, 104 and 114). Elsewhere in the BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 171 Fic. 2. Distribution of adults of Oithona similis in 1932. May—June: number per minute. August: number per cubic meter. APRIL—JULY | NUMBER PER MINUTE 1000 AUGUST - SEPTEMBER NUMBER PER CUBIC METER & fy : a w o/ G “ Fic. 3. Mean numbers of adults and young of O. similis in seven sections be- tween Casco Bay and Cape Spencer in 1932. Sections: A. Casco Bay (Stations 23A-25A, 26); B. Penobscot Bay (Stations 27-29); C. Mt. Desert (Stations 17, 30-31) ; D. Moose Peak (Stations 32-33) ; E. Passamaquoddy Bay (Stations 5-84, 35); F. Pt. Lipreau (Stations 13, 36-37) ; G. Cape Spencer (Stations 9-11). 172 BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE as) Bay they were frequently very abundant, with hauls in August yielding up to 35,970 early nauplii per cubic meter (Fig. 9C), but these were always in the path of the entering drift. Breeding Seasons Like the other dominant boreal copepods, Calanus finmarchicus and Pseudocalanus minutus, Oithona begins spawning in the Gulf of Maine in March and continues in a succession of generations until September. There is a seasonal delay of about one month in the appearance of eggs in the eastern Gulf and Bay of Fundy and this regional variation per- mits the distinguishing of two breeding stocks, a western and an eastern. Also, as in the case of the two above-mentioned species it must be borne in mind that these stocks of Otthona are not permanently restricted to their respective localities, but merely represent delayed maturation of that portion of the winter population which happens to be located in the eastern part of the region during the spring, and earlier maturation in areas to the westward where the water mass responds more rapidly to vernal warming. However, since the augmentation starts at sig- nificantly different times in the two regions, with substantially the same interval of development, subsequent broods of the two stocks continue to appear at distinct periods irrespective of where they are dispersed. Annual Cycle in the Two Breeding Stocks Very little has been published on the biology of this species. Gran (1902) found some indication that the development of small species like Oithona similis, Microsetella atlantica and Oncea conifera is com- pleted in a much shorter time than if larger species like Calanus fin- marchicus, C. hyperboreas, Metridia longa, and Eucheta norvegica. Ruud (1929) reported that the stock of O. helgolandica (similis) doubled in number off More in 1925 from the end of May to the end of July. In 1926 a small stock at the beginning of March increased steadily until May 12. A second increase took place later, reaching a maximum on July 20. In 1927 conditions were similar except that the second maximum attained the highest peak of the year on July 31. Bogorovy (1932) found copepodite stages IJ] and IV predominating in the Barent’s Sea at the end of July. In the Gulf of Maine and Bay of Fundy the maxima of different larval stages, following the appearance of the first brood of the stock maturing in the western Gulf in March, indicate a developmental period throughout the region of about two months during the early season, and about six weeks in summer. 174 (CISVAINIIOS) ja LES There appear to be at least three and possibly four broods of the western stock in March, May, July, and September. This stock, which in 1932 was represented by a maximum of late nauplius and copepodite APR 16 ST A2i3i8 W, Ww SEPT 16 ST IIA 1000 1000 E. E.N. U.N. C. ADE. E.N. LN. C. AD E—E ENULN C. ADE. EN LN. C. AD. Fic. 4. iline, 4): Fic. 4. Composition of the population of Oithona similis, in the outer part of the western area in 1932, showing the succession of generations of the western (WV) and eastern (E) stocks. April-June: number per minute. August-September : number per cubic meter. Stages recorded as, E-egg sacs, E.N.—early nauplius (I-III), L.N.—late nauplius (IV-VI), C—copepodites (I-V), 4.D.—adults. Fic. 5. Composition of the population of Oithona similis at Station 5 in the New Brunswick area of the Bay of Fundy in 1931 and 1932. Symbols as in Fig. 4. BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 75 stages on April 16, had by May 2 advanced largely to late copepodite stages and by May 31 matured and was spawning (Fig. 4). The sec- ond brood had reached late nauplius-copepodite stages by the end of June. The third brood, apparently developing more rapidly, was dom- inated by late nauplii in mid-August and what appeared to be a fourth brood was in egg and early nauplius stages on September 15-16 (Fig. 5). There are three broods indicated in the eastern stock. Eggs found between April 20—May 2 had presumably advanced to copepodite stages by the end of May (Fig. 4). A second brood of eggs appeared a month later, and a third brood by mid-August. That portion of the stock remaining on September 15-16 in the Bay was apparently repre- sented by late nauplius and early copepodite stages. TABLE | ’ Dominant stages of the eastern and western stocks of P. minutus and O. similis occurring at comparable periods in 1932. Western Stock Eastern Stock 1932 P. minutus O. similis P. minutus O. similis iNpralgllOne oe Late nauplius | Late nauplius- | Egg Egg copepodite May 19-22..... Late copepodite| Adult Late nauplius | Late nauplius Wleny Sil, sooseee Adult Ege Late nauplius- | Copepodite early copepo- dite June 20-31..... Late nauplius | Late nauplius- | Egg Adult-egg copepodite JlyasORPe ee Egg—early Early nauplius | Adult Adult nauplius August 11...... Latenauplius |Latenauplius- | Adult-egg Adult-egg copepodite September 15-16} Egg Ege—early Late nauplius | Late nauplius- nauplius copepodite The successive generations of the two breeding stocks of Oithona as shown in Table I and in Figs. 4 and 5 compare very closely with those of Pseudocalanus (Fish, 1936), Figs. 2-4). It will be noted that in the western stock Oithona in 1932 appeared to be slightly more advanced than Pseudocalanus at comparable periods throughout the summer, while in the eastern stock very little difference was detectable. Productivity and Mortality in Developing Larve Oithona similis was frequently taken during the breeding pericds with one or two attached ovisacs. However, by far the larger number 176 (CISUNIRIEIES) Jf. IMUSIaL was found floating free in the water. Whether or not the sac normally remains attached to the female for only a short period of time to insure fertilization, and then becomes detached to permit the formation of others, is not known. It is, of course, possible that those in the samples were broken off in the process of capture, but considering the large number of larve produced by a relatively small parent stock in August, several sacs of eggs must have been produced by each female at that season. Unlike Pseudocalanus, where the ovisac was never found around detached eggs, the ovisac of Ozthona remains intact during the period of incubation and no doubt serves to protect the eggs from small predatory forms until they have hatched, a condition also found in Microsetella. From April until the end of June in 1932 the stock of Oithona in- creased as shown in Table II, but considering the relative number of TABLE II Mean numbers of larve and adults of Oithona similis in the total region. April— June: number per minute. August-September: number per cubic meter. Month Larve Adults 1932 April 1,112 879 May 5,918 2,519 June 8,919 3,387 August 34,852 2,938 September 3,262 332 1931 August 33,925 3,550 September 4,825 94 adults and young, it does fot appear to have been an unusually prolific species during this period. In midsummer, however, with the maturing of the second western brood, a third crop of larve appeared, outnum- bering the combined young of all other species of copepods both in 1931 and 1932 (Tables II and IV). It is not possible to say whether the parent stock of the third eastern brood is equally productive, because spawning had just started at the time of the last cruise in the Gulf, in mid-August. Like Pseudocalanus, the greatest mortality in developing larve of Oithona seems to take place in the nauplius stages. This can be seen by comparing the abundance of nauplii in August with that of copepodites and adults in September (Fig. 9 and Table I1). In the Bay of Fundy, where the mean number of late nauplius and copepodite stages in Au- BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 177 gust amounted to 26,655 per cubic meter, September collections showed a decline to 1,720 late larve and 332 adults per cubic meter. A similar decline was also found in September, 1931. Fic. 6. Occurrence of copepodite stages of the western stock and egg sacs of the eastern stock in April, 1932. Number per minute of towing. 178 CEUAINIEE S wilinkt 1S kt Dispersal and Regional Abundance of Eggs and Larve Dispersal in 1932.—The method of tracing dispersal by monthly changes in the distribution of developmental stages of the different breeding stocks has been used in the present analyses. The general surface circulation and expected drift of eggs and larve within the Gulf and Bay of Fundy has previously been described (Fish, 1936D). olde SS os a i ee == Fic. 7. Occurrence of egg sacs representing the second brood of the western stock and copepodite stages of the first brood of the eastern stock in May, 1932. Number per minute of towing. BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 179 In the case of O. similis there was an apparent difference in the gen- eral distribution of larve of the two breeding stocks. Spawning of the western stock began seaward of the 100-meter contour and subsequent generations remained for the most part concentrated offshore. The eastern stock, on the other hand, although emanating largely from the eastern basin, soon became widely dispersed in the coastal zone, and succeeding broods, particularly in the more advanced larval stages (Fig. 7B), were until August at times more abundant inside than outside of the 100-meter contour. Western Stock—Spawning offshore in the western basin and in the region of Georges Bank in March, by April 16 (Fig. 64) larve in late nauplius and copepodite stages were found most abundant in the offing of Casco Bay and seaward in an are corresponding with the general surface drift to the outer banks. Smaller numbers occurred in the eastern basin, the Bay of Fundy and along the south coast of Nova Scotia. Like Calanus and Pseudocalanus, in outer waters Oithona simi- lis appears to reproduce off Halifax as early as in the western basin of the Gulf. This brood remained centered in the western area, although by late -May when the stock had matured and was spawning (Figs. 2A and 7A), a considerable number, having presumably circled the outer Gulf, were found entering the Bay of Fundy along the Nova Scotian side. The largest single haul of adults (possibly eastern stock; see p. 181) was made at the innermost station off Casco Bay, but egg sacs were for the most part concentrated offshore westward from Mt. Desert. The second brood greatly increased the size of the western stock and by the end of June had become generally distributed offshore in the Gulf. The larve, at this time in late nauplius and copepodite stages (Fig. 84), were abundant in all hauls seaward of the 100-meter line, particularly in the Casco section and along the Nova Scotian side of the Bay of Fundy. The fact that the numbers of both late nauplii (17,938 per minute) and copepodites (32,404 per minute) at the entrance to the Bay (Station 84) exceeded those found anywhere in the inner Gulf may indicate that immigrants from the western area are supplemented in June by contributions from another important production area, prob- ably in the vicinity of the outer banks.* Once the second brood had become generally dispersed offshore very little change was noted thereafter in the relative regional abundance of comparable stages during the remainder of the summer. In mid-August copepodites of the third and largest brood (Fig. 9B) were distributed in 3 Since immigrant larve of other species, abundant at the time, showed no evi- dence of unusual concentration at Station 8A (Fish, 1936a and b), this condition in the case of Oithona does not appear to have been due to local current action. 180 (GlaUNRIEIS) I 1S isl much the same manner as similar stages of the preceding brood in June (Fig. 84), being abundant at all offshore stations. Late nauplii of the same brood (Fig. 94) were somewhat more restricted, with concentra- tions west of Penobscot Bay and in the drift entering the Bay of Fundy. COPEPODITES fe e fh 1726 Fic. 8. Occurrence of copepodites of the second brood of the western stock and egg sacs of the second brood of the eastern stock in June, 1932. Number per minute of towing. BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 181 The coastal area east of Penobscot Bay and including the Bay of Fundy would appear to be relatively unproductive. The Bay apparently derives at times an abundant supply of nauplius stages from the eastern basin (Fig. 9C), but the eastern coastal region of the Gulf was found to be populated for the most part with a relatively small immigrant stock in more advanced stages (copepodites and adults). The large numbers of nauplii entering the Bay of Fundy along the Nova Scotian side must decline rapidly in the cold turbulent waters for there is little evidence that they survive the circuit of the Bay and are dispersed westward in the drift along the coast of Maine. The most probable source of the late stages found in the latter zone, east of Mt. Desert, is that portion of the drift from the outer Gulf and eastern basin passing westward across the entrance to the Bay. By mid-September the Oithona population had greatly declined in the Bay of Fundy. Observations were not made in the Gulf after Au- gust, but in view of the absence of an appreciable influx in the entering drift it is probable that the Bay reflected conditions throughout the region. Eastern Stock—The first brood of the stock maturing in the eastern part of the inner Gulf and Bay of Fundy was just appearing near its southern and western margins in late April. Eggs were found in the eastern basin, over German Bank, and in a limited portion of the western basin, but where vernal warming was delayed in the more turbulent eastern coastal zone of the Gulf and the Bay there was as yet no evidence of propagation. Egg sacs west of Mt. Desert were all found in the drift from the eastward and were probably derived from a parent stock trans- ported from the eastern basin (Fig. 6B). No eggs were found in the coastal zone anywhere in the Gulf. By late May this first brood consisted largely of late copepodite stages which had become generally dispersed both in inshore and off- shore wate1s, being most numerous well inshore in the western area (Fig. 7B). Some in the warmer western coastal zone had apparently passed through the final moult, for as indicated by the scarcity of eggs (Fig. 7A) at the innermost station off Casco Bay (Station 254) adults there (Fig. 24) were sexually immature and probably represented for the most part members of the eastern stock.* Further evidence that only those eggs transported beyond the eastern coastal region before hatching survive, is seen in the composition of the eastern stock at this time. Since propagation had not: begun in the inner Gulf east of Mt. Desert by the end of April, the product of this 4 The western stock offshore had already spawned at this time (Fig. 7A). 182 CHARTES i PUSE LATE NAUPLII EARLY NAUPLII © 16232 Fic. 9. Occurrence of late nauplii and copepodites of the third brood of the western stock, and early nauplii of the third brood of the eastern stock in August, 1932. Number per cubic meter. BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 183 late seasonal spawning a month later should have been represented by late nauplius stages, but very few were found. The young occurring in this region were copepodites assumed to be immigrants from more favorable production areas. The distribution of eggs of the second eastern brood in late June (Fig. 85) corresponded closely to that of the parent stock in the pre- vious month (Fig. 74). They were most numerous at the innermost stations in the Gulf and along the course of the drift into the Bay. In mid-summer Oithona appeared to withdraw from the coast and the third eastern brood, in egg and nauplius stages, was found in Au- gust centered in the eastern basin and extending eastward into the Bay ‘easice i Mean numbers of Oithona similis in the Gulf of Maine in 1931 and 1932. Number per cubic meter. Gulf of Maine Western Area Central Area Stage 1931 1932 1931 1932 (Aug. 21-25) (Aug. 11-14) (Aug. 26) (Aug. 8-15) Deer care ee emai 1,947 1,855 1,409 1,224 Early nauplius........ 11,765 15,363 8,044 10,981 Late nauplius........ 12,018 15,529 21,133 11,175 Copepodite.......... 7,908 7,217 9,220 5,144 /ANGRD Uae eM esto unease 3,537 2,557 3,575 3,090 of Fundy and westward along the outer stations to the western area (itis 2G) As in the case of the western stock, there must have been a great decline in the third eastern brood in late summer. In the Bay of Fundy by mid-September the survivors in late nauplius and copepodite stages were found entering along the Nova Scotian side in numbers not ex- ceeding 3,000 per cubic meter.® Annual Variation—lIn the Gulf the mean numbers of larve and adults in August of the two years were of a somewhat similar order. In 1931 (August 21-26) there were 35,044 larve and 3,556 adults per cubic meter, and in 1932 (August 8-15), 32,705 larve and 2,824 adults. However, the mean for the Bay of Fundy at the beginning of September 5 Within the region of observations in the Gulf in 1932, copepodites of the eastern stock in May (Fig. 7B) and eggs in June (Fig. 8B) were most abundant in the coastal zone westward from Mt. Desert. 6In August in this area early nauplii were found up to 35,970 per cubic meter. 184 CHARLES J; PISH in 1931 was more comparable with that of mid-September than of mid- August of the following year. In 1932 larve declined from 37,716 on August 17-21 to 3,262 per cubic meter on September 15-16, and adults from 3,120 to 332. The decrease had apparently already taken place in the Bay by September 1-5 in 1931, when the mean number of larve, 4,826, was almost as low as in the previous year in mid-September and the number of adults, 94, even lower. Considered by areas, in the Gulf the total numbers and relative per- centages in comparable stages exhibit no very striking differences. (Table III). A similar condition in the case of Calanus (Fish, 1936a, p. 130) and Pseudocalanus (Fish, 1936b, p. 202) has been interpreted to mean that in 1931 the season was delayed about two weeks. Adults were slightly more numerous in both the central and western areas in 1931. The mean number of larve was higher in the western area in 1932, and in the central area in 1931. TasBLe IV Mean numbers of larval stages of dominant species of copepods in the total region in 1932. April-June: number per minute. August-September: number per cubic meter. 1932 Calanus finmarchicus |Pseudocalahus minutus Oithona similis iNprilenen gr ean cee 2,007 7,126 1,112 IN anya eet ee edie 7,024. 11,484 5,918 ULOVERE Sl eeercepes, my een ee 2,156 21,303 8,919 (Nictist OM 977 5,639 34,852 September............ 41 3,506 3,262 Aprl-Septs! Leuba. (Mean) 2,441 9,812 10,813 RELATIVE NUMERICAL STRENGTH OF O. SIMILIS IN THE ZOOPLANKTON POPULATION Owing to its small size, Oithona similis is not taken in representative numbers in plankton nets of the type usually used in American waters and for this reason former available records do not indicate accurately its relative abundance in the boreal population of the western Atlantic. Its presence in large numbers in collections from the Newfoundland region whenever there was a considerable amount of ctenophore debris (Pinhey, 1926) would indicate that it is abundant in those waters in late summer, but was captured only when the meshes of the net became partially clogged. As far south as Chesapeake Bay O. similis has been found by Wilson (1932) to form an important addition to the winter plankton. In the eastern Atlantic Gran (1902) reported that this spe- BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 185 cies also plays an important role off Norway, and Ruud (1929) in the same region found it comprising up to 60 per cent of the copepod stock in July, 1925. In the present investigations Oithona similis rarely appeared in the coarse meter net samples, but an estimate of its importance in the region during the spring and summer months can be obtained by comparing the relative abundance of larve with those of other dominant species. In making such a comparison it must be borne in mind that the relative numbers of young do not necessarily indicate that adults occur in the same proportion. The mortality of developing young appears to be very much greater in some species than in others and, like Pseudocalanus (Fish, 19360), a very small percentage of Oithona larve apparently reach maturity. Again, because of its small size Oithona in the adult form must rank far below Calanus and Pseudocalanus in importance. However, it is not the adult stock of Oithona but the tremendous num- bers of larve produced by this prolific species which appears to be of greatest significance in the economy of the region. Particularly in August it must form a most important source of food for those species feeding on larval copepods. In Table IV it will be seen that, based on the mean numbers of young between April and September (1932), Oithona ranks with Pseudocalanus and far exceeds Calanus. Of added importance in the economy of the region is the fact that the seasonal maxima of the three species occur at successively later periods during the summer. Whereas Calanus reaches its peak in May (first brood), Pseudocalanus is most abundant in June (second brood), and Oithona (third brood) in August.’ In the latter month the mean of Oithona larve for the entire region amounted to 34,852 per cubic meter. SUMMARY 1. Oithona similis, although almost world-wide in distribution, 1s centered largely north of Cape Cod on the western Atlantic coast, and south from this point is gradually replaced by O. brevicornis. 2. The adult stock in the Gulf of Maine appears to approach the coast from offshore with rising temperatures in the spring. 3. Propagation begins in the Gulf in March and continues in a suc- cession of generations until September. 4. Two breeding stocks are distinguishable, due to an average dif- ference of about one month in the time of vernal propagation in the 7 To this seasonal progression may be added a September maximum of Centro- pages typicus larve throughout the Gulf and Bay of Fundy. 186 CHARLES J. Fish eastern and western parts of the region. This difference is reflected in subsequent generations after dispersal. 5. There appear to be four broods of the western stock, in March, May, July, and September, and three in the eastern, in April, June, and August. 6. A developmental period of two months during the early season and about six weeks in summer is indicated. 7. Oithona is at times far more prolific than either Calanus or Pseudocalanus, but mortality is relatively much higher, depletion being greatest in nauplius stages. 8. The western stock remained for the most part concentrated off- shore, but the eastern, particularly in advanced stages, was more gen- erally distributed, sometimes being more abundant inside than outside of the 100-meter contour. 9. There is some evidence that in June the local stock in the inner Gulf is supplemented by important contributions from another produc- tion area, probably in the vicinity of the outer banks. 10. There is no indication of successful propagation in the turbulent coastal region of the Gulf east of Mt. Desert or in the Bay of Fundy. 11. The larve of Oithona afford the most abundant source of food in the region in mid-summer for those animals feeding on larval plank- tonic organisms. Of added importance in the natural economy of the region is the fact that the seasonal maxima of the three numerically dominant species occur at successively later periods during the summer, Calanus in May, Pseudocalanus in June, and Oithona in August. BIBLIOGRAPHY Bicetow, H. B., 1926. Plankton of the offshore waters of the Gulf of Maine. Bull. U. S. Bur. Fish., vol. 40, Part II, 1924 (1926). Washington. Bocorov, B. G., 1932. Materials on the biology of copepods of the Barent and White Seas. Bull. State Oceanographic Institute. Moscow. Part 4. Fisu, C. J., 1925. Seasonal distribution of the plankton of the Woods Hole region. Bull. U. S. Bur. Fish., 41: 91, 1925 (1926). Washington. Fisu, C. J., 1936a. The biology of Calanus finmarchicus in the Gulf of Maine and Bay of Fundy. Biol. Bull., 70: 118. Fisu, C. J., 1936b. The biology of Pseudocalanus minutus in the Gulf of Maine and Bay of Fundy. Biol. Bull., 70: 193. Gran, H. H., 1902. Das Plankton des Norwegischen Nordmeeres von biologischen und hydrografischen Gesichtspunkten behandelt. Report on Norwegian Fishery and Marine Investigations, Vol. II, Part II (1909), No. 5. Ber- gen. HerpMAN, W. A., I. C. THomeson, anp A. Scott, 1898. On the plankton collected continuously during two traverses of the North Atlantic in the summer of 1897; with descriptions of new species of Copepoda; and an appendix on dredging in Puget Sound. Proc. and Trans., Liverpool Biol. Soc., vol. 12, Session 1897-98 (1898), p. 33. Liverpool. sf) BIOLOGY OF OITHONA SIMILIS, GULF OF MAINE 187 Pinuey, K. F., 1926. Entomostraca of the Belle Isle Strait Exposition, 1923, with notes on other planktonic species. Contr. Canad. Biol., N. S., vol. 3, No. GnOttawale: Ruup, J. T., 1929. On the biology of copepods off M¢gre, 1925-1927. Rapports et Procés-Verbaux, Conseil perm. internat. pour Vexploration de la mer, vol. 56. Scort, T., 1905. On some Entomostraca from the Gulf of St. Lawrence. Trans. Nat. Hist. Soc. Glasgow, vol. 7, N. S., Part I, 1902-1905 (1907), p. 46. Witson, C. B., 1932. The copepod crustaceans of Chesapeake Bay. Proc. U. S. Nat. Musewm, vol. 80, No. 2915. Washington. THE EFFECT OF ULTRACENTRIFUGING UPON CHICK EMBRYONIC CBEES, WERE SPE CLA Rr Dk NiCr TO WHE, RESDING «SNC US NID satis: MOOD SAIN IDILIS; H. W. BEAMS AND R. L. KING (From the Department of Zoology, State Umversity of Iowa) Few, if any, problems of the cell are more perplexing to the cytolo- gist than those of the structure of the “resting” nucleus and of the mitotic figure. Some of the structures involved can be seen in the living cell, but for others we must rely almost exclusively upon fixed and stained preparations for our information. Because of this fact, marked difference of opinion exists concerning those structures that can only be seen in the fixed preparations; many investigators consider them artifacts which have no counterpart in the living cell. Within the past four years, work in this laboratory with the ultra- centrifuge has to some extent concerned itself with the problems of the physical nature of protoplasm, particularly its components and inclusions in certain somatic tissues (Beams and King, 1934a, b; 1935). In the present study we propose to limit our description mainly to the effects of ultracentrifuging upon the “resting” nucleus, chromosomes, and mitotic apparatus, hoping thereby to contribute to the knowledge of their physical constitution. MaTERIAL AND METHODS The method of centrifuging is, for the most part, similar to that already described (Beams and King, 1934a). Forty-eight to 96-hour- old chick embryos were removed from the egg and placed in physiologi- cal salt solution warmed to 102° F. They were then immediately trans- ferred in the physiological solution to the rotor of the air-driven ultra- centrifuge, where they were whirled for 10 minutes at approximately 150,000 times the force of gravity and immediately thereafter fixed in Bouin’s solution. Sections were cut and were stained in Delafield’s or in Heidenhain’s hematoxylin. DESCRIPTION In Fig. 1 is shown a mesenchyme cell of the sclerotome region of an 80-hour chick embryo, illustrating the usual position of the inter- 1 Aided by grant from the Rockefeller Foundation for work on cellular biology. 188 EFFECT OF ULTRACENTRIFUGING UPON CELLS 189 6 b kinetic or “ resting”? nucleus and the normal distribution of its various components, namely, nucleoli, chromatin, “reticulum,” nuclear “ sap” (karyolymph) and nuclear membrane. It is quite evident that in the nuclei of these, unlike the condition for other cells (Herrick, 1895; Gray, 1927), the distribution of the various components is not noticeably atf- fected by the normal pull of gravity. However, when the force of gravity is increased 150,000 times, a marked displacement and strati- fication of the nuclear materials in the order of their relative specific gravity is evident, as illustrated in Figs. 2, 3, 4, and 5. In Fig. 2 the nucleolus is shown displaced to the heavy or centrifugal pole, the chromatin (basichromatin) and “ reticulum” or “linin” (oxychroma- tin) materials are next in order and the non-staining nuclear “sap” or karyolymph is concentrated at the lighter or centripetal pole of the nucleus. In this preparation the nucleus as a whole has not been greatly displaced within the cell. However, in Figs. 3, 4, and 5, a progressively more extensive effect is evident in that the nucleus is stretched in a direc- tion parallel with that of the centrifugal force. The condition in Fig. 5 is extreme, as the nucleus has been extended to over four times its nor- mal length; this illustrates beautifully the elastic properties of the nu- clear membrane and the materials included in it. We have never en- countered in any of our experiments evidence to support the view that the nucleus is anchored in position by cytoplasmic strands from the cell membrane, as has been figured for cells from the hair of the squash by Heidenhain (1907). A somewhat different effect is seen in the resting nucleus of the red- blood corpuscle, where a separation of the chromatic from the achro- matic elements of the nucleus is effected (Fig. 10). The darkly staining mass at the centrifugal pole contains the chromatic moiety while the rounded clear body at the opposite or centripetal pole is the achromatic or nuclear “sap” portion. This condition has been noted by us in centrifuged paramecium (King and Beams, 1934): such animals have been observed to live for several days, but apparently do not divide normally (unpublished data). This phenomenon clearly demonstrates that, in these cells too, the chro- matic elements of the nucleus are distinctly heavier and the achromatic materials lighter than the cytoplasm. In reference to the spindle in the following description of dividing cells it is evident that we are dealing with what Schrader (1934) has termed half-spindle and continuous fibers. However, since our mate- rial is not favorable for a careful differentiation of these components, we are not here concerned with the origin of the spindle components but will simply speak of them collectively as spindle fibers. 190 H. W. BEAMS AND R. L. KING myers, oermyer nd it : : , IPILAN IIS, IL BEE Ci Ow ULTRACENTRIFUGING UPON CELLS 191 Many different effects of centrifugal force upon the metaphase mitotic spindle are noted, depending to a large degree upon the orienta- tion of the spindle as regards the direction of the force. For instance, in Figs. 6, 7, 8, and 9, the force has been applied at right angles to the spindle and rather marked effects are observed. In Fig. 6, instead of the mitotic spindle being displaced in toto through the cytoplasm of the cell as is often the case in other material (Lillie, 1909; Spooner, 1911; Morgan, 1927), the two halves of the spindle with their associated centrioles are often stretched and bent centripetally through an angle of approximately ninety degrees; the attached metaphase chromosomes in this particular case have not been greatly displaced, nor has the spindle returned to its normal position after release of the centrifugal force. However, in Figs. 7, 8, and 9, similar but somewhat more marked ef- fects upon the spindle are noted. In Figs. 7 and 8 the bulk of the chro- mosomes has been displaced centrifugally and torn free from one half of the spindle; the other half of the spindle is greatly stretched and ap- parently was in the process of separating from the chromosomes when the centrifuge was stopped. It is of considerable interest that after such treatment the detached halves of the spindle still possess well- EXPLANATION OF FIGURES All the drawings were made by Miss Gladys Larsen from mesenchyme cells of approximately 80-hour-old chick embryos unless otherwise indicated. The figures have been magnified approximately 2,000 times. The direction of the centrifugal force in all cases has been toward the bottom of the plate. Fic. 1. Control cell showing the usual distribution of the “resting” nuclear elements. Fics. 2, 3, 4, and 5. Centrifuged cells showing a progressively more drastic effect upon the “resting”? nucleus and illustrating the relative specific gravity and elastic properties of its materials. Fics. 6, 7, 8, and 9. Cases illustrating the bending and stretching of the spindle toward the centripetal pole, displacement of the chromosomes from the metaphase plate centrifugally and disintegration of the spindle, leaving the centri- oles free in the cytoplasm. Fic. 10. Blood corpuscle. The nucleus has been separated into a chromatic portion (at the centrifugal pole) and an achromatic portion (at the centripetal pole). Fics. 11, 12, and 13. Cases depicting the displacement of portions of the chromosomes from the spindle. Always a portion of each chromosome remains attached to the spindle fibers. Separation of some of the spindle fibers are also observed. Fic. 14. Cell showing displacement of the chromosomes along the centrifugal half of the spindle. Fie. 15. Cell showing displacement and apparent disintegration of the cen- trifugal half of the spindle. Fic. 16. Cell with its spindle displaced to the centripetal pole. Portions of the chromosomes have been torn from the spindle. Fic. 17. Cell with an apparently normal centripetal half of a spindle, but with the lower centrifugal half missing. 192 H. W. BEAMS AND R..L. KING defined centrioles and spindle fibers. Here, as in other cases to be men- tioned later, fragments of the chromosomes still remain attached to the spindle fibers; this definitely indicates that the union between spindle substance (fibers) and chromosomes is stronger than the internal co- hesion of the chromosomes. Furthermore, it is of interest that, when they are free of most of their chromosome material, the half spindles move readily to the centripetal pole of the cells, and hence must be lighter than the surrounding cytoplasm. In the final stages the entire achromatic spindle, save for the centrioles, frequently undergoes a marked physical disintegration (Fig. 9). Only faintly can one observe a portion of the spindle fibers on the centripetal side of the chromosome mass. The two centrioles seem to be completely detached from the spindle substance and are free to move in a centripetal direction. They are not, however, as near the centripetal pole of the cell as might have been expected from the condition in Figs. 7 and 8. In still other material subjected to the same treatment, cells may be found where most of the chromosomes have been thrown off the equa- torial plate of the spindle into the centrifugal end of the cell; the spindle remaining apparently normal in the centripetal end of the cell with very small portions of the chromosomes still attached at its equatorial plate (Fig. 11). Probably the spindle has been greatly distorted, as shown in Fig. 6, during the action of the force, and subsequently has returned to normal upon cessation of centrifuging. Frequently, in similarly treated cells, the spindle has been separated, since part of the spindle fibers can be seen to be still attached to the displaced chromosomes (Figs. 12 and 13). The presence of fragments of chromosomes re- maining attached to the spindle at the equatorial plate after such treat- ment strongly suggests that the spindle and the union between the spindle substance and chromosome is able to withstand greater distortion before breaking than the substance of the chromosomes themselves. When the mitotic spindle is oriented approximately parallel to the centrifugal force, quite different effects are found. In Fig. 14 the force has acted slightly obliquely from the upper left, forcing the bulk of the chromosomes centrifugally and to the right side of the lower half of the spindle. Here too, as in other cases mentioned above, the fiber attach- ment portions of the chromosomes are not displaced from the equatorial plate. Sometimes most of the chromosomes are thrown completely free from the spindle into the centrifugal half of the cell (Fig. 15). We were unable to determine whether or not the centrifugal portion of the spindle in this instance has disintegrated or is only masked by the displaced chromosomes. In any case, the centripetal half of the spindle seems to be normal and to possess the usual chromosome fragments on EFFECT OF ULTRACENTRIFUGING UPON CELLS 193 the equatorial plate at the points of spindle attachments. Other cells, treated in a like fashion, often have only part of their chromosomes dis- placed from the spindle (Fig. 16). In such cases the intact spindle, with its remaining chromosomes and fragments of displaced chromo- somes, usually moves into the centripetal half of the cell, the displaced chromosomes to the centrifugal half. Finally cells are found which have a normal centripetal half of a spindle with its metaphase plate of chro- mosomes while the lower or centrifugal half of the spindle is apparently absent altogether (Fig. 17). This condition is strikingly similar to that recently reported for the onion root tip by Luyet (1935). In such a case, the question immediately arises whether the lower half of the spindle has been forced centripetally, as is known to be the case in Fig. 6, and subsequently lost to view, as has been suggested by Luyet; or whether it is no longer distinct from the surrounding cytoplasm, due perhaps to the pressure developed between the chromosomes and the cell membrane. We are inclined to accept the latter view, because we have never seen the lower half of the spindle in the large number of such cases examined. In this connection the observations of Harvey (1935) are of interest: she observed that the mitotic spindle of certain sea-urchin eggs moved to the centripetal pole upon centrifuging and that the astral rays became much shortened on the compressed or centripetal side of the aster. The question as to the survival of these chick cells has not been satis- factorily answered as yet. We hope that sometime in the near future attempts may be made to grow centrifuged chick cells in tissue cultures. However, we already know that root tip cells of wheat seedlings do re- cover after similar distortion by ultracentrifuging.? DIscUSSION Notwithstanding the fact that thousands of papers have appeared describing and illustrating the nucleus in practically every type of cell since it was first discovered by Robert Brown in 1831 (Wilson, 1925), there still remains no general agreement as regards its real physical constitution. For example, one reads in Wilson’s “ The Cell in Develop- ment and Heredity,” that the resting “vesicular nucleus, as seen in sec- tions, usually shows four distinct components, namely: an enclosing wall or membrane; a nuclear framework usually described as a network or reticulum, though by some observers regarded as an alveolar struc- 2Eges of Ascaris suum will recover and divide after ultracentrifuging at 400,000 times gravity for an hour; we have recently centrifuged Ascaris eggs at 150,000 times gravity for 10 days during which time they divided. (In press, Science, 1936.) 194. H. W. BEAMS AND R. L. KING ture; the nuclear sap, enchylema, or ground substance which occupies the interstices of the framework; and one or more nucleoli, massive and usually rounded bodies suspended in the framework.” On the other hand, some investigators express the opinion summarized by Gray (1931) in his “ Textbook of Experimental Cytology” that “ Since all nuclei exhibit a visible granular or fibrillar structure after coagulative fixation, it is generally supposed that the structures seen in preserved preparations or in moribund nuclei are to be regarded as purely artificial products of coagulation, which cannot be correlated with the fundamental structure of the living nucleus. This view, developed many years ago by Hardy (1899), is now accepted by the majority of animal cytologists (Champy, 1913; Policard, 1922; Lewis and Lewis, 1924; Chambers and Rényi, 1925; Schitz, 1925; Burrows, 1927).” It will be clear to those who analyze the literature on the physical nature of the nucleus that those who support the former view are mainly morphological cytologists, who have worked chiefly with fixed materials (however, see Bélar, 1929; Cleveland, 1935) and that those who sup- port the latter view are chiefly experimental cytologists, who have worked mainly with surviving cells. In other words, the question re- solves itself into whether or not the chromatin, “ reticulum,” nucleolus and nuclear ‘“‘ sap” seen in fixed preparations of the “ resting’ nucleus, correspond to the condition in the living nucleus to the same degree as do the chromosomes and other elements of the dividing nucleus. The results reported in this paper seem to bear upon these questions, insofar as we have demonstrated that after centrifuging the living nucleus and immediately fixing and staining it, all the structures men- tioned above are present, but displaced and stratified in the order of their relative specific gravity. This can only be interpreted to mean that the nucleus is not a homogeneous body, although it often appears so in the living material. It means, we think, that it is composed of materials of different relative density, yet of similar index of refraction, which usually prevents its elements from being seen in the living nucleus. Moreover, it is significant that these displaced nuclear elements appear morphologically similar in structure to those of the uncentrifuged fixed nucleus. Furthermore, the chromatin and “reticulum” materials in the centrifuged and subsequently fixed preparations take up the same relative position in the nucleus as does the prophase chromatin and chromosomes and hence must be of a similar specific gravity. There- fore, these results as well as the observations of the majority of cy- tologists have convinced us that well-fixed preparations of the “ resting ” nucleus probably give as reliable a picture of the condition during life as do fixed preparations of the prophase nuclei and chromosomies. EFFECT OF ULTRACENTRIFUGING UPON CELLS 195 9 We, of course, realize that the structure of the “resting” nucleus no doubt varies greatly in different types of cells. For example, the chromatin in the nuclei of certain insect ova (Gardiner, 1935) is very scarce (diffuse?) as determined by the Feulgen method. In contrast to this, however, the nuclei of the salivary glands of many dipterous larvee give a very marked reaction as of very orderly arranged segments of chromatin on an achromatic framework (“reticulum”). It would seem almost incredible that in the nuclei of the salivary glands of, for instance, Chironomus larve (King and Beams, 1935), all the material of the nuclei, except for the nucleolus, is evenly dispersed in a homog- enous mass throughout the nucleus and subsequently clumped by fixa- tion into long chromosomes, the identity of which may be recognized from cell to cell. In this connection, Doyle and Metz (1935) have re- cently shown, in the living nucleus of the salivary glands of Sciara larvee, that no chromosomes or chromatic segments, so characteristically seen in aceto-carmine preparations, are to be observed. They never- theless conclude from their studies that the chromosomes and chromatin segments are always present in the living nucleus, but that their refractive index is such that they are not optically differentiated. Némec (1929) concludes from his work on centrifuged plant nuclei that a definite “reticulum” is present which is attached to the nuclear membrane. From our studies we cannot be sure just what constitutes the so-called “ reticulum”; that is, whether it represents an achromatic framework in the nucleus impregnated with chromatin or whether. it consists entirely of chromatin material. We are inclined to think that the “reticulum ” in preserved cells has not been precipitated by fixation from a homogenous mass, but that it corresponds to a special material similarly distributed in the living “ resting ” nucleus. When an inquiry is directed toward the structure and function of the mitotic figure, particularly the centrioles and spindle, one encounters an enormous amount of literature containing a most extraordinary group of conflicting contentions (see Wilson, 1925; Sharp, 1934; Schrader, 1934 and Bleier, 1930, for a review of the literature). It is beyond the scope of this paper to deal generally with a discussion of the mech- anism of mitosis, but our results do seem to bear upon certain points of the structure of the mitotic apparatus which will be discussed below. The reality of centrioles in living cells has recently been seriously questioned by Fry (1929). However, the centrioles in our material still possess their typical characteristics after being moved through the cytoplasm to the centripetal pole of the cell where they are in some cases detached and rest freely in the cytoplasm. This we believe is 196 H. W. BEAMS AND R. L. KING good evidence for their existence in the living cell. Furthermore, the fact that they have been so clearly demonstrated in the living cell by Cleveland (1935) and their genetic continuity so carefully followed by Heuttner (1933) leaves little doubt that they are actually present. As regards the spindle, one group of investigators of which Chambers (1924) and Lewis and Lewis (1924) are typical, holds that the mitotic spindle is optically homogeneous, and the appearance of fibers in fixed preparations is an artifact, without any correspondence to structures in the living cell. However, in contrast to this view is the one derived by Cleveland (1935) from his work on living protozoan cells. Here “there is not the slightest doubt regarding the existence of the cen- trioles, the formation of the achromatic figure from the centrioles, the fibrillar nature of the achromatic figure, and the role of the achromatic figure in nuclear division.” Our work points to the fact that the spindle and centrioles in chick cells are, in general, distinctly lighter than the chromosomes and for the most part lighter than the surround- ing cytoplasm, as is evident by their frequent movement to the centrip- etal pole of the cell. The fact that the spindle is capable of consid- erable bending and distortion, without becoming detached or dissolved from the chromosomes or centrioles, argues that it is a body of con- siderable form and structure rather than an artifact. However, from our work we cannot definitely say that fibers are actually present in the living spindle, but certainly the structure of the spindle is capable of considerable tension and plasticity ; it may be partially split and still possess its typical appearance in fixed preparations. These observations, we believe, support those of Cleveland (1935), who was able to estab- lish by pulling the centrioles that the spindle fibers are real and defi- nitely attached to the chromosomes. Recently Schrader (1934) has centrifuged dividing cells of crustaceans, molluscans and insects, with much less force than used by us. He states that “ metaphases so treated continue their mitotic activities and it is concluded that the half-spindle components as seen in fixed preparations are not coagulation artifacts but have a morphological basis in the living spindle.” Furthermore Wyckoff (1934), who attempted to photograph (with ultraviolet light) the living spindle, says “you surely see something; but whether it is the individual fibers or the result of a state of strain in the viscous protoplasm, I don’t know.” Lewis (1934) has recently reported that when the metaphase spindle of chick cells in tissue culture is exposed to a hypotonic medium, the fibers disappear, the chromosomes are dis- persed from the equatorial plate and the progress of division stopped. However, if the hypotonic medium be replaced in time by an isotonic EFFECT OF ULTRACENTRIFUGING UPON CELLS NO medium, a new spindle forms, the chromosomes collect on the equator- ial plate, and division follows. If the cell is fixed and stained after the action of the hypotonic medium but before the addition of the isotonic medium, no spindle fibers are to be found, although the centrospheres seem not to have disappeared. These observations indicate definitely that the spindle is essential for division in chick cells, and that it 1s capable of being dissolved and reformed while the cell is in the meta- phase state of division. In any case, whatever the structure of the spindle may be, we are in agreement with Seifriz (1929) “that it makes little difference whether we regard the spindle fibers as threads, rows of granules or lines of force. If fixation causes previously existing lines of force to appear as fibers this is sufficient, for, after all what are lines of force if not the linear orientation of particles, whether electrons or protoplasmic granules.” CONCLUSIONS 5} 1. Chick embryonic cells, both “resting ”’ and in stages of division, have been centrifuged at approximately 150,000 times gravity for 10 minutes. 2. The elements of the “resting ’’ nucleus were stratified in the order of their decreasing specific gravity as follows: (a) nucleoli; (0) chromatin and “ reticulum ’”’; (c) nuclear “ sap.” 3. Evidence is presented to support the view that fixed preparations of the “resting ”’ nucleus give as characteristic a representation of the elements during life as do fixed preparations of prophase nuclei and metaphase chromosomes. 4. The metaphase spindle has been distorted, its chromosomes dis- placed, its fibers separated, and, in some cases, it has been completely disintegrated by centrifuging. 5. These results support the claim that spindle fibers and centrioles as seen in fixed preparations have a morphological basis in the living spindle. ” IEAPAD DRA IE Ueda, CIA) Beams, H. W., anv R. L. Kine, 1934a. The effects of ultracentrifuging upon the Golgi apparatus in the uterine gland cells. Anat. Rec., 59: 363. Beams, H. W., anv R. L. Kine, 1934b. Effect of ultracentrifuging on the mito- chondria of the hepatic cells of the rat. Anat. Rec., 59: 395. Beams, H. W., anv R. L. Kine, 1935. The effects of ultracentrifuging the spinal ganglion cells of the rat, with special reference to Nissl bodies. Jour. Comp. Neur., 61: 175. Bexar, K., 1929. Beitrage zur Kausalanalyse der Mitose. II. Untersuchungen an den Spermatocyten von Chorthippus (Stenobothrus) lineatus Panz. Arch. f. Entw. Mech., 118: 359. 198 H. W. BEAMS AND R. L. KING Bierer, H., 1930. Untersuchungen iiber das Verhalten der verschiedenen Kern kkomponenten bei der Reduktionsteilung von Bastarden. La Cellule, 40: 83. Cuampers, R., 1924. The Physical Structure of Protoplasm as Determined by Micro-dissection and Injection. General Cytology (Edited by Cowdry), DuZIee CLEVELAND, L. R., 1935. The centriole and its role in mitosis as seen in living cells. Science, 81: 598. Dove, W. L., anp C. W. Metz, 1935. Structure of the chromosomes in the sali- vary gland cells in Sciara (Diptera). Biol. Bull., 69: 126. Fry, H. J., 1929. The so-called central bodies in Cerebratulus eggs, etc. Biol. Bull., 57: 131. Garpiner, M., 1935. The origin and nature of the nucleolus. Quart. Jour. Micr. Sica iu SZ3" Gray, J., 1927. The mechanism of cell-division. IV. The effect of gravity on the eggs of Echinus. Brit. Jowr. Exper. Biol., 5: 102. Gray, J., 1931. A Textbook of. Experimental Cytology. Cambridge. Harvey, E. B., 1935. The mitotic figure and cleavage plane in the egg of Pare- chinus microtuberculatus, as influenced by centrifugal force. Biol. Buill., 69: 287. Hemenuatin, M., 1907. Plasma und Zelle. (Jena.) Herrick, F. H., 1895. Movements of the nucleolus through the action of gravity. Anat. Anz., 10: 337. Huettner, A. F., 1933. Continuity of the centrioles in Drosophila melanogaster. Zeitschr. f. Zellforsch. u. Mikr. Anat., 19: 119. Kine, R. L., anp H. W. Beams, 1934. Somatic synapsis in Chironomus, with spe- cial reference to the individuality of the chromosomes. Jour. Morph., 56: 5/7: Kine, R. L., anp H. W. Beams, 1934. Effect of ultracentrifuging on Paramecium. Proc. Soc. Exper. Biol. and Med., 31: 707. Lewis, M. R., 1934. Reversible solation of the mitotic spindle of living chick em- bryo cells studied in vitro. Arch. exp. Zellforschg., 16: 159. Lewis, W. H., anp M. R. Lewis, 1924.° The behaviour of cells in tissue cultures. General Cytology (Edited by Cowdry). Liuiz, F. R., 1909. Karyokinetic figures of centrifuged eggs; an experimental test of the center of force hypothesis. Biol. Bull., 17: 101. Luyet, B. J., 1935. Behavior of the spindle fibers in centrifuged cells. Proc. Soc. Exper. Biol. and Med., 33: 163. Morean, T. H., 1927. Experimental Embryology. Columbia University Press. Nemec, B., 1928. Uber die Beschaffenheit der achromatischen Teilungsfigur. Arch. exp. Zellforschg., 5: 77. Nemec, B., 1929. Uber struktur und aggregatzustand des Zellkernes. Proto- plasma, 7: 423. Scuranver, F., 1934. On the reality of spindle fibers. Biol. Bull., 67: 519. Serrriz, W., 1929. The structure of protoplasm. Biol. Rev., 4: 76. SHarp, L. W., 1934. Introduction to Cytology. McGraw-Hill. Spooner, G. B., 1911. Embryological studies with the centrifuge. Jour. Exper. Zool., 10: 23. Witson, E. B., 1925. The Cell in Development and Heredity. Macmillan, N. Y. Wycxorr, W. G., 1934. Ultraviolet microscopy as a means of studying cell struc- ture. Cold Spring Harbor Symposia on Quantitative Biology, 2: 39. ipsa ClSlOF HAV VOW MEER UPON tHE, FISSION RAT BAND) PE bib bh C Vier Ey Ok Elks CILIATE UROLEPTUS MOBIENS ARLENE JOHNSON DELAMATER (From the Department of Zoélogy, Columbia University, New York City) Since the discovery of the heavy isotope of hydrogen, a great number of “biological” experiments have been conducted with a variety of concentrations and with a diversity of complex phenomena such as growth, longevity, enzymatic reactions, oxygen consumption, lumines- cence, and permeability relations. For reviews of the current literature Barnes and Jahn (1934) or Fox (1934) may be referred to. In gen- eral, the work with low concentrations of deuterium shows a diversity of results; contradictions frequently occur as to the effects upon the same biological processes. A depressing effect has not usually been found. On the other hand, indications of stimulation or of the absence of effects exist. In the work with Protozoa, a stimulating effect in respect to popula- tion numbers was reported with Euglena gracilis in D,O of 1.000061 density by Barnes (1934). This would indicate that some life processes are speeded up in this weak concentration; Barnes and Larson (1934) also hold that the life span of primitive cells such as Spirogyra is in- creased by dilute solutions (e.g. 0.06 per cent) of deuterium. It ap- peared to have little or no effect upon the rate of cell division or upon the rate of elongation of the cells. The autodigestive processes of the planarian, Phagocata gracilis, during starvation were considerably slowed down in 0.06 per cent. With a concentration of 0.13 per cent to 0.47 per cent D,O, the effects were “ progressively obscured” due to the increase of bacteria and molds. Clearer results probably could have been obtained by a frequent change of the solutions. Harvey (1934) as well as Taylor, Swingle, Eyring, and Frost (1933) reported no toxic effects with low concentration upon Paramecium caudatum; however, no actual account of fission rates or length of life was given. Barnes (1935) also stated that “ workers who claim that dilute heavy water has no influence on living organisms have not allowed sufficient time for the appearance of recognizable effects.” It was with this very purpose that an analysis of the fission rate of the ciliate protozoan, 199 200 ARLENE JOHNSON DELAMATER Uroleptus mobilis, during its complete life cycle was made. Since fission rate is one of the indications of the state of vitality of this protozoan, it was proposed to find whether a dilute concentration of deuterium oxide would alter this vitality over a long period of time, and if it would produce a stimulating effect such as has been reported upon the flagellate, Euglena, by Barnes. Secondly, it was proposed to note any changes in the total longevity of the culture. This type of isolation work involves no obscuring of effects by molds or bacteria since each animal is placed daily into fairly constant, fresh medium. With intermediate concentrations, Taylor, Swingle, Eyring, and Frost (1933) reported that Paramecium caudatum may survive for about 72 hours in 60 per cent to 70 per cent D,O, although many die after 48 hours. A concentration of about 50 per cent was selected to test the effects upon the fission rate of Uroleptus mobilis and its length of survival. The contractile vacuole rate was also observed as an additional method of noting any changes upon direct immersion into 50 per cent D,O. The results will be compared below with those ob- tained by Barnes and Gaw (1935) for the immersion of Paramecium caudatum into 30 per cent D,O. The work with high concentrations has yielded relatively consistent results, namely that the protoplasm of all organisms cannot survive for more than short periods. Several Protozoa have been tested. Harvey (1934) found that Amaba dubia, as well as an unnamed species, was killed in 6 to 20 hours, and that although Euglena proxima exhibited avoiding reactions, then rounded up and became immobile soon after immersion in 90 per cent D,O (a certain proportion remaining mobile, however, for three days), activity can be regained after five days by a return to normal water. With Euglena gracilis, recovery was possible after nine days in the D,O, thus indicating that the animals were not “irreversibly injured.” However, whether the treated animals would show delayed effects in their subsequent life history was not tested. Taylor, Swingle, Eyring, and Frost (1933) found a similar decrease in motility for Euglena viridis in 92 per cent D,O, although recovery tests were not made. Paramecium caudatum swam slowly upon immersion into 97 per cent D,O; in two hours the contractile vacuole enlarged with a cessation of contraction; blisters formed and it died within less than 24 hours (Harvey). Taylor, Swingle, Eyring, and Frost state that more time was required to kill Paramecium caudatum than frogs, fish, or flatworms, and they note that “ Paramecium shows more resistance to heavy water than highly organized animals studied.” If Uroleptus mobilis had been selected by chance instead of Paramecium, it would have been found that no more time for disintegration was needed, since EFFECTS OF HEAVY WATER ON FISSION RATE 201 this ciliate usually collapses at once. This scattered work points to the fact that until a systematic immersion of various types of Protozoa is made to answer such questions as the correlation of the kinds of pel- licles with the degree of resistance or whether animals in the encysted state may be more resistant to high concentrations, no general compari- sons can be accurately made. I wish to express my appreciation to Dr. Gary N. Calkins for the suggestion of this work and for his helpful advice, to Dr. James Curry for the making of the dilutions, to Dr. T. M. Sonneborn for his criti- cisms, and to Dr. H. S. Jennings for permission to complete the work at the Johns Hopkins University. The deuterium oxide was supplied by the Chemistry Department of Columbia University upon a grant _ from the Rockefeller Foundation. MATERIAL AND MErHops Cultures in 0.44 Per Cent Deuterium Oxide The descendants from each member of two pairs of conjugating Uroleptus mobilis which were taken from stock mass cultures were selected for study. From each single ex-conjugant one series of de- scendants was carried in isolation cultures throughout the life cycle; these series will be designated as A, B, C, and D respectively. Series A and B are derived from the members of one of these conjugating pairs; C and D from the other. In each series there were six isolation lines, each line being one continuous sequence of descendants. The medium for the normal control lines was made in the same wav as that of Calkins for Uroleptus (1919) ; 50 mg. of timothy hay were boiled for 5 to 10 minutes in 50 cc. of Great Bear Spring water to which a few grains of white flour were also added. A second type of medium was prepared in the same manner, except that redistilled water was used instead of the usual spring water. For the medium of 0.46 per cent D,O, a modification in preparation was made due to the sufficient but limited supply of the material. In order to produce a corresponding concentration of daily food supply in the 0.46 per cent D,O, 50 cc. of distilled water with the same proportions of hay (i.e. 50 mg.) and flour as above were boiled to 10 cc. The same amount of hay was again added and the liquid was further boiled to 1 cc. and allowed to remain for 24 hours. The D,O medium was then made up in the proportions of one drop of concentrated medium to 24 drops of 0.46 per cent D,O; the concentration of D,O was thus lowered to about 0.44 per cent D,O. The 0.46 per cent D,O came from the same lot as that used by Curry, Pratt, and Trelease (1935) (for preparation of which, see same). 202 ARLENE JOHNSON DELAMATER All lines of each series were run for the first four ten-day periods in normal medium. During Period 5 in Series B, C, and D, and Period 6 in Series A, two lines of each series were transferred into the redistilled water medium and two lines into the 0.44 per cent D,O by gradual changes extending over six days. For protection against evap- oration and change of concentration of cultures, the depression dishes were fixed with vaseline covers. The cultures were continued in the same type of medium until the end. When one line died, a substitution of an animal from another line in the same type of medium was made. The average fission rates for ten-day periods were calculated after Calkins (1919). If a line was ended by death of an animal during the ten-day period, it was not used in the calculation. If death oc- curred in all lines which were in the same medium at various times during a ten-day period, the average was made by averaging the fission rates of the preceding and succeeding periods (Calkins). Since the animals were under unfavorable conditions of transportation during Period 8, the values were also calculated in this way. For the first four periods, the six lines of one series were averaged together. During Periods 1 to 4, the animals remained at room temperature which was around 24° C. An extreme temperature variation killed some of the animals in Period 4; the cultures were thereafter kept at a temperature of 24° to 28° C. Although this temperature variation doubtless accounts for some of the variation in the fission rate from period to period, it introduces no error into comparisons made syn- chronously, since all cultures were exposed to the same temperature at any one time. Cultures in 48.1 Per Cent Deuterium Oxide The descendants from one ex-conjugant of a conjugating pair which was selected from a stock culture of Uroleptus mobilis were carried in six lines for the controls. All lines which were started in the 50 per cent D,O medium or in the special control water were derivatives from the control lines; in all cases, six lines of each type were carried. The normal and heavy water media were made up in the same way as those in the former experiment except that 50 per cent D,O was used instead of 0.46 per cent D,O. The 50 per cent D,O was prepared by the dilution of purified 95 per cent D,O. For a part of the experiment, the animals were sub- jected to a special type of control water which was put through the same process as that used for the preparation of the deuterium of high concentrations, except that it had a concentration of deuterium which EFFECTS OF HEAVY WATER ON FISSION RATE 203 corresponded to ordinary distilled water. After dilution with the neces- sary food solution, the 50 per cent water was about 48.1 per cent. From Period 10 on, distilled water was substituted for the spring water in the controls. From Period 1 to 6, the temperature ranged from 24° C. to 28° C. During Periods 6 through 8, the animals were transported several times and remained at room temperature. After Period 8, they were kept at 222i 27-7. In order to test the abundance of the bacteriological food supply in the 48.1 per cent D,O, four samples of the medium after 24 and 48 hours were plated upon nutrient agar which was made up with the 50 per cent D,O, and were compared with four samples from the control lines made upon normal agar upon the same day during Period 25 and at the same temperature. As much care as possible was taken to plate approx- imately the same amounts of fluid by following one type of procedure, although this type of quantitative work cannot be accurate even at its best. In twelve cases, the contractile vacuole rate of Uroleptus mobilis was observed before, after, and during immersion into 48.1 per cent D,O. The specimens came from the isolation lines of the group of animals which was used as controls in testing the effects of 48.1 per cent D,O upon the fission rate; they were taken at random, but all came from Period 19 in which the average fission rate was 13.1. Before observing, the animals were placed in several drops of fresh, normal medium for one to two hours. The animals were placed in one drop of the 48.1 per cent D,O medium in a depression slide covered with a vaselined coverglass for observation. The D,O medium was made up with nu- trient in the same manner as in testing the fission rates. Upon immer- sion of the animal, the fluid was carefully mixed with the pipette. Upon removal, the animal was briefly washed through three or four drops of the normal medium, before observing. For any continued observation the animal was transferred to fresh medium every two hours. The figures used represent the averages of the number of seconds between successive contractions of the vacuole, based upon about ten consecutive measurements recorded by the use of two stop watches. RESULTS Low Concentration In mass cultures from all three types of media, 1.e., the spring water, redistilled water, and the 0.44 per cent D,O medium, normal and similar 204 ARLENE JOHNSON DELAMATER amounts of conjugants as well as encystments were obtained during the life cycle. Thus 0.44 per cent D,O has no striking effect, as re- vealed by this general method, upon the ability of animals to conjugate or encyst. . In considering the effect of the low concentration of 0.44 per cent D,O upon a measurable function such as the fission rate of the ciliate, Uroleptus mobilis, no changes beyond the normal variation occurred, not only during a short time such as ten to twenty days, but also during a prolonged contact for at least ninety days with this low concentration Series D Series B 5 10 us) Fic. 1. Average number of fissions for four series: A, B, C, and D, during the life cycle in media made with spring water ( ), redistilled water (-—-—), and 0.44 per cent D.O (——— ). Ten-day periods (abscisse) are plotted against the average number of fissions for ten days (ordinates). of D,O. No accumulative or delayed effects were found, as may be seen from a comparison of the graphs for the four series in Fig. 1. For the most part, the 0.44 per cent D,O lines followed those in redistilled water more closely than those in the normal control medium. Except for short divergences in these two media such as in Series A, Period 6; Series B, Periods 10 and 14; Series C, Period 12; and Series D, Periods 10 and 12; the rates were almost identical throughout the seventy or more days. In all media the appearance of the animals was normal. BEBE CISION HEAVY WATER ON FISSION RATE 205 Furthermore, in considering the total number of generations, it was found that the average number for the animals in 0.44 per cent D,O cultures was intermediate in three series (A, B, and C). In Series D, the average of 233.5 generations in 0.44 per cent D,O was slightly less than that of 242.5 generations for the redistilled water cultures or 254.5 for the normal controls. These differences in all cases le within the limits of normal variation. In no series is there even a slight indi- cation that the total longevity is prolonged or cut short by the contact of the animals with this low concentration of D,O. The percentage of animals surviving in the 0.44 per cent D,O for seven periods after immersion was of the same magnitude as in the other two media for the same length of time. During this time, not iS) 10 I5 20 25 30 Fic. 2. Comparison of fission rate of Group A ( ) in normal medium, and Group B (—--) in special control water put through the same process as the 48.1 per cent D.O. Return to normal medium -——. Abscissa = ten-day periods. Or- dinate = mean number of fissions in ten days. more than two deaths occurred in any one line; there was only one © death in all four series during the transitional period in the 0.44 per cent D,O lines. Intermediate Concentration Conjugation and encystment tests made on animals which were living in normal spring water or in special control water usually yielded many viable conjugants and cysts. In the greater proportion of cases in the 48.1 per cent D,O, conjugation was not obtained since the conditions of the medium became toxic before enough animals could be obtained by their slow division to produce conjugating conditions. In the few cases where enough animals were obtained, conjugation always occurred in 206 ARLENE JOHNSON DELAMATER good proportions but the ex-conjugants never survived. In two to three days, after conjugation in the culture, the ex-conjugants became small, either subsequently disappearing or forming what appeared to be cysts. Attempts at ex-cystment were unsuccessful. Similar results for the ex-conjugants from 128 conjugating pairs which were isolated in fresh D,O medium were obtained. These also could not be revived in 48.1 per cent D,O and disappeared after a time. Twenty-four conjugating animals were placed in normal medium; 3 of the ex-conjugants died within ten days, but the other 21 survived. Their subsequent fission rates, however, were not tested. The normal control animals will be designated as Group A; deriva- tives from them, as Group B, as shown in Fig. 2. The former group tS) 10 15 20 25 50 Fic. 3. Effects of 48.1 per cent D.O on fission rate and length of life cycle of animals immersed in it at various ages. Group A ( ), in spring water; Group C (---), D (-——), and E (-:-), in 48.1 per cent D.O. Group C introduced into D:O when mature, Groups D and E introduced into D2zO when old. Abscissa = ten-day periods. Ordinate = mean number of fissions in ten days. survived for 31 ten-day periods and its fission rate was practically the same as that of Group B. This similarity existed both in Periods 11 to 13, when Group B was carried in special control water; and during the following six periods, when Group B was carried in normal medium. This shows that any changes found in the 48.1 per cent D,O medium are not due to the presence of injurious agents, since this special control water was carried through the same process as the 95 per cent D,O from which the 50 per cent D,O was diluted. At the end of Period 2, when the animals were in the forty-fifth to forty-eighth generations, derivatives (Group C) from each of the con- trol lines were placed in 48.1 per cent D,O, and allowed to remain until they died (Fig. 3). During the first fifty days (1.e., Periods 3 to 8) EFFECTS OF HEAVY WATER ON FISSION RATE 207 the fission rate decreased to about half that of the normal animals so as to approximate that of old animals in a declining period such as No. 22 or 23. After a slight rise in Periods 8 to 10, the rate continued near the new level for about ninety days to Period 20. A decrease set in which ended in the death of the animals in about fifty more days. The slight increase in the rate may be interpreted as a delayed adjust- ment of the animals to the new environment, but due to the chemical or physical effects of the D,O, the adjustment or return to the normal level was impossible. From Periods 15 to 20, the rates of both the control animals and those in 48.1 per cent D,O were at the same level because the former were in the early stages of their decline. It is also seen from Fig. 3 that although 48.1 per cent D,O depressed the fission rate greatly during “ maturity ” (1.e., that time during the life cycle in which the animals are dividing at a high rate), the animals showed “aging ”’ (the first signs of a general decline in the fission rate) at about the same time as the controls. However, “ senescence ” (period of very low fission rate which is followed by death) was cut much shorter. Not only did the animals which were placed in 48.1 per cent D,O during a period of high vitality divide more slowly but they also did not live as long. The animals began to die off in Periods 23 to 24, with an end to all lines in Period 25; while the controls did not start to die off until Periods 27 and 28, and ended in Period 31. Consequently, since the fission rate was lower in 48.1 per cent D,O and since the length of life was shorter, the total number of generations was less, an average of 220.6 generations as compared to one of 366.1 for those in normal medium. That these changes, in deuterium of this concentration, were not due to a continued insufficient diet can be assumed from the results of bac- terial checks on the medium. In each case of the samples taken after 24 and 48 hours from the normal and heavy water cultures, the number of bacterial colonies upon nutrient agar was approximately the same. Also the same types of colonies in the same proportions appeared. A\I- though no further examination of exact ratios was made, this roughly indicated that the bacterial food supply was at least ample in quantity ’ and quality. When normal old animals (Group D) in the 328th to 346th genera- tions at the end of Period 21 were placed in 48.1 per cent D,O, the fission rate fell to 3.0 and soon to zero (Fig. 3). By Period 25 all lines were dead, thus living only for a short time. Although these animals were dividing at a rate of 12.0 fissions per ten days just before immer- sion as compared with that of 16.3 fissions per ten days for the first 10 Group c 20 15 10 U 4 4 ‘ t r-o-----24 e- ----5 0 5 IO 15 20 25 30 Fic. 4. Fission rate and total longevity of mature animals after release from long exposure to 48.1 per cent D.O for various lengths of time. Normal controls, Group A ( ); Group © in 48:1 percent D20) (==) = Group c, ce cx andi c: after return to normal medium (———). Abscissee = ten-day periods. Ordinates = mean number of fissions in ten days. 208 EFFECTS OF HEAVY WATER ON FISSION RATE 209 group which was immersed after Period 2, it does not seem probable that this difference would account for the short survival time of Group D, but that the latter is due to the age of the animals upon immersion. Similarly, when a third group (Group E) was placed in 48.1 per cent D,O at the end of Period 23 (Fig. 3), the survival time was less than one period. The longest length of life of any individual was thirteen days while the average was five days for this group. Again after Period 27 when the normal group was very old, similar results were obtained. Derivatives from the animals in 48.1 per cent D,O were placed back into normal medium at five points, i.e., at the beginning of Periods 11, 13, 22, and 23 (Groups c,, ¢, cs, ¢, respectively). Im each case the fission rate rose during the first period of this return (Fig. 4) ; except for Group c, it did not closely approximate the normal rate at the same time. In Groups c, and c, the rate rose to a level which was above that of the normal controls for the same period. Those in the first lot to be removed (c,) were placed in the special control water which was later substituted by the normal medium after Period 13. The other groups were placed directly in normal medium. Since the first group was discontinued after Period 19, no further statements can be made concerning it. None of these released groups which were continued to the end of the life cycle lived as long as the normal control animals. Some were also removed at the end of Period 25; they died within a few days with no divisions, thus following the same course as the ani- mals in the 48.1 per cent D,O at that time. Since all of these groups died before the animals in the normal medium, they must have been slightly weakened by the long contact with the D,O. On the other hand, when derivatives (Group d,) were removed at the end of Period 24 from the group of old animals which had been placed into 48.1 per cent D,O from normal medium after Period 21 (Group D), they lived as long as the normal controls (Fig. 5). The release was accompanied by a stimulation which caused the fission rate to rise to 7.7 during the first ten days while the controls (Group /Al)) were dividing at a rate of only 2.1. However, this high rate declined in the next sixty days so that the animals died at the same time as the normal controls. In spite of their age, the relatively short time of ‘thirty days in D,O had no permanent harmful effects, although the rate was strongly suppressed while in this medium. For a secondary indication that the intermediate concentration changes life processes, a few measurements of the contractile vacuole tate were made. In transferring a normal Uroleptus mobilis to 48.1 210 ARLENE JOHNSON DELAMATER per cent D,O from the normal medium, there was an immediate de- crease in the rate (i.e., within 15 minutes or less). Out of the twelve cases examined, there was one exception in which no regulation occurred and the animal suddenly disintegrated within two hours while still in the heavy water. Whether or not this lack of regulation was the cause of the death cannot be said. The general decrease was greater than any sudden fluctuations which occur in the normal rate. For six ani- mals measured over a period of 4 to 5 hours with observations every 20 minutes in the normal medium, the greatest change was 3.9 seconds in 20 minutes; while upon transfer into 48.1 per cent D,O, the de- crease varies in extent from 6.1 seconds to 21 seconds for different animals. The majority fell between 7 and 15 seconds, although an in- 20 15 igo) 5 10 15 20 25 30 Fic. 5. Fission rate and total longevity of old animals after release from a short exposure to 48.1 per cent D.O. Normal medium, Group 4 ( ). Group D, animals put into D,O when old (——— ). Group d,, animals returned to normal medium after 30 days in D2O (-::). Abscissa = ten-day periods. Ordinate = mean number of fissions in ten days. sufficient number were tested for averages. Consecutive immersions were not tried. Immediately upon immersion, in some cases, the con- traction of the vacuole was very irregular, although within half an hour it became fairly constant. The degree of decrease did not depend upon the initial rate. Therefore each animal seemed to regulate itself ac- cording to its individual state at the time of immersion. The obtained variation in the decrease occurs within a temperature range of 2° C. It is very difficult to assign one value to. the rate while the animal is in any effective concentration of deuterium for a short time because the rate may fluctuate. The initial decrease may not be lasting or con- stant. In considering four animals which were kept in the D,O from four to six hours, one of these animals regained and increased above its normal initial rate in less than three hours after immersion, although EFFECTS OF HEAVY WATER ON FISSION RATE Pll its original decrease upon immersion was as great as that of the other three animals which continued to show a low rate in the D,O until their release. A more extensive experiment would probably show more of this type. Also in the seven other cases examined over a much shorter period of time, there were two which gained the approximate original level in less than one hour and one which showed an increase in the rate just before it was released. Here also the initial change was as great as in the four other cases whose rates did not increase while in the D,O. Whether any of these animals would have reached the original level after four to six hours cannot be told. There is also the possibility that these fluctuations are only preliminary and that over a longer period of time such as 24 hours a marked difference in the 48.1 per cent D,O might show. iABrE i Differences in contractile vacuole rate between animals living in normal water (Group A) (Group c) and those in 48.1 per cent D.O (Group C). Average contractile Number of animals vacuole rate observed seconds Normal controls. Groupware ant acaeis enna ee yews 19.7 13 In 48.1 per cent D,0. (Groups ee enh eae 28.9 27 This possibility is further substantiated by data obtained from measuring the contractile vacuole rates in samples of animals taken from Group A (normal medium) and Group C (48.1 per cent D,O) during Periods 9-15 inclusive, as well as from Group c, (normal medium after eighty days exposure to 48.1 per cent D,O) in Periods 12-15 inclusive. Table I shows that there is a clear distinction be- tween the average rate of animals in normal medium and those in the heavy water; the animals which had been returned to normal water for ten days (Group c,) had already gained the normal rate. Upon removal from exposure to D,O for a short time there was a corresponding change in the contractile vacuole rate which varied in its magnitude for various animals. In nine cases out of the eleven which were tested (above), the change was as great or greater than the high- est variability while the animal was still in the deuterium, so that re- moval as well as immersion may cause a real change in the rate. Varia- DN ARLENE JOHNSON DELAMATER tion was present here also. In most cases the change restored the rate to or near the original level in half an hour or less. High Concentration When normal vegetative animals were immersed in 98 per cent D,O an immediate quick shrinkage occurred, resulting in the formation of spherical or somewhat truncated shapes with the cilia still beating. Usually in five seconds to one minute these balls disintegrated. In a few cases the animals survived in the shrunken state for more than two hours. One animal which became about half of its original size during four minutes in 98 per cent D,O gradually regained its normal appear- ance after transferal to normal medium and lived for thirty hours. Usually there was no recovery from these short exposures. Even al- though most specimens die at once, some distinct variability is apparent in Uroleptus mobilis. DISCUSSION The facts indicate that neither the amount of conjugation or encyst- ment in mass culture, nor its total longevity, nor the fission rate of Uroleptus mobilis during its life cycle are altered by a low concentra- tion of deuterium and they are in harmony with indications from other work that low concentrations neither accelerate nor slow down certain life processes. On the other hand, these results are in contrast both to the stimulating effect upon the increase in population numbers of Euglena gracilis in mass cultures as reported by Barnes (1934), and to the in- crease of the life span of primitive cells such as Spirogyra (Barnes and Larson, 1934). It is interesting to note that very young ex-conjugants from animals which have conjugated in the 48.1 per cent D,O never survive if they are allowed to remain in this medium, but will live if the conjugants are transferred to normal hay infusion. Assuming from this latter fact that the conjugation process in the D,O is regular, which should be checked as far as possible, the high susceptibility of conjugants and their young ex-conjugant derivatives is evident. Jollos has long main- tained that conjugants as well as recent ex-conjugants are more sensi- tive to unfavorable conditions than animals during other stages. How- ever, mature animals such as those which were placed in the 48.1 per cent D,O twenty days after conjugation and subsequent fission to 45 to 48 generations in the normal medium, were more able to withstand the effects of this agent, although it is evident that their period of “senescence ’’ was cut short. Furthermore, old animals, after 328 to EFFECTS OF HEAVY WATER ON FISSION RATE Pals) ! 346 generations in the normal medium, have only a short survival in the deuterium. Although the discrepancy exists that the young ani- mals conjugated in the D,O and that the mature and old animals were derived from conjugation in normal hay infusion, these facts are in general accordance with the effects found by Gregory (1925, 1926) upon very young, mature and old Urolepti by continued treatment with beef extract or di-potassium phosphate, and by brief treatment with di-sodium phosphate, 1.e., that both young and old series are immedi- ately depressed while mature animals are stimulated in these media. Immature and old animals are not as able to adjust themselves to rapid changes in the environment. Thus these results not only show the actual effects obtained with an intermediate concentration of deuterium but also incidentally further substantiate the theory of Calkins (1926) that changes occur in the derived organization of Uroleptus mobilis throughout its life cycle. That 48.1 per cent D,O is clearly of a depressing nature is seen from the facts that both fission rate and total longevity are less in this medium. Also mature animals which have been in contact with this concentration for a long time, as from 90 to 230 days, are weakened. Although an immediate stimulation which may cause the fission rate to rise above that of the normal controls at the same time may occur upon release from heavy water, they do not live as long after they are returned to the normal medium as the control specimens which have never been exposed. On the other hand, even old animals are not weakened by shorter contacts as for 30 days, since they afterwards live as long as the controls; as above, the fission rate may even be temporar- ily stimulated upon release from heavy water to a rate above that of the normal controls. This clearly shows the necessity of carrying out work of this type with Protozoa over a long time in order to obtain the full results which may not be present at an earlier stage. This inter- mediate concentration, in contrast to the low concentration used, has a clearly depressing effect upon the protoplasm of Uroleptus mobilis. The changes produced are not due to the lack of sufficient amounts or types of the bacterial food supply, nor are they due to any injurious impurity in the D,O, since the fission rates of the animals in the special control water, which had been run through the same series of distillates as the D,O, were equal to the normal rate. With the few contractile vacuole rate measurements as a second way of testing the effects of an intermediate concentration, a slight decrease upon immersion is usually found. The great variation in the degree of decrease points to the fact that the evident attempts at regulation are 214 ARLENE JOHNSON DELAMATER a function of the particular individuality of the protoplasm at this time. Each animal regulates itself according to its state, which is unknown to the observer. The decrease in contraction may continue within the first four to six hours, or it may regain the original rate which the vacuole possessed before immersion. However, it might reach a steady decreased rate or a complete recovery in all cases after a longer period such as twenty-four hours. That the former condition is the more likely is seen from the fact that animals which had been running in isolation cultures in the 48.1 per cent D,O for sixty days were con- tracting at an average rate of 28.9 seconds as compared with that of 19.7 seconds for the control lines in normal medium. From the work of Barnes and Gaw on Paramecium caudatum, a decrease was reported after twenty-four hours immersion in 30 per cent D,O. Finally, upon removal from D,O, there may be a corresponding increase in the con- tractile vacuole rate which usually brings it back to or near its original level in those animals whose rate has remained lowered in the heavy water solution. In working with the contractile vacuole rate of Paramecium cauda- tum, Barnes and Gaw (1935) state that “in a race of Paramecium caudatum the contractile vacuole empties every 18.9 seconds in 30 per cent D,O and 11.3 seconds in controls at 18.8° C.” They apply the Arrhenius equation to their data and find that in 30 per cent D,O, the graph representing the log K as a function of 1/T is an unbroken line throughout the entire temperature range which was used, and that E& (or the energy of activation of the controlling catalyst) equals 22,000 calories. Moreover, they found that in normal medium, the rate varies with the temperature, and from the rates assigned specific values to E at these temperature ranges. Since E below 16° C. was similar to the E in 30 per cent D,O, they concluded that deuterium has similar effects to those of low temperatures. In the first place, the extremes of 18.9 seconds and 11.3 seconds for contraction are within the limits of normal variation for Paramecium ‘caudatum. The normal extremes can be far greater than the differences here stated by Barnes and Gaw. In studying the extremes of these variations, Putter (1900), for example, found that in a total of 1,100 pulsations observed, the interval between pulsations varied from 6 sec- onds to 43 seconds; in the same animal, they varied as much as 21 seconds. Frisch (1935) further observed among other facts that swim- ming, for example, produces a marked decrease in the rate. The de- crease begins suddenly when swimming commences and stops suddenly when it stops. Averages for several animals gave the results of 27.5 EFFECTS OF HEAVY WATER ON FISSION RATE Dries seconds at rest before swimming, 157.7 seconds swimming, and 28.1 seconds at rest after swimming. Furthermore, the rate varies when animals are at rest, depending upon the time of feeding. Secondly, it is difficult to apply the Arrhenius equation to the con- tractile vacuole rate due to this variation. At the present time, we do not know the nature of the reaction or reactions which govern the con- tractile vacuole discharge. All that we can say is that the filling of the vacuole and its discharge may be the results of numerous types of reac- tions, of a possible catalytic nature; whether the same causative reactions are affected by deuterium and by temperature cannot be stated. Finally Barnes and Gaw make no designations as to which of the two vacuoles in Paramecium were observed. Fortner (1924) as well as Frisch (1935) have noted that the posterior vacuole contracts more frequently than the anterior one, although in a few animals the reverse was true. Among a group of twenty animals, Frisch shows in his Table I (p. 13) that the variation in the difference of contraction may range from .2 second to 13.3 seconds in cases where the posterior vacuole is contracting the more rapidly. Since so little is known at present about the function or regulation of the contractile vacuole as a typical but not universal character of the protozoan cell, it does not present itself as a favorable function for the basis of suitable experiments to ascertain the effects of D,O. In considering the effects of a high concentration, this protozoan, like other organisms, cannot survive for more than a very brief time in 98 per cent D,O. In contrast to Euglena gracilis (Harvey) an ir- reversible injury is present. CoNCLUSIONS 1. No indications were found for the stimulation or depression of the fission rate of Uroleptus mobilis in 0.44 per cent D,O during that part of the life cycle which covered about ninety days. 2. On the other hand, with a concentration of 48.1 per cent D,O both the fission rate and the total longevity were decreased. The degree of the decrease depends to some extent upon the age of the animals since the previous conjugation. Young and old animals survive for only a short time such as thirty days after immersion into 48.1 per cent D,O; while mature animals live for many days. This shows the necessity of knowing the history and of controlling the experimental material in such organisms as Protozoa. Short exposures to 48.1 per cent D,O such as for thirty days even with old animals have no harmful effects, as evidenced by the fission rate and the length of life after 216 ARLENE JOHNSON DELAMATER return to normal water, whereas long exposures such as eighty days with mature animals result in a decrease in the total longevity. 3. In general, a lowering of the contractile vacuole rate occurs in 48.1 per cent D,O. The extent of the decrease varies from animal to animal. However, due to the small amount of knowledge upon the exact causative reactions which govern the contraction as well as the normal variation in the rate for different animals, the function is not favorable for study in this respect in this protozoan. 4, Ninety-eight per cent D,O causes an irreversible injury resulting in immediate death, although there is evidence that some few animals can withstand the effects for a few hours longer than the majority. BIBLIOGRAPHY Barnes, T. C., 1934. The effect of heavy water of low concentration on Euglena. Science, 79: 370. Barnes, T. C., 1935. The biological effects of heavy water. Am. Jour. Sci., 30: 14. Barnes, T. C., anp H. Z. Gaw, 1935. The chemical basis for some biological ef- fects of heavy water. Jour. Am. Chem. Soc., 57: 590. Barnes, T. C., AND T. L. JauHn, 1934. Properties of water of biological interest. Quart. Rev. Biol., 9: 292. Barnes, T. C., AND FE. J. Larson, 1934. The influence of heavy water of low con- centration on Spirogyra, Planaria and on enzyme action. Protoplasma, 22: 431. Carxins, G. N., 1919. Uroleptus mobilis Engeim. II. Renewal of vitality through conjugation. Jour. Exper. Zool., 29: 121. CALKINS, G. N., 1926. Biology of Protozoa. Lea and Febiger. Curry, J., R. Pratt, anv S. F. Trezease, 1935. Does dilute heavy water influence biological processes? Science, 81: 275. Fortner, H., 1924. Uber die physiologisch differente Bedeutung der kontraktilen Vakuolen bei Paramaecium caudatum Ehrenberg. Zool. Anz., 60: 217. Fox, D. L., 1934. Heavy water and metabolism. Quart. Rev. Biol., 9: 342. FriscuH, J. A., 1935. Contractile vacuole of Paramecium. Regulation and func- tion. . Dissertation, The Johns Hopkins University Library. (To be pub- lished. ) Grecory, L. H., 1925. Direct and after effects of changes in medium during dif- ferent periods in the life history of Uroleptus mobilis. Biol. Bull., 48: 200. Grecory, L. H., 1926. Effects of changes in medium during different periods in the life history of Uroleptus mobilis. Biol. Bull., 51: 179. Harvey, E. N., 1934. Biological effects of heavy water. Biol. Bull., 66: 91. Putter, A., 1900. Studien tiber die Thigmotaxis bei Protisten. Arch. f. Anat. u. Physiol., Physiol. Abt., Supplbd., S. 243. (See Frisch. Original article not read.) Taytor, H. S., W. W. Swincie, H. Eyrine, Anp A. H. Frost, 1933. The effect of water containing the isotope of hydrogen upon fresh water organisms. Jour. Chem. Phys., 1: 751. Pit SOC UR AN DEF UNCRON OF TH GUE INSURE PERCHES (EMBIOTOCIDZ) WITH REFERENCE TO THEIR CAROTENOID METABOLISM+ ROBERT T. YOUNG AND DENIS L. FOX (From the Scripps Institution of Oceanography of the University of California, La Jolla, California) In examining freshly-caught surf perches (Embiotocide) for trema- todes, one of us (Young) encountered an interesting yellow or orange pigmentation of the gut, especially in the rectal segment of these fishes. Chemical and spectroscopic studies revealed the carotenoid (specifically xanthophyllic) character of much of this pigment. In freshly-caught specimens of these fish and of the spot-fin croaker (Roncador stearnsi) the rectum is frequently but not always deeply pigmented throughout its length to within a few millimeters from the anal opening, ranging in color from a brilliant orange to a deep brown. The upper intestine may also be colored to some extent, but it is the rectum which is most strik- ing. This observation suggested to us an experimental study of the role of the gut in the metabolism of these pigments, and for this purpose we selected Cymatogaster aggregatus,? a small perch which occurs abun- dantly in San Diego Bay and elsewhere along the west coast of North America. It has an average weight of 15-20 grams.* It is silvery in color with yellow markings on the sides of the body, which alternate with dusky bars. The description of the gut which follows is based partly on that of Cymatogaster and partly on that of Embiotoca jacksom and Abeona minima. The structure being very similar in all, the description of one will serve equally well for that of the others. Several of the standard fixatives were used for the histological work including Bouin, Zenker, Fleming and Carnoy; while paraffin sections 1In this work, Young was responsible for the descriptive and histological por- tions, while the chemical analyses and spectroscopic observations were made by Fox. The experimental work was carried out jointly. 2 A number of observations and experiments have also been made on other spe- cies, but Cymatogaster was the fish mainly used, and the one selected for all the quantitative studies reported here. 3 Jordan and Evermann give its length as 6 inches but this is much above the average of our specimens, which is about 8-10 cm. Divi 218 ROBERT T. YOUNG AND DENIS L. FOX were stained mainly in Delafield and Haidenhain’s hematoxylin, counter- stained with eosin, and a few in Giemsa, as recommended lyaelbeer With the latter stain we obtained some very fine results, especially of the mast cells and of the ganglion cells in the myenteric plexus. Besides studying the vascular supply of the gut in fresh material, we also injected two specimens of Embiotoca with India ink, one of which gave a very successful result. The general anatomy of the digestive tract in surf perches is shown in Fig. 1. The bile duct enters the stomach just behind the pharynx, so that a true stomach is wanting. The rectum forms about one-tenth of the total length of the intestine, and is demarcated from the rest of the latter by a distinct sphincter. Ss L BG Fic. 1. Digestive tract of Abeona X 4. B, bile duct; G, gall bladder; L, liver; RK, rectum; S, sphincter. Internally the lining of the gut is similar throughout. It is elevated in narrow folds or ridges that follow an irregular, zig-zag course and in general run parallel to the walls of the gut, but whose form and direc- tion is largely determined by the condition of the organ itself. If the latter is filled with food, the folds flatten out and assume a more trans- verse direction; if the gut is empty the folds heighten and follow more nearly a longitudinal course. At the apices of the angles formed by their bends the folds frequently send out offshoots which may end blindly, or occasionally anastomose with neighboring ridges (Fig. 4). The blood vessels of the gut have been studied mainly in Embiotoca jacksoni instead of Cymatogaster, because of the greater ease of in- jecting them in the larger fish. After penetrating the two muscular layers of the gut wall the vessels branch frequently to form a plexus at the base of the mucosa (Fig. 2) whence smaller twigs extend through the latter to terminate in a very fine plexus at the summit of the folds, just below the epithelium (Fig. 3). In sections prepared from a specimen of Embiotoca injected with India ink, these twigs can be followed in 4 Vade Mecum, Ed. 9, p. 503. CAROTENOID METABOLISM IN SURF PERCHES 219 4 5 Fic. 2. External vascular plexus from an injected specimen of Embiotoca. Fie. 3. Internal plexus of same. Fic. 4. Intestinal ruge of Cymatogaster, from a tangential section. The stratum granulosum is the loose layer between the epithelial folds, while the com- pactum is the darkly stained line at their base. An anastomosis between two ruge is shown at X. Three trematodes are seen in the lumen between the ruge. X 130. Fic. 5. Section of a ruga from the rectum of Cymatogaster. S, stratum com- pactum; G, stratum granulosum with mast cells. X 560. 220 ROBERT T. YOUNG AND DENIS L. FOX sections close to the epithelium but careful search fails to reveal any penetration of the latter by the capillaries. There are several accounts in the literature, however, of the penetration of the epithelium by capil- laries in the gut of the Cobitide.® The wall of the gut is similarly constituted throughout, so that it is very difficult to tell what region any section represents. The absence of a muscularis mucose removes any distinction between mucosa and sub- mucosa, regarding which considerable confusion exists in the literature (see Blake, 1930, and Rogick, 1931). Greene (1912) divides this layer (mucosa + sub-mucosa) in the in- testine of the salmon, where a muscularis mucosz is absent, in contra- distinction to the stomach where it occurs, into three sub-layers: tunica propria, stratum compactum and stratum granulosum. Both of the former authors describe a basement membrane of the epithelium which (see Blake, Joc. cit., p. 53) is formed from the tunica propria in the stomach of Centropristes. In our own material basement membrane and stratum compactum are indistinguishable, the latter being well developed and sharply set off from the rest of the connective tissue, or stratum granulosum (Fig. 5). This is so named by Greene because of the presence within it of large numbers of granular cells whose “ special and characteristic feature . . is the presence of the cytoplasmic granules that always crowd them, unt- formly filling the cell body ”’ (doc. cit., p. 83). In the salmon these cells are not restricted to the stratum granulosum, but occur also in the stratum compactum, and occasionally in the tunica propria and the sub-mucosa. Greene suggests that these cells form an internal secretion, probably lipase, which 1s discharged “ directly into the surrounding tissue spaces from which distribution takes place’ (loc. cit., p. 84). These cells have been described in detail by Bolton (1933) as mast cells whose “true function seems as yet to be as obscure as that of the mast cell of the higher vertebrates .. . (but which) may play some role in connection with the elaboration of zymogen in these organs” (loc. cit., pp. 576-7). In our material, cells which are undoubtedly the same as those de- scribed by these authors occur in considerable numbers scattered through the connective tissue of the mucosa. In some places they occur in groups, in others as isolated cells. They are most abundant in the rec- tum, but are widely distributed throughout the gut. In places the gran- ules are scattered through the connective tissue, as though set free by the rupture of the cells, but for the most part, the cells are discrete, varying in shape from globular or ovoid to ameboid. Whether they > See especially in this connection Abolin (1924). CAROTENOID METABOLISM IN SURF PERCHES Papa have any relation to metabolism of the carotenoids in the gut, we cannot say. The carotenoids are found in the mucosa of the gut, as is easily demonstrated by washing this tissue with a fine jet of water applied with sufficient force to remove it in patches, leaving it intact elsewhere. Where the latter has been removed the pigment is absent, but is present elsewhere. And when the mucosa begins to break down post mortem the pigment is readily washed out with it. Whether the pigment is restricted to the epithelium, or is present in the connective tissue also we do not know. However, when a brightly pigmented rectum is dis- sected so as to expose the mucosal folds, the pigment appears as lines or streaks in the latter, and since these are composed mainly of epi- thelium it is probable that this is the chief, if not the only locus of the pigments. The source of the pigment is the food. The diet of the surf perches is a varied one, consisting of mussels, clams, barnacles, worms, many kinds of Crustacea and even fish. Cymatogaster, on which our experi- ments were mainly performed, feeds extensively on little green and brown shrimps (Hippolyte californiensis) ®° together with a species of Crangon, copepods, etc. When brought into the laboratory from the bay the recta are characteristically bright orange in color, while the upper intestine is salmon colored, fading out posteriorly, however, so that the last half centimeter or so is usually colorless, or very pale. The color of the upper intestine is more or less affected by the rich vascular supply of this region, so that in a faintly pigmented gut it is not always easy to determine how much of the color is due to carotenoid and how much to blood pigments. In other surf perches (Embiotoca, and Hy- perprosopon) taken from the surf, and in specimens of Abeona which have been kept in aquaria and fed upon various diets, or left unfed, the color of the rectum varies from yellow, through orange to olive or brownish, depending upon the kinds of pigment, derived from food or other sources, or possibly upon chemical changes which the pigments may have undergone. The rectal pigments which depart in color from pure yellows or orange are not carotenoids, as was demonstrated on several occasions by treating the ground tissue directly with lipoid solvents which failed to extract pigment. Water, or alcohol extracts assumed a pale cloudy. yellow tint, but no pigment could be transferred from these solutions to ligroin (Eastman Kodak Co., B.R. 70-90° C.). By no means all of the available carotenoid pigment in the food is 6 The authors acknowledge their thanks to Dr. Martin W. Johnson, of the Scripps Institution of Oceanography, for verifying the identity of this shrimp. DDD, ROBERT T. YOUNG AND DENIS L. FOX absorbed, for the partly digested food and the feces of fish which have been fed shrimps are always colored, and brightly colored bits of Gari- baldi tissue may be found in the feces of fish which have been feeding thereon. Even in a fish whose intestine exhibits no appreciable color the food may retain much of its pigment, as was shown by an Abeona whose gut was practically colorless, but which contained the brightly colored remains of a shrimp. Carotenoids may be recovered in con- siderable quantity from the feces of perches which consume shrimp and other crustacean food. (See below.) QUALITATIVE AND QUANTITATIVE STUDIES OF THE YELLOW PIGMENTS Several chemical analyses and spectroscopic observations * were made of the carotenoid pigment or pigments in each of the following: the en- tire intestines or the recta alone, the gut contents and feces, the shrimps upon which the perch feed, and the fish bodies minus viscera. Rectal Segments A number of these were removed from several specimens (some freshly caught and others having been fed in the aquarium for some days their natural diet of shrimps) and analyzed.* Trituration with a little sand and absolute methyl alcohol caused this orange colored tissue to yield its carotenoid material promptly and completely to the solvent, rendering it yellow orange in color. The pigment remained hypophasic, i.e., dissolved entirely in the lower aqueous-alcohol layer when shaken with mixtures of 90 per cent CH,OH and ligroin, whether the alcoholic solution had or had not first been hydrolyzed for some hours in warm alcoholic KOH. Acidification or alkalization did not alter this charac- teristic. The pigment was easily transferred from the alcoholic solu- tion (neutralized to preclude the formation of stubborn emulsions if necessary ) to ligroin by adding sufficient water and shaking the mixture. It was quantitatively removable from alcohol-free ligroin solutions by adsorption on a column of dry, powdered CaCO,, upon which it exhib- ited a pale yellow-red band at the top of the column, with a rose-colored band just beneath. 7 Spectroscopic observations in this work were made with the large Bausch and Lomb spectrometer No. 2700. Absorption maxima were estimated, and the average of a number of readings taken, in each determination, since the accessory spectro- photometric equipment for securing more precise data was not at hand at the time that this work was done. 8 In some analyses the whole intestines, from esophagus to anus, were used; in others, the rectal segments only. Often the intestine exclusive of the rectum ap- peared quite pale and yielded little or no pigment. Even when the absorbent por- tion of the gut was distinctly pigmented, no differences were detected in the nature of the extracted carotenoid material. CAROTENOID METABOLISM IN SURF PERCHES 223 The carbon disulphide solution of the unseparated pigments was deep red orange in color, and exhibited absorption bands whose maxima averaged approximately: I 499.4 mp, II 471.6 mu. Ethereal solutions of the pigments shaken with concentrated HCl remained colorless. The foregoing properties of the pigment in the intestinal tissue (especially concentrated in the rectum) indicate that it is a carotenoid, or more probably a mixture of two carotenoids judging from the two rings in the CaCO, column, of a xanthophyllic nature, and unesterified. The absorption spectrum corresponds rather closely to those of taraxan- thin and violaxanthin (Zechmeister, 1934; Lederer, 1935) of the C,,H,,O, series; the additional fact that concentrated HCl imparted no blue color to ethereal solutions of the pigment indicates that the pre- ponderant carotenoid is not violaxanthin but probably chiefly taraxanthin or a carotenoid indistinguishable from the latter by these various tests. No attempt was made to separate the adsorbed pigments in the chromatogram, since the total quantity of available pigment was very small, and the paleness of the primary ring indicated a relatively small quantity of this fraction. Furthermore, the spectroscopic picture of the rectal pigment was nearly identical with that of the shrimp xantho- phyll, that of the fish skin xanthophyll, and with the body xanthophylls of other species of fish (see Fox, 1936). Why the rectal and shrimp xanthophyll chromatogram showed a red component, while the skin xanthophyll (hydrolyzed) chromatogram consisted in a yellow ring is yet unexplained, but it is possible that the xanthophyll stored in skin is either altered chemically without materially shifting the absorption bands, or rigidly separated from another carotenoid having a very similar spectrum. Intestinal Contents; Feces No differences were detected in the nature of the carotenoids of the ingested food removed from the gut and those found in voided feces. Absolute methyl alcohol extracted quantitatively all of the yellow col- ored fat-soluble compounds. Shaking a 90 per cent CH,OH solution of the pigments with an equal quantity of ligroin resulted in most of the colored compounds remaining in the lower (alcoholic) layer, with some pigment being transferred to the ligroin phase above. Selective solubility in ligroin is characteristic of carotenes, esterified xanthophylls, free (neutral or acidic) astacene, and certain astacene esters. (See Zechmeister, 1934; Lederer, 1935; Fox, 1936.) That no free astacene was present in the ligroin was demonstrated by adding a drop of strong lye to the aqueous alcohol layer which did not, upon shaking, remove any 224 ROBERT T. YOUNG AND DENIS L. FOX pigment from the ligroin phase. A portion of the whole extract was hydrolyzed in warm alcoholic KOH, usually for some hours, after which three components were easily separated, whose descriptions fol- low. (1) A red substance, insoluble in ligroin or absolute CH,OH, re- mained in flocks on the bottom of the flask. When collected on filter paper and washed several times with alcohol and ligroin the material was finely divided and bright red in color, resembling that of ripe tomatoes or red peppers. It remained completely insoluble in neutral ligroin, CH,OH, or CS, but was readily dissolved in any of these solvents if acidified by adding a trace of glacial acetic acid. It exhibited, there- fore, the properties of a fat-soluble acidic compound capable of forming a potassium salt which was insoluble in the various organic solvents. The acidified product yielded bright yellow ligroin solutions, salmon- colored CH,OH solutions, and a beautiful rose pink to magenta color in CS,. It was hypophasic when shaken with equal quantities of ligroin and 90 per cent CH,OH. Dilution of the alcohol layer caused the free pigment to be transferred to the ligroin phase. Solutions in ligroin or CS, (always freed from the last trace of alcohol and acetic acid) when passed through a CaCO, column left the pigment quantitatively adsorbed to this salt in a single rose-colored ring at the top of the column. It was readily desorbed, as are the xanthophylls, by ligroin containing a trace of CH,OH. This pigment was even adsorbed to some extent by pure, dry, crystalline NaCl imparting to this salt a delicate rose tint. It was desorbed by wetting the salt with a little water. The pigment proved to be rather unstable chemically; when recov- ered from solvent and stored dry in a partially evacuated chamber, the red color faded completely in a few days. The deep red solution of this rather unusual, acidic carotenoid in CS, solution gave a very poorly- defined absorption shadow down in the violet, the average maximum absorption region, taken from numerous readings, lying close to 499.5 my, which is somewhat lower than that of astacene (510 mp in CS, ac- cording to Lederer, op. cit.). While also acidic, this compound differs markedly from astacene in other ways as well as in its absorption maxi- mum, viz.: its preferential solubility in alcohol when subjected to parti- tion between alcohol-ligroin mixtures; the insolubility of its potassium salt in alcohol (astacene, preferentially soluble in ligroin, can be trans- ferred immediately to the alcohol layer by adding a drop of lye and reshaking (Zechmeister, of. cit.) ), its instability in air, and its quanti- tative adsorption from ligroin or CS, solutions by CaCO,. Were it not for the acidic character of this red carotenoid, it would correspond rather closely to Lederer’s glycymerine, which is neutral, shows a single CAROTENOID METABOLISM IN SURF PERCHES 225 absorption maximum at 495 mp in CS,, preferential solubility in 90 per cent CH,OH, adsorbability on CaCO,, and other characteristics which liken it to the xanthophyll group of pigments. Until further study may provide more information concerning this interesting pigment, no pro- visional name will be assigned to it. (2) When the filtrate from the crop of red flocks was shaken with ligroin, a relatively small quantity of yellow pigment remained in the ligroin layer and was not removable therefrom by subsequent washes with 90 per cent CH,OH, whether neutral, acidified, or alkalized. The ligroin solution was freed from alcohol and acetic acid with re- peated water washes and then tested chromatographically by passing through a CaCO, column. No pigment whatever was adsorbed. Transferred to CS,, after evaporating the ligroin in a current of illumi- nating gas (to prevent bleaching from oxidation), the solution had an orange color, and exhibited absorption bands whose average maxima lay respectively at: I 511.7 my»; II 480.9 mp, which values correspond closely to the first two principal maxima in the visible region for CS, solutions of a-carotene (i.e., I 510 my; Il 478 mp) and depart some- what from those of @-carotene (I 521 mp; II 485 mp).® That it is a carotene is shown by its neutral character, its preferential solubility in ligroin instead of in 90 per cent CH,OH, and its non-adsorbability on CaCO. Its absorption spectrum would indicate that the preponderant part of it was closely allied to, if not identical with a-carotene. (3) The hypophasic (alcohol-soluble) portion of the original filtrate of the hydrolyzed mixture yielded no further pigment to washes of ligroin applied subsequently to the first separation. After transference to ligroin by diluting the alcohol phase, shaking, separating, and finally washing the ligroin solution free from all traces of alcohol, the pigment was quantitatively adsorbed by CaCO, most frequently as a single yel- low band (sometimes as a reddish band) at the top of the column. The orange colored CS, solution of such pigment fractions gave absorption bands with average maxima readings at the following points: I 496.9 mp; Il 467.4 my. In its neutral character, hypophasic property, and its quantitative adsorption upon CaCO,, this pigment, like that of the intestine, showed xanthophyllic properties. Also, its absorp- tion spectral bands showed maxima which agreed rather closely with those of the intestinal pigment although the latter’s maximal readings were placed slightly more toward the red. While the intestinal pig- ment always showed two rings, i.e., a yellow, and a red one just below it in the CaCO, chromatogram, the xanthophyll fraction of the feces 9 These figures are for visual spectroscopic readings, but see Smith (1936) re- garding photoelectric spectrophotometric methods which give shifts in the positions of absorption bands. 226 ' ROBERT T. YOUNG AND DENIS L. FOX and food material in the gut did not always show two such rings, but frequently only the yellow one. The presence of the red ring as well as the yellow, indicating another compound, very possibly depended upon the stage of digestion and relative degree of absorption of the com- pounds which may have already occurred by the time that the feces or intestinal contents were collected. Since this feature in the chromato- gram characterized the xanthophyll of both the intestinal tissue of the fish and the whole shrimps which formed its raw food, and hence the source of the carotenoids, one might have expected to find some such inconsistency in the relative amounts of the xanthophylls in the corre- sponding fraction of the food boli or feces. The red component in the chromatogram might have been either another xanthophyll or an incom- pletely hydrolyzed ester, since the esters were shown to be less readily sorbed by CaCO, and to show various color variations between yellow and red in a chromatogram. Shrimps These were gathered from the same locality as that from which the perch were taken, constituted the natural diet of the latter, and were taken from the lot which were fed the experimental animals whose tis- sues were subsequently analyzed. Two or more species were repre- sented, but the predominant form was Hippolyte californiensis. Some two dozen of them were dropped into pure CH,OH, whereupon they quickly lost their original color, whether brown, green, or dull red, and exhibited bright red patches upon various parts of their bodies. They were ground in a mortar with sand, and yielded all their carotenoid pig- ments to the alcohol. All pigment was quantitatively epiphasic, being readily extracted from the alcohol by ligroin; shaking the ligroin solution of carotenoids with additional 90 per cent methanol did not cause any transfer of pig- ment back to the latter solvent. After hydrolysis for some hours in warm alcoholic KOH, the shrimp pigments were, as in the extracts of the intestinal contents or feces, resolvable into three components, which were virtually identical in chemical and spectroscopic properties with the respective components of the latter, as follows: (1) A red, acidic carotenoid, whose potassium salt was insoluble in alcohol, ligroin, or CS,, but which when free was readily soluble in any of these ; hypophasic property in ligroin + 90 per cent methanol systems ; quantitative adsorption by CaCO, as a rose-red ring at top of column; poorly defined absorption band with maximum at approximately 494.5 mp. 227 CAROTENOID METABOLISM IN SURF PERCHES Sjustsid JO SUOT}e1]UIIUOD dAT}EIII BJeUNXOIdde AjTaIOW ynq so r “((uoszeM “MM “D Aq uMeIq) Sunox “[ ‘y pue xoy “7 ‘q) yWUeND synjosqe jou ‘sJouUeW [eUOT}UBAUIOD sy} UI Yuesetdat SassoId OT, ‘suspydogiy st ‘pazsjorpAy ussq sulAey “uoursid oy) Jey} sojouap (#7) { paylseyso ‘oudj01"9 fo SI JUoUIsId oY} yey} so}ouep (7) ‘snpohaihbn sajspbojnmi) yoied jans ay} Aq paysasut sprousjo1e9 durstys fo a3ef YT, “9 “DIY AYVOLISNYY L ++ (H) SITTAHGOHINYX 334 S3NSSiIL WWL33y goo01@ VIA AIN3NVWY3d ++ (3) TVAHPOHIANVX C3131 31S53 3Sa dil ainisisinn (‘913 ‘NIMS) S3NSSIL AQgo@ SANssit 24 hour station. Observations every 4hours. August 2Oand 21, 1935. 350 fe. 3100fc. 3200fc. loooft. © fe. 0 fe. 150 Fe. INTENSITY OF LIGHT (FOOT CANDLES) Fic. 1. This figure shows the vertical distribution of Gonyostomum semen from the surface to the bottom of Cedar Pond during two 24-hour periods. All the observations recorded in this figure are for the vertical distribution at Station D. In each observation the number of organisms in 2 ml. of water are shown. The degree of illumination in foot-candles at the surface of the water is shown for each observation. bottom. Ordinarily there is a sharp decrease from the surface to the 0.5-meter level after which the decrease is much less, more gradual, and somewhat irregular, as can be seen from Table I. A daily increase in dissolved oxygen has been found to occur on sunny days during the morning or morning and afternoon, especially at 0.5 meter (Table 1). It is at the 0.5-meter depth that the maximum numbers of individuals of Gonyostomum (Fig. 1) were found during 292 ROPe COWEES AN Dy © FE Bin AM Bi Ie the morning hours and it is reasonable to suppose that the increase in O, is due largely to the photosynthetic activity of that organism, although another supposedly chlorophyll-bearing flagellate found associated with Gonyostomum may possibly be another small source. It will be noted from Table I that the oxygen content decreases fairly regularly at the 0.5-meter level a short time after the Gonyostomum individuals begin their descent to the lower layers during the late afternoon (Fig. 1). The average of the dissolved oxygen values on August 14 and 15, for Station D, from surface to bottom, at 1 P.M., as compared with the average at 1 A.M., are 2.7 p.p.m. and 2.1 p.p.m. Similar averages for August 20 and 21 are 2.7 p.p.m. and 2.0 p.p.m. These differences be- tween the O, values at 1 P.M. and 1 A.M. are presumably due in a large part to a change in the photosynthetic activity of Gonyostomum. The dissolved O, of the water of this bog pond and especially of the surface and 0.5-meter levels is considerably depleted early in the morn- ing, e.g., 1 A.M. (Table I), due to the respiration of organisms and oxi- dation of organic matter in the water and at the bottom. During the morning or the morning and afternoon the dissolved O, in the water increases as the result, mostly, of the photosynthetic activity of Gony- ostomum. The rise of temperature during this time favors the escape of O,, but notwithstanding this, the increase in O, or “ oxygen pulse,” as it is sometimes called (Welch, 1935), is distinctly evident. The data obtained with a Weston photometer for the illumination at the surface of the water are shown in Fig. 1 and Table I. The relation of these values to the distribution of Gonyostomum will be taken up in the next section. DISTRIBUTION OF GONYOSTOMUM On an exceptionally bright morning it was found that at seven dif- ferent stations distributed along a line extending across the bog pond the largest counts of swimming individuals of Gonyostomum were at a depth of 1.0 meter (Fig. 2). Judging from this widespread similar vertical distribution in the brownish colored water of this pond, it would seem then that under the physical and chemical conditions of that day and that time of day, a 1.0-meter depth from the surface offered the optimum conditions for these organisms. On other somewhat less bright mornings the maximum numbers were found at a depth of 0.5 meter (Fig. 1), while on a cloudy morning the maximum counts were at the surface. Muttkowski (1918) finds that the level for opti- mum photosynthetic activity for green plants in Lake Mendota, which is a clear, uncolored lake, is between 3 and 5 meters. Juday and ENVIRONMENTAL CONDITIONS IN A BOG POND 293 Schomer (1935) obtained similar results, by different methods, for other lakes. The dependence of Gonyostomum on light as a stimulation for mi- gration upward is strongly indicated, as one can see from the results shown in Fig. 1. At 5 o’clock in the morning, with only 200 to 300 foot-candles of illumination at the surface of the water, the Gonyosto- mum individuals were distinctly massed in the region of 0.5 meter below the surface and at 9:00 in the morning and at 1:00 in the afternoon, with an increasing illumination resulting in 3,000 foot-candles, 0.5 meter still seemed to offer the optimum conditions. But even at 1:00 P.M., on August 20, the counts indicated downward migration, since the number of organisms at 0.5 meter has decreased a little, while at Sta. A B C C, D E F G 10 M 00 Numbers of organisms in 2ml. of water are shown Fic. 2. This figure deals with the vertical distribution of Gonyostomum semen from the surface to the bottom of Cedar Pond from 9 A.M. to 10 A.M. on August 25, 1935. The vertical distribution is shown at Stations 4, B, C, C,, D, E, F, and G which lie along a straight line extending across the middle of the pond. The day was exceptionally bright and there were no clouds. nearly all of the lower levels the numbers have increased. By 5 P.M., with a surface illumination of only a few foot-candles, the migration downward is indicated again by the higher counts at the lower levels and this condition continues at 9 P.M. when the photometer failed to show any appreciable illumination. It may be seen that in the case of both of the twenty-four-hour periods during which studies were made, at 5 P.M. and 9 P.M., the bottom (3.0 meter) counts were the largest when com- pared with those for any other time during the whole twenty-four hours at that depth. This increase in the lower layers with a corresponding decrease in the upper layers is almost positive proof that there has been an extensive migration or a quiescent sinking of the individuals of Gony- ostomum toward the bottom. In this paper the bottom of the pond is spoken of as being at 3.0 meters—the lowest depth at which satisfactory 294. R. P. COWLES AND C. E. BRAMBEL samples could be taken in routine work. Between this depth and 3.5 meters, where there is a somewhat solid bed of decaying vegetation, there is a mush of dark brown mud and frequently of vegetation where, ordinarily, it is a very laborious process and one of questionable accuracy to make counts of organisms. It is quite surprising to find that by 1 A.M. both on August 15 and August 21, individuals of Gonyostomum began to increase again in the upper layers and decrease at the lower levels (Fig. 1), indicating an upward migration. This migration, at 1 A.M., was more marked on August 15 than on August 21 and correlated with this difference, our records show that there was bright moonlight on the night of August 14-15, while on the night of August 20-21 there was much fog although at times the moon was to be seen. While our observations concerning the effect of moonlight on the behavior of Gonyostomum are meager, they suggest that these organisms react to the moonlight and migrate toward the surface, when there is moonlight during the early morning hours. Such a suggestion as to the cause of this behavior is supported by the observations just mentioned, concerning the illumination due to moonlight on two different nights, as compared with the numbers of Gonyostomum individuals to be found in the upper levels of this pond. In the opinion of the authors, the observations:made and our knowl- edge of the structure and behavior of Gonyostomum indicate that light is an important factor in the upward migration of that organism. It would seem that having spent several hours in darkness and, as a result, having probably become exceptionally sensitive to light, it tends to migrate upward till it reaches the proper combination of illumination, free CO, and O, content, and temperature, which are necessary for the required photosynthetic activity. The downward movement in the early afternoon, sometimes as early as 1 P.M. when the light is still bright, is rather surprising. One might suspect that having carried on the process of photosynthesis long enough to satisfy their carbohydrate needs, the Gonyostomum individuals swam or sank quiescently toward the bottom. Drouet and Cohen’s (1935) observation and our own have shown that they do sink to the bottom of jars in the laboratory. We are dealing with a periodic phenomenon here—a migration to the upper levels during the morning and a movement toward the bottom in the afternoon and evening. The theory that this may be regulated by some internal condition of the organism, by some periodic physiologi- cal change which regulates the periodicity in behavior, must not be dis- regarded. There is considerable evidence to show, in the case of cer- ENVIRONMENTAL CONDITIONS IN A BOG POND: 295 tain snails and green unicellular forms, that periodic changes in behavior which are correlated with definite periodic changes in the environment occur even when the organisms are no longer subjected to these changes in the environment. Mast (1920), studying a euglenoid form, Lepo- cinchs texta, found a periodic reaction toward light and it is very prob- able that the periodic movement of Gonyostomum away from the light, at least, involves a periodic physiological change in the organism—pos- sibly a cessation of photosynthetic activity when sufficient sugar has been formed, such as we have mentioned above, or a reproductive change or both. In fact, it is known that Gonyostomum sinks to the bottom and undergoes longitudinal division. It has been shown by several investigators (see Mast, 1911) that changes in temperature may bring about reversal in the direction of the response to light, but the diurnal changes in the bog pond under con- sideration do not seem to be important since the decrease in the water temperature as day changes to night is not very great at any one level except at the surface, where, as our studies show, there are ordinarily few individuals of Gonyostomum. However, the range of temperatures from surface to bottom is large (Table 1) and it is evident that Gony- ostomum is subjected to rather large changes in water temperature dur- ing its movements, so that since a higher temperature often causes green, flagellate forms to react positively to light, the temperature may be a factor which helps to determine the upward migration. Since we are dealing with an organism which carries on photosynthe- sis, water temperature must necessarily be an important factor, for it is known that photosynthesis takes place within certain temperature limits. It is surprising to find that during the daytime the congregation of Gony- ostomum individuals in large numbers along a vertical gradient almost invariably takes place in regions where the temperature is about 24° C. It may be assumed, although not proved, that during the daytime where this congregation in large numbers occurs, photosynthesis is at its height. Numerous observations, not recorded in the text, in which the organisms under discussion were found congregated together during the daytime in large numbers in water of about 24° C., point strongly to temperature as being a factor in determining the vertical distribution, but it does not rule out other factors such as the sunlight, reproductive periodicity, CO, con- tent and O, content of the water, etc., because we find that during part of the twenty-four hours, namely, the hours of darkness, when photo- synthesis has stopped or has been reduced to a minimum, the congrega- tion of Gonyostomum individuals in largest numbers does not take place in water of a temperature of 24° C. There is no intention of offering these results as proof that Gony- 296 Rae. COWLES ANDEC] ES BRAM BIL: c ostomum reaches its “ optimum” photosynthetic activity at about 24° C., since we have no experimental work to support it and since it is well known that such a so-called “ optimum’ may vary with the amount, intensity, or quality of other factors concerned in photosynthesis. How- ever, our water temperature records must be taken into consideration in any attempted explanation of the vertical distribution of this organism. Without having determined experimentally the amount of CO, used by Gonyostomum in photosynthesis, but judging from the abundance with which this organism occurs at most levels in the pond, it would seem that there is a plentiful supply of CO, for their needs from the surface to the bottom. It will be remembered that practically all of this is in the form of free CO, and bicarbonate CO,, both of which can be used by chlorophyll-bearing organisms. The congregation of Gonyostomum in- dividuals in rather large numbers at lower levels during the night, when the free CO, content of the water may reach 75 p.p.m. or more, indi- cates that such amounts of free CO, are not particularly detrimental to these green flagellates. In fact the decreasing amounts of free CO, during the daylight morning hours, from 5 A.M. to 9 A.M.., at all levels and the increasing amounts of free CO,, with one exception, during the dark morning hours, from 1 A.M. to 5 A.M., indicate that photosynthesis takes place at all depths during the daytime. These conditions may be seen in Table I where the total CO,, free CO,, and bicarbonate CO, are given in parts per million for August 20 and 21, at various depths and times of day. It must be pointed out that our results are not offered as conclusive proof that photosynthesis takes place at all depths during the daytime. In fact, we know that at times bubbles of gas rise to the surface of the pond which should tend to cause irregularities. However, we think that our results indicate that photosynthesis may occur at all levels. Since the knowledge of how free CO, and bicarbonate CO, are used by aquatic, chlorophyll-bearing organisms is quite uncertain and since our studies of the relation between the CO, content of the water and photosynthesis have not been experimental in nature, we shall have noth- ing more to say concerning the relation between the CO, content of the water and the vertical distribution of Gonyostomum except in so far as to point out that the amount of free CO, is one of the factors governing photosynthetic activity. We can say but little concerning the effect of the oxygen content of the water on the movements of Gonyostomum. However, if it can be demonstrated that this organism behaves like the chlorophyll-bearing Paramecium bursaria which, according to Engelman (1882) moves to- ENVIRONMENTAL CONDITIONS IN A BOG POND 297 ward the light, when there is little oxygen in the water in which it is swimming, so that under the increased intensity the organism will produce more oxygen for respiration as a result of increased photo- synthesis, then the small amounts of dissolved oxygen in the lower layers may possibly be considered as a factor in determining the upward migra- tion of Gonyostomum individuals during the early morning hours. Our observations furnish no evidence to support the idea that the organism under consideration migrates vertically due to a search for organic food. SUMMARY 1. The vertical distribution of the green, euglenoid-like organism, Gonyostomum semen, in a bog pond, has been studied at intervals of four hours, throughout twenty-four hours, with reference to air tem- perature, water temperature, illumination at the surface of the pond, total-, free- and bicarbonate CO, in the water, O, in the water and the pH value of the water. 2. The practically undisturbed condition of the water in the pond made it possible to demonstrate an oxygen and a carbon dioxide “ pulse.” These are mainly the result of the photosynthetic activities of Gonyosto- mum. 3. The acidity of this bog pond is partly due to the acid water which — enters from the sphagnum bog. 4. A striking chemical characteristic of the pond is the sudden in- crease in the free CO, content of the water. between 0.5 and 1.0 meter. 5. Many individuals of the Gonyostomum population of this pond move upward during the early morning hours and congregate in maxti- mum numbers somewhere between one meter and the surface. The level at which this occurs seems to be dependent on the proper combina- tion of light, water temperature, and CO, content of the water. During the late afternoon and night many individuals move downward so that the maximum counts are found in the lower layers. The upward move- ment of individuals of Gonyostomum from the lower levels is probably due to an increase in the light at the source combined with an increasing gradient of light, temperature, and CO, from the bottom to the surface, acting on Gonyostomum individuals which have been living in darkness. The downward movement is probably due to some physiological con- dition resulting from a sufficient amount of photosynthetic activity or a reproductive tendency or both which is followed by a quiescent sinking. 298 R. P. COWLES AND C. E. BRAMBEL REFERENCES BereMan, H. F., 1921. Am. Jour. Bot., 8: 50. ' BrampeL, C. E., anp R. P. Cowtes, 1935. Anat. Rec., 64, No. 1, Supplement No. 1, Dec. 25: 116. Cuapman, R. N., 1931. Animal Ecology. Ciark, W. M., 1928. The Determination of Hydrogen Ions. Williams and Wil- kins. Baltimore. k CxiarKE, G. L., 1934. Ecol. Monog., 4: 530. Cow tes, R. P., Aanp C. E. BrAmset, 1934. Anat. Rec., 60, No. 4, and Supplement : 97. Cow es, R. P., anp C. E. Bramset, 1935. Anat. Rec., 64, No. 1, Supplement No. 1, Dec. 25: 116. Cow tes, R. P., anp A. M. ScHwitTatxa, 1923. Ecology, 4: 402. Dancearp, P.-A., 1889. Le Botaniste, Premiére Série: 21. Drovet, F., anp A. CoHEN, 1935. Biol. Bull., 68: 422. ENGELMANN, T. W., 1882. Pfliiger’s Arch., 29: 391. Ester.y, C. O., 1907. Am. Jour. Physiol., 18: 47. Esterty, C. O., 1917. Univ. Calif. Publ. Zool., 16: 381. Esterty, C. O., 1919. Jbid., 19: 1. Jennincs, H. S., 1906. Behavior of the Lower Organisms, 250-259. Columbia University Press. New York. JEwELL, M. E., ann H. W. Brown, 1929. Ecology, 10: 473. Jupay, C., anp H. A. ScHomer, 1935. Biol. Bull., 69: 76. Kixucui, K., 1930. Quart. Rev. Biol., 5: 189. Kuess, G., 1883. Untersuch. aus dem Bot. Inst. su Tubigen, 1: 311. Mast, S. O., 1911. Light and the Behavior of Organisms, 272-279. Wiley. New York. Mast, S. O., 1920. Anat. Rec., 17: 345. MurtrKkowskI, R. A., 1918. Trans. Wis. Acad. Sci., 19, Part 1: 384. Parker, G. H., 1902. Bull. U. S. Fish Comm., 1901, 21: 103. Puuiip, C. B., 1927. Ecology, 8: 73. Rose, M., 1925. Arch. d. Zool. Expér. et Gén., 64: 387. RussEL., F. S., 1927. Biol. Rev. and Biol. Proc. Camb. Phil. Soc., 2: 213. SOUTHERN, R., anpD A. C. GarDINER, 1932. Proc. Roy. Irish Acad., 40: 121. UTERMOHL, H., 1924. Schrft. f. Siisswasser-u. Meereskunde, Hit. 9, 11. Jahrg.: 260. We tcu, P. S., 1935. Limnology, 114-115, 94-97, 219-235. McGraw-Hill Book Co. New York. WHIPPLE, G. C., 1927. Microscopy of Drinking Water, 237-248. Worruincton, E. B., 1931. Int. Rev. der gesam. Hydrobiol. u. Hydrograph., Bd. 25: Hit. 5/6, 394-436. PROMMIN Ele s SB ENDEN Een 2ROROPIEA SM L. V. HEILBRUNN UNIVERSITY OF PENNSYLVANIA Bleven years ago, Leathes and Raper (1925) wrote: “ Clearer. ideas as to the physical relation between insoluble fat and the aqueous proto- plasmic systems of the living organism, and of the part played by choles- terol in this relationship, are indispensably necessary before we can answer some of the commonest questions in physiological inquiry.” The sentence is as true today as when it was written. Although it is now generally realized that protoplasm is more than a solution of proteins, and that non-aqueous lipids are essential to its life, biologists and physi- ologists have scarcely given thought to the problem of how the lipids are related to the rest of the protoplasm. Most students of the cell are scarcely aware of the existence of the problem. And yet the patholo- gists have known for many years that the fats and fat-like substances of protoplasm are so bound or united to proteins as to be for the most part non-recognizable in the living or stained cell. It is only when degenera- tion occurs that the lipids are freed from their union with protein. In such degenerated cells, fats previously concealed may occupy a large part of the cell volume. Obviously, such a freeing of lipids results in an impairment or a loss of protoplasmic activity, and this bears witness to the importance of the lipids for the vital machinery. The manner in which lipids are bound to proteins in living cells is thus a question which merits investigation, and as will be shown later, it is a question which can be studied experimentally. Unfortunately, however, one difficulty in the way of progress is the lack of definite information as to the manner in which various lipids may unite with proteins in vitro. Thus Handovsky (1933) states: “ Die Verbindungen die einzelne Eiweisskorper mit anderen Eiweisskorper oder mit anderen lyophilen Kolloiden (z. B. Lipoiden) eimgehen, sind biologisch sehr wichtig, aber wenig exakt untersucht.” The experiments to be reported in this paper were begun as the re- sult of a chance observation. Sea urchin (Arbacia) eggs which had been exposed for 4 hours to a solution of oxalated sea water made by adding 1 part of m/4 ammonium oxalate to 2 parts of sea water were centrifuged in an Emerson electric centrifuge (force about 7,000 times gravity). It was then found that the gray cap of these treated eggs was 299 300 LV. BEI ERUNN very much larger than that of normal eggs. As is well known, the gray cap of centrifuged Arbacia eggs consists of lipid substances. There was thus an apparent increase of the lipid portion of the protoplasm. In the original experiment fertilized eggs were placed in the solution, but later observations showed similar, though not as pronounced effects with unfertilized eggs. With the unfertilized eggs, somewhat longer ex- posures are necessary. It was first thought that the action of the oxalated sea water was due to the oxalate ion. However, experiments with potassium oxalate gave no support to this view, and it was realized that the effect produced by the ammonium oxalate was due to the ammonium ion. Indeed, various types of ammonium salts all give essentially the same results. In addi- tion to the oxalate, the acetate, chloride, and sulfate were successfully used. Fis. 1 Fic. 2 Fic. 3 The details of the experiment are very simple. Eggs are placed in mixtures of sea water and isotonic solutions of an ammonium salt, and are centrifuged after the lapse of a suitable time interval. Solutions of ammonium salts alone can not be used, for they cause a coagulation of the protoplasm, and the eggs then resist. stratification on centrifugal treatment. In most of the experiments, one part of an isotonic solution of ammonium chloride was added to 3 or 4 parts of sea water, and eggs were then immersed in this mixture. If unfertilized eggs are used, in 4 or 5 hours, centrifugal treatment shows an increase in the volume of the gray cap. A very marked increase appears after 8 or 10 hours. A typical result is shown in Figs. 1-3. These show eggs which were centrifuged for 3 minutes in an Emerson electric centrifuge at about 7,000 times gravity. The control eggs shown in Fig. 1 were allowed to lie in sea water for 8 hours before being centrifuged ; they show a typical PROTEIN LIPID BINDING IN PROTOPLASM 301 small gray cap. Figs. 2 and 3 show the appearance of eggs centrifuged after exposure to 70 parts sea water plus 30 parts 0.53 m NH,C1 for 8 hours. The centrifugal treatment of the experimental eggs was the same as that of the control, but the eggs exposed to the ammonium salt show a greatly enlarged gray cap. Indeed the gray cap is so large that the hyaline zone is reduced to a narrow streak. It might be thought that in these eggs with enlarged gray cap, the increase in volume is due to a greater scattering of the materials of the cap. However, when the eggs are centrifuged for short or long periods, the volume of the gray cap remains essentially the same. Such centrifu- gal treatment can not be continued indefinitely, for the eggs eventually break into two; but with the most vigorous centrifugal treatment pos- sible, the gray cap of the eggs exposed to ammonium salts is always much larger than that of normal eggs. The behavior of Arbacia eggs toward solutions of ammonium salts is not peculiar to this type of material. When eggs of the clam Cu- mingia are exposed to mixtures of sea water and isotonic ammonium chloride or oxalate, there is in this case also an increase in the width of the fatty zone when the eggs are centrifuged. This increase is not as striking as in the case of Arbacia. In Cumingia eggs, the normal gray cap or zone of fatty materials appears rather dark under the microscope. What appears to be the added accumulation after treatment with ammo- nium salts is somewhat lighter or more transparent. A similar distinc- tion can sometimes be observed in Arbacia eggs. Ammonium salts may also cause an increase in the visible fat of Ameba proteus. In experiments performed in May, 1935, amcebe were kept in dilute solutions of ammonium chloride for 3 or 4 hours and were then centrifuged vigorously (5 minutes at 7,000 times gravity). The solutions used were m/60 or m/120 NH,Cl, at a pH of approximately 6.0. In one case, 1 part of m/60 NH,Cl was added to 2 parts of m/40 NaCl (pH 6.5). Numerous experiments were tried. In every case, care was taken to select cultures which normally had no free fat or lipid in them, as occasionally cultures are found which contain free fat even without treatment. Normally, when an Ameaba proteus is centrifuged vigorously, the crystals of the protoplasm accumulate in the heavy or centrifugal half of the cell and the rest of the cell is clear and trans- parent (see Heilbrunn and Daugherty, 1932). However, after treat- ment with ammonium salts, the centrifuged amcebz had approximately a fourth or more of the cell filled with particles lighter than the rest of the fluid protoplasm. These light particles stained brilliantly with Sudan III and were thus clearly lipoid. Amcebe constitute a much more variable material than Arbacia eggs. 302 L. V. HEILBRUNN It has already been noted that occasional cultures show free fat even without treatment with ammonium salts. On the other hand, in experi- ments done in March, 1936, no free fat could be demonstrated even after treatment with ammonium salts. However, in June, 1936, experiments with ammonium chloride were successful in 80 per cent of the amcebe tested. Perhaps there is a seasonal variation. Whether this is true or not, one can distinguish three types of amcebz. In the first type, free fat is present normally; in the second, free fat appears only after treat- ment with ammonium salts; and finally there are amcebz in which there is no free fat even after exposure to ammonium salts. Similar variations appear to occur in frog muscle. Students in my laboratory have studied the effect of ammonium salts on the detectable fat in frog muscle fibers (Rana pipiens). Mr. H. Blumenthal exposed gastrocnemius muscle to isotonic solutions of ammonium salts mixed with Ringer’s solution, and he found on fixing in osmic acid solutions that there was a great increase in the free fat. Similar results were also obtained with heart muscle of the frog. Mr. Newman has repeated Blumenthal’s experiments using single muscle fiber preparations from the frog sartorius, and staining with Sudan III. In the winter, the nor- mal muscle fibers are without free fat, but show a large amount after exposure to ammonium salts. In the spring, the normal muscle fibers have free fat even before treatment. In the late summer, it is possible that free fat may be absent even after ammonium treatment. This would accord with the observation of Bell (1911) that during the summer months it is not ordinarily possible to demonstrate fat-staining granules in Rana pipiens muscle, although such granules are easy to observe during the spring and fall. It is a remarkable fact that in various types of protoplasm, in sea urchin eggs, in clam eggs, in amceba, and in frog muscle, treatment with ammonium salts results in an increase in the free fat or lipid. In the sea urchin egg, this increase in fatty materials is indicated by a very marked increase in the volume of granules lighter than the rest of the protoplasm. That these lighter granules are fatty seems reasonably certain. They could scarcely be protein, for proteins constitute the heaviest rather than the lightest constituents of the protoplasm. An attempt was made to stain the lighter granules of the sea urchin egg with Sudan III. This did not prove very satisfactory, perhaps because of the small size of the granules. In ameceba, on the other hand, the light granules set free following treatment with ammonium salts stained very readily with Sudan III. So, too, in the muscle experiments, fat could be demonstrated both by osmic acid and by Sudan III. It might be argued that in the experiments with Arbacia eggs, the PROTEIN LIPID BINDING IN PROTOPLASM 303 increase in volume of the oil cap was due to an increased affinity of the oil for water. However, if Arbacia eggs are broken into two fragments by very vigorous centrifuging, and the heavier fragments which are free from fat or oil are then treated with ammonium salts, there is an accumu- lation of lighter particles at the centripetal pole of the heavy fragments when these are centrifuged. This indicates that the lipid material which appears following ammonium salt treatment can originate from proto- plasm freed from its original fatty substances. Moreover, it may be noted again that in amceba, cells totally without free fat may be made to show a considerable quantity after treatment with ammonium salts. How can one account for the increase in lipid materials? Although change from protein to fat is possible, it can readily be shown that no such transformation is responsible for the results outlined above. By gathering large masses of eggs, it was possible to make quantitative determinations of the total lipid content of normal eggs as compared with the lipid content of eggs treated with ammonium salts. In order to insure an effect of the ammonium salts, the eggs must be placed in large flat dishes so as to avoid crowding during the experiment. A con- centrated egg suspension obtained from a hundred or more sea urchins was divided into two equal portions. One portion was allowed to remain in sea water. To the other portion, 0.53 m NH,Cl was added in the proportion of 3 parts of the solution to 7 parts of sea water. After 8 hours, the eggs from each portion were collected and analyzed for total lipid. In one experiment, extraction was performed with a modified Soxhlet apparatus, in a second the wet material was extracted with alco- hol and ether. Both tests showed almost identical quantities of total lipid in eggs treated with ammonium salt and controls. These tests were performed with the assistance of Mr. Samuel Koppelman. As a result of the chemical analyses, it may be concluded that the action of ammonium salts is to free bound lipid from union with pro- tein. That living cells normally contain a relatively large amount of fat or lipid in the bound state is abundantly clear from the literature on fatty degeneration. Cells may appear to be practically free from fatty substances, whereas upon degeneration they become filled with numerous fatty droplets. And yet, often in these cells which have undergone fatty degeneration, no increase in total lipid can be demonstrated by chemical methods. Thus, obviously, bound fats or lipids occur in protoplasm. In blood, also, the lipids are to a large extent bound to the proteins. The mechanism of action of the ammonium salts is to some extent clear. Jacobs (1922) has shown that when cells are exposed to a me- dium containing ammonium salts, the pH of the protoplasm rises. This is due to a hydrolysis of ammonium salt to ammonium hydroxide and 304 L. V. HEILBRUNN acid. The ammonia or ammonium hydroxide is then able to penetrate the cell, whereas the inorganic acid remains outside. ; That the effect of the ammonium salts is in reality due to an alka- linization of the protoplasm is supported by two sets of observations. In the first place, ammonium hydroxide also causes an increase in free or visible lipid. Thus sea urchin eggs were exposed for several hours to solutions with enough ammonium hydroxide added to bring the pH to 8.9 in one case, and to 9.2 in another. To avoid excessive precipitation of magnesium hydroxide, the ammonium hydroxide was not added to sea water, but to a sodium chloride-calcium chloride mixture (125 cc. 0.53 m NaCl + 1.5 cc. 0.3 m CaCl,). When the eggs were centrifuged after exposure to these solutions, there was a marked increase in the volume of the oil cap, and the appearance of the eggs was like that shown in Figs. 2 and 3. In the second place, it was discovered that an excessive amount of carbon dioxide in the medium surrounding the cells prevented an increase in free fat or lipid. Thus when egg cells were crowded, even though they were in contact with the proper strength of ammonium chloride, no increase in visible fat occurred. Also when carbon dioxide was bubbled through the medium, the action of the ammonium chloride was inhibited. Carbon dioxide readily penetrates cells and it doubtless tended to prevent the alkalinization of the protoplasm which might otherwise have been caused by the ammonia diffusing into the cells from the ammonium chloride solution. It is quite conceivable that alkalinization in itself might cause a liberation of lipid from protein-lipid combination. Thus Theorell (1930), who has done some of the most interesting work in the little- studied field of protein-lipid binding, states: “In der Literatur findet man mehrere experimentell mehr oder weniger wohlbegrtindete Angaben, das sowohl Cholesterin als auch Phos- phatide an die Globuline, besonders an das Euglobulin, gebunden seien. Gabe es eine solche Verbindung, ware es ja nicht unwahrscheinlich, dass sie bei dem isoelektrischen Punkte der Globuline dissoziiert ware, wo ja ihr Minimum an elektrischer Ladung und Hydratation zu finden ist.” Theorell himself studied the binding of globulin and cholesterol and found that although the cholesterol was bound at either side of the iso- electric point, at the isoelectric point itself there was no binding. How- ever, in the case of albumin the situation is not quite so simple, and a recent paper of von Pryzlecki, Hofer, and Frajberger-Grynberg (1935) indicates the complexities of protein-lipid binding. In protoplasm, it is quite possible that increase in pH does bring a shift toward the isoelectric point, of at least certain of the protein con- PROTEIN LIPID BINDING IN PROTOPLASM 305 stituents. In this connection, it is interesting to remember that there is evidence that protoplasmic particles migrate to the cathode when an electric current passes through a cell (see Heilbrunn, 1928). This would indicate that there are proteins present which are on the acid side of the isoelectric point. SUMMARY 1. When various types of protoplasm are treated with dilute solutions of ammonium salts, there is an increase in the free fat or lipid. 2. This effect is due to an alkalinization of the protoplasm. ILA BIR AMICON, OMAN) Bett, E. T., 1911. The interstitial granules of striated muscle and their relation to nutrition. Intern. Monatschr. f. Anat. u. Physiol., 28: 297. Hanpovsky, H., 1933. Allgemeine Chemie der Proteine. In Oppenheimer’s Handbuch d. Biochemie d. Menschen u. Tiere. Second edition. Erganz- ungswerk, vol. 1, p. 320. HeEILBRUNN, L. V., 1928. Colloid Chemistry of Protoplasm. Berlin. G. Born- traeger. 1928. Heitprunn, L. V., anv K. DaucHerty, 1932. The action of sodium, potassium, calcium, and magnesium ions on the plasmagel of Ameeba proteus. Physiol. Zool., 5: 254. Jacoss, M. H., 1922. The influence of ammonium salts on cell reaction. Jowr. Gen. Physiol., 5: 181. LeaTHEs, J. B., AND H. S. Raper, 1925. The Fats. Longmans, Green and Co. Second edition. London, 1925. PrzyLecki, S. J. von, E. Horer, AND S. FRAJBERGER-GRYNBERG, 1935. Uber Af- finitaten zwischen Proteinen und Fettsauren, Fetten, oder Lipoiden. Buo- chem. Zeitschr., 282: 362. THEORELL, H., 1930. Studien tiber die Plasmalipoide des Blutes. Biochem. Zeitschr., 223: 1. FREE CALCIUM IN THE ACTION OF STIMULATING AGENTS ON ELODEA CELLS DANIEL MAZIA AND JEAN M. CLARK (From the Zoélogical Laboratory, University of Pennsylvama) INTRODUCTION A number of investigations have been made in this laboratory on the colloid-chemical changes in the protoplasm that are involved in cell stimu- lation. The results conform to the hypothesis, discussed most fully by Heilbrunn and Daugherty (1933), that a primary effect of stimulation is the release of Ca ion in the cell. MHeilbrunn, Mazia, and Steinbach (1934) (fuller accounts to be published soon) reported that direct analy- sis after ultrafiltration showed that in the fertilization of sea-urchin eggs and in the injury of muscle, the free Ca content of the cells indeed in- creases. On the other hand, changes which are interpreted as evidence of Ca release are prevented by oxalates (Heilbrunn and Young, 1930; Heilbrunn and Daugherty, 1933). At the present stage of the problem, a qualitative, intracellular method would be even more valuable than our quantitative ultrafiltration method for a general testing of the hypothesis of Ca release. The chem- ical accuracy of the latter method is offset by the fact that masses of cells must be used, and must be subjected to violent means of killing. An intracellular method in which it is not necessary to add reagents is ideal. Just such a method is provided in the present case by plant cells which contain high concentrations of soluble oxalates in their sap. An increase in the free calcium concentration of the cell is immediately indicated by the formation of calcium oxalate crystals. Although there is a large literature on the natural occurrence of calcium oxalate, dating back to Malpighi (1675) (cf. Patschovsky, 1920), there are very few papers on its experimental production in cells in which it does not normally occur. Osterhout (1910) used crystal formation as a test of Ca penetration into root hairs of Dianthus. The use of the cells of Elodea, in which crystals do not occur. nor- mally, was suggested to us by the work of Nadson and Rochline-Gleich- gewicht (1927) and Rochline-Gleichgewicht (1930). These authors noted the appearance of CaC,O, crystals in cells of E. densa, E. cana- densis, and Pterygophyllum hepaticefolium after ultraviolet irradiation, after exposure to radium emanation, and after plasmolysis. 306 * FREE CA AND STIMULATION OF ELODEA CELLS 307 IDENTIFICATION OF CRYSTALS Calcium oxalate was produced, in our experiments, as tetragonal crystals of the trihydrate and the monoclinic crystals of the monohydrate. The former are the most characteristic, as well as the largest and most easily recognized. Large crystals of both types are shown in Fig. 1. In addition to such crystals, sometimes small, needle-like crystals of calcium oxalate are formed. Their habit cannot be determined. Although such appearances correspond to descriptions of calcium oxalate in the literature, and to our own in vitro products, it seemed worth while to verify this. The solubilities of experimentally produced crystals in 5N H,SO, and in 10M acetic acid were compared. The crys- tals dissolved in the sulphuric, but not in the acetic acid. This behavior is characteristic of calcium oxalate. But the most elegant method for the identification of microscopic crystals is that of refractive index determination. Such determinations were made for us by Mr. A. W. Postel, of the Geology Department of this University. They demanded the greatest skill and patience, as the largest crystals obtainable in Elodea cells are much smaller than those usually found in petrological material. The indices of refraction were obtained by the immersion method. The crystals were bathed in liquids of known refractive index, until those liquids were found which had the same indices as the crystals; in such liquids all relief between crystal and liquid vanishes. In the present work the Becke line method was used (for description see Chamot and Mason, 1931, Vol. 1). Since the crystals in our cells were anisotropic, the measurements had to be made with the polarizing micro- scope, as the values of the several indices of the same crystal can be ob- tained only by placing the vibration planes of the crystal successively in a position parallel to the vibration plane of the lower Nicol prism. The largest crystals were obtained in leaves plasmolyzed for about thirty minutes in 0.5M sucrose, and deplasmolyzed in tap water. These leaves were dehydrated in ethyl followed by isoamyl alcohol. The matching liquids were a series of mixtures of isoamyl alcohol and alpha-monochlor-naphthalene (halowax oil), ranging in refractive index from 1.440 to 1.600 in gradations of about 0.005. The liquids were introduced into the dehydrated cell by the “sinking method ” used for clearing microscopic preparations (Lee, 1900, p. 79). After the leaf had sunk through a column of a mixture, it was immersed for some time in a fresh quantity of the same liquid. Five-cc. samples were used. The leaf was now ready for examination with the polarizing micro- scope. When matches were obtained, the exact refractive indices of the 308 DANIEL MAZIA AND JEAN M. CLARK liquids were determined by means of a goniometer with a hollow prism for the liquid. Exact values were obtained for the tetragonal crystals. The indices that could be measured came out as follows: ; (3) € 1.491 1.540 1.505 1.499 1.495 These figures are in agreement with those given by Wherry and Keenan (1923). The shape of the other crystals, as well as the fact that they had one index of refraction higher and one lower than the highest index of the tetragonal crystals, was taken as sufficient evidence that they were monoclinic crystals of calcium oxalate monohydrate. | The optical properties of our crystals, therefore, identify them as calcium oxalate. Parallel determinations on artificially prepared cal- cium oxalate monohydrate and trihydrate verified this conclusion. ACTION OF STIMULATING AGENTS Agents Used Stimulating agents may be defined as physical or chemical agents which call forth the characteristic responses of irritable cells. In working on a general theory of stimulation, it is necessary to use those agents which affect the most familiar types of responsive cells, such as muscle, nerve, and unfertilized egg cells. Among such general stimu- lants are electric shocks, radiations, changes in temperature, mechanical shock, a variety of chemical agents, and osmotic changes. To test Heilbrunn’s hypothesis, which was the purpose of this work, cells of Elodea canadensis were subjected to all these influences to see whether the latter caused a release of calcium in the cell interior. EXPLANATION OF PLATE I Fic. 1. Crystals produced by plasmolysis in 0.5M sucrose and deplasmolysis in pond water. Crystals at lower left tetragonal, others monoclinic. Fic. 2. Crystals produced by two condenser shocks. Twenty volts, 2 micro- farads. Fic. 3. Deposit of crystals toward anode (left in all figures). Current 0.75 milliampere for 15 seconds. Fic. 4. Deposit of crystals toward anode, under higher magnification. Same current as in Fig. 2. Fic. 5. Electrical breakdown of plasmolyzed cells. Normal cell marked by arrow. Fic. 6. Same field at lower focus, showing crystals which have settled down within the tonoplasts. FREE CA AND STIMULATION OF ELODEA CELLS 309 Priate | 310 DANIEL MAZIA AND JEAN M. CLARK Effect of Electric Current Condenser Discharges.—In studying electrical stimulation it seemed desirable to use shocks of brief duration in order to minimize such fac- tors as electroendosmosis. A condenser of 2-microfarad capacity was charged from a “B” battery that gave about 20 volts, and discharged through Cu-CuSO, electrodes of simple design. Glass tubes of 6 mm. bore were filled at their tips with a 3-4 per cent agar gel made up in tap water or 0.1M NaCl. Over the gel was placed saturated CuSO, solu- tion, and the copper wires were dipped into this. The Elodea leaf was dried on filter paper and placed on a glass slide. The electrode tips were placed a few millimeters apart, because the resistivity of the tissue was found to be so great that effective currents could not be passed through the whole length of the leaf with the appa- ratus used. When more than one shock was used, the shocks were spaced at intervals of one second or two seconds. At first, 5-10 shocks were used. These caused the deposit of many crystals in the vacuoles although the protoplasm streamed vigorously, more vigorously if anything, after the shocks. It was soon found that definite production of crystals results from application of a single shock, or two shocks. Cells that had received two shocks are shown in Fig. 2. In general, more shocks produced more crystals. One had to use more than 20 shocks to kill some of the cells. The crystals were localized at the anodal end of the cell, as Fig. 2 shows. Later they became scattered as a result of streaming motion, but if observed immediately after stimulation, it could clearly be seen that calcium oxalate forms at the anodal end on stimulation. Continuous Direct Current—When a 110-volt source of D.C. was used, the current could be passed through the whole leaf, and the cells observed during the passage of the current. The same electrodes were used as in the condenser experiments. When a current of 1 milliampere is made and instantly broken, the immediate effects are a slight but definite jerking of the protoplasm away from the cathodal end of the cell, and the appearance, very soon, of small crystals of calcium oxalate in the vacuole at the anodal end. If the current is allowed to flow for a second or two, and then broken, the same effects become more pronounced. At the cathodal end the proto- plasmic layer thickens markedly and the convex contour of the tonoplast becomes visible. At the anode the protoplasmic layer becomes thin, and the protoplasmic strands in the vacuole are broken, the protoplasm being drawn into the main mass. More crystals appear in the vacuole at the anodal end. If at this stage the cells are treated with 0.33M sucrose, FREE CA AND STIMULATION OF ELODEA CELLS Sui they plasmolyze, but the protoplasts adhere to the cell wall at the anodal end. If a larger amount of current is used, a layer of small crystals forms along the extreme anodal boundary of the protoplast, appearing to be contiguous to the cell wall (Figs. 3,4). A latent period may intervene between the breaking of the current and the appearance of this crystal deposit. The formation of this anodal deposit of crystals marks the beginning of the disintegration of the cell membrane. This first ap- pears toward the anode. With further passage of the current, the dis- integration passes in a wave over the surface of the protoplast, from anode to cathode. Occasionally, it has seemed that it went in the re- verse direction. As the wave passes, calcium is released. Apparently free oxalate is available at the site of the disintegration, for the wave leaves behind it a wake of calcium oxalate crystals on the surface of the protoplast. The wave passes rather slowly, its speed depending on the current applied. If the current is broken, it stops, and is seen as a tenuous, refractive boundary, the edge, apparently, of a perfectly trans- parent layer. The wave reported by Weis (1925) in Allium cepa cells may be similar, though from his description we would judge that he was ob- serving a cathodal movement of the tonoplast. Another phenomenon observed in cells subjected to severe currents was the frequent pulling-in of the protoplast along the sides of the cells, particularly at points where it was convex. Once a deposit of crystals has formed along the anodal end of the protoplast, it is no longer possible to plasmolyze the cells; they do not recover their semipermeability on standing. The chloroplasts shrink and assume puckered shapes. To determine the relationship between membrane disintegration and calcium release, cells were plasmolyzed with 0.5M sucrose, and the proto- plasts were subjected to the electric current in this medium. Since the conductance is low, most of the current is shunted through the cells, so that small currents will have severe effects. Currents of 0.5—-1 milliampere were used. When the circuit is made and instantly broken, all the protoplasts jerk toward the anode (cf. Rubinstein and Uspenskaya, 1934). Slightly more current causes some cells to break down; others are more resistant, and some never break down. The cells which do break down rupture first at the anode. The plas- malemma breaks, at one point as a rule, but sometimes at several. Pro- toplasmic granules flow out, and a few calcium oxalate crystals are seen to form outside the protoplast. The tonoplast protrudes through the oz DANIEL MAZIA AND JEAN M. CLARK point of rupture, enlarging the break in the protoplasmic layer. The clear contents of the vacuole become filled with crystals, beginning at the point of exposure, and a wave of plasmalemma disintegration passes over the surface from anode to cathode, leaving the tonoplast to round up again (Figs. 5,6). The protoplasmic layer, also broken at the anode, “peels” off from the tonoplast, and is clumped at the cathodal end of the cell. Sometimes the tonoplast is freed in two parts, or fragments break away, becoming small, clear spheres containing crystals. When the current is allowed to flow for 15—30 seconds, the tono- plasts, too, break down, contracting first, then disintegrating after anodal rupture. Our experiments did not reveal any natural polarity in the cells of the Elodea leaf. The results were the same whether the current was passed from base to tip or vice-versa. If the leaf was placed transversely in the electrical field, crystals were in this case also deposited at the anodal end. The explanation of the anodal calcium release cannot be discussed at length here. An obvious suggestion is that a local increase in acidity decreases the capacity of the proteins to bind Ca. However, the ob- servations of Kuhne (1864), Bethe (1915), Mast (1931), led to the conclusion that plant cells subjected to direct currents become more alkaline at the anode, and acid at the cathode. However, Blinks (1932) attributes the apparent polar change in color in the natural anthocyanin indicators used by the above authors to electrophoresis rather than to change in reaction. We performed a number of experiments to see whether acid could cause the setting free of Ca and the precipitation of calcium oxalate. Leaves were placed in 0.1N or 0.2N acetate buffers of pH 3.8, 4.2, 4.6, 5.0, 5.8. These external pH values tell nothing about the effect on the intracellular pH, for the acetic acid penetrates rapidly while the sodium acetate penetrates slowly if at all. The leaves in these solutions were examined at intervals up to two hours. Various morphological changes were observed, but in no case did crystals appear. Therefore, the hy- pothesis that local acid accumulation is responsible for calcium release when direct currents are applied was not verified. Effect of Oxalate-—The vacuole of the Elodea cell contains a fairly high concentration of free oxalate. There is also reason for believing that there is oxalate in the protoplasm (see discussion). Yet, if the observations of Heilbrunn and Daugherty (1933) are of general sig- nificance, we must expect that immersing the Elodea cell as a whole in an oxalate solution should prevent calcium release. In testing this, leaves were washed in 0.05M sodium oxalate before they were exposed FREE CA AND STIMULATION OF ELODEA CELLS hs} to electric currents. Since this solution has such a high conductivity as compared to the normal medium of the cells, they were removed from it after a minimum of ten minutes exposure and placed in 0.001M so- dium oxalate or in distilled water. The same strengths of electric cur- rent were applied to these cells as had been applied to cells not washed in oxalate. The results were consistent in many experiments. No crystals were produced no matter how long the current was passed. The plasma- lemma did disintegrate in a wave beginning at the anode, but even this -was not accompanied by crystal production. Mechanical Stimulation Mechanical shock is a very general stimulant. Its effects on cells are discussed in a review by Lepeschkin (1936) and by Angerer (1936). The latter, working on Amoeba, has found that the viscosity changes associated with mechanical agitation are similar to those described by Heilbrunn and Daugherty (1933) for ultraviolet irradiation, and there- fore in accord with the hypothesis that calcium release is a primary effect of the stimulation. Our method of applying the stimulus was very primitive. A leaf was placed between two microscope slides and rapped sharply with heavy forceps. In one experiment in which the leaf was rapped 50 times and examined immediately, more than half of the cells showed very clear crystal formation, most of the crystals being of the needle-like _ variety. The cells were not killed, but seemed to be streaming at a somewhat more rapid rate than normally. The effect of mechanical stimulation was tried on leaves from many different specimens of Elodea. Crystals always could be obtained, although sometimes very many shocks were necessary. Irradiation with Ultraviolet Light The valuable paper of Nadson and Rochline-Gleichgewicht on the production of calcium oxalate crystals as a result of ultraviolet irradia- tion has already been mentioned. There was no difficulty in verifying the observations of these authors. Leaves were irradiated from a Cooper-Hewitt quartz mercury vapor lamp, at a distance from the are of 25 centimeters. The leaves were exposed for 2.5 minutes, sur- rounded only by an adhering film of water. On examination within five minutes after treatment, they showed a definite but not profuse precipitation of calcium oxalate in the vacuoles of the cells. Most of 314 DANIEL MAZIA AND JEAN M. CLARK the crystals were tetragonal. Three minutes irradiation caused a much heavier precipitation. At the same time leaves were irradiated after having been immersed for 15 minutes in 0.05M sodium oxalate. In these leaves even 10 min- utes after irradiation there was no sign of calcium oxalate in the vacuoles of the cells. Therefore, just as in the case of electrical stimulation, treatment of the outside of the cell with oxalate prevents the release of calcium ion in the interior. Heat Heat stimulates many living systems; it is listed by Bethe (1915) among the general protoplasmic stimulants. In our preliminary experi- ments Elodea leaves were immersed in water at 100° C. This treatment killed the cells without producing any crystals. We then tried heat treatments of the same order of magnitude that have been found to stimulate other cells, particularly egg cells. Heating at 35° for more than an hour caused no crystal production. At 40° many crystals, a mixture of fine needles and large tetragonal crystals, were formed after 60 minutes exposure. At 45° exposures of 20 min- utes duration produced many crystals in all the cells. These cells plas- molyzed normally, after the heat treatment, in 0.3M sucrose. At 50°, 8 to 10 minutes exposure caused the production of crystals without destroying the ability of the cells to plasmolyse in the 0.3M sucrose solutions. At 60°, however, all the cells died after exposure of a min- ute or two—and no crystals were produced! To determine the effect of oxalate on the reaction of the cells to heat, leaves were washed in 0.05M sodium oxalate, and subjected to the heat treatments described above while they were still in the sodium oxalate solution. No crystals were produced. We attempted to decide whether the effects described above were due to the absolute heat treatments we used, or whether temperature change was the causal factor. We studied the effect of rapid cooling by im- mersion in an ice-water mixture. No crystals were produced by 10-, 20- and 60-minute immersions. When the cooled cells were placed in water at room temperature, no crystals were produced. Therefore, high temperatures themselves were responsible for crystal production by heating. Osmotic Changes Bethe, in the list of general stimulants cited above, includes osmotic influences ; hypertonic solutions particularly are stimulating agents. FREE CA AND STIMULATION OF ELODEA CELLS Silla If cells are plasmolyzed in 0.5M sucrose and deplasmolyzed in pond water, they are then found to have crystals in their vacuoles. When the process is followed under the microscope, it is observed that the crystals do not appear until deplasmolysis begins. The first effect observed, upon addition of the deplasmolyzing agent, is a strong contraction of the tonoplast. This is observed, even after complete, convex plasmolysis. It is strikingly rapid and vigorous, and even after harsh plasmolysis in 1.3M CaCl, deplasmolysis is preceded by further contraction. After strong plasmolysis, e.g., after 10 minutes or more in 0.5M sucrose, deplasmolysis leads to the death of most of the cells of Elodea leaves. The protoplasm expands in a series of slight jerks, which are due to the rupture, at various points, of the plasmalemma. The vacuoles of such cells acquire many crystals, even when a deplasmolyzing agent such as 0.25M sucrose or. distilled water, lacking any calcium, is used. In this type of fatal plasmolysis, the tonoplast breaks down as well as the plasmalemma. When weaker plasmolysis is used, such as follows exposure for less than 10 minutes to 0.5M sucrose, or 10-15 minutes in 0.33M sucrose, de- plasmolysis is. smooth and the cell may be replasmolyzed. Less crystals are obtained in these cells, but there is always a definite precipitate. Sometimes these crystals may adhere to the vacuole wall and be carried about as the protoplasm streams. When such cells are bathed in 0.05M CaCl,, no further precipitation of calcium oxalate is observed. Many experiments on the effects of very weak plasmolysis were performed. When cells are treated with a 0.2M NaCl solution, there is a preliminary stage, lasting about three minutes, in which the vacuole shrinks markedly, but the plasmalemma does not separate from the cell wall. Ifa hypotonic solution is added as soon as this preliminary stage of plasmolysis is established, the tonoplast expands and the cell resumes its normal appearance without the production of any crystals. If, how- ever, 2-3 minutes are allowed to elapse before the 0.2M NaCl is changed for pond water or any other hypotonic medium (the exchange is ef- fected under the microscope simply by drawing the new medium under the coverslip with strips of filter paper), the protoplast, which had not yet contracted in the hypertonic medium, separates from the cell wall, and within a few seconds expands again. In cells in which this transient plasmolysis occurs, crystals are formed. It seems, therefore, that actual shrinkage of the protoplast is necessary to cause calcium release after treatment with hypertonic solutions. It is immaterial whether this plas- molysis is produced while the cell is in the hypertonic medium, or. as a result of the “ contraction” that occurs on transfer from the hypertonic medium to the hypotonic. 316 DANIEL MAZIA AND JEAN M. CLARK Other experiments on the effect of mild plasmolysis were performed with the balanced mixture (0.18M NaCl plus 0.005M CaCl,) as plas- molyzing agent, and the same mixture diluted to half strength as deplasmolyzing agent. The time of exposure to the hypertonic solution was varied. The cells showed vigorous protoplasmic streaming after deplasmolysis, and the vacuoles contained crystals, usually small ones but sometimes numerous. We were unable to correlate time of plas- molysis with the amount of precipitate, although a quantitative method might reveal such a correlation. The precipitate could not be increased by washing the cells with a 0.05M CaCl, solution, except in those few cells that had been broken even by the mild treatment. These experiments answer the objection that the crystal formation may be ascribed to penetration of calcium from the external medium, after rupture of the plasmalémma by our treatment. First, in the ex- periments with mild plasmolysis there is no indication of membrane rupture. Second, although after strong plasmolysis more crystals can be obtained when 0.05M CaCl, is used as deplasmolyzing agent, very many crystals are obtained when distilled water, 0.1M NaCl, or 0.25M sucrose are used. Finally, when 0,05M sodium oxalate was used as deplasmolyzing agent, fewer crystals were produced, but there was a definite, sometimes an abundant, precipitate. The effect of oxalate on the calcium release in plasmolysis-deplas- molysis was investigated. Cells were immersed for various periods in a solution of 0.5M sucrose in 0.1M sodium oxalate. With: prolonged immersion—30 minutes or more—the cells gradually undergo “ systro- phy,” in which the main mass of cytoplasm and cytoplasmic inclusions flows together at one part of the cell, while the remainder appears as a thin layer around the now-visible tonoplast (cf. Kiister, 1929, p. 172). If the cells at any time after plasmolysis in this solution are treated with distilled water, they deplasmolyze. Typically no crystals are pro- duced. In this part of the work two types of experiments are to be distin- guished: (1) experiments in which plasmolysis is brief, or in which the weakly hypertonic solution 0.33M sucrose in 0.1M Na,C,O, is used; and (2) experiments in which strong plasmolysis, leading to systrophy, is obtained in 0.5M sucrose in 0.1M sodium oxalate. In the first group of experiments, deplasmolysis was smooth and the cells were not killed; they could be replasmolyzed. A number of deplasmolyzing agents were tried : distilled water, pond water, 0.1M NaCl plus 0.005M CaCl,, 0.05M sodium oxalate. No crystals were obtained with any of these. There- fore the results are in accord with those obtained with electrical, thermal, FREE CA AND STIMULATION OF ELODEA CELLS Sil 7) ultraviolet stimulation: washing in oxalate prevents the release of cal- cium which otherwise takes place. In the second group of experiments, in which deplasmolysis of the systrophied cells does involve rupture, no crystals were obtained when distilled water or 0.05M sodium oxalate was used, but crystals were obtained when pond water or 0.005M CaCl, was used. This result is enlightening, for it shows that the tonoplast is permeable to Ca; observa- tion showed that the crystals were formed as soon as the plasmalemma ruptured. To test whether the oxalate actually removes the Ca from the surface, or whether the presence of the oxalate ion itself is responsible for the prevention of crystal formation, some cells were first plasmolyzed in 0.5M sucrose in 0.1M sodium oxalate, then washed in 0.5M sucrose alone, then deplasmolyzed in 0.25M sucrose or in distilled water. No crystals were produced. Therefore, the oxalate does remove the Ca or whatever is responsible for the Ca release, from the surface. lf it is Ca that is removed from the surface by oxalate, one might be able to restore it, and thus to reverse the effect of the oxalate. Leaves treated with 0.05M sodium oxalate for 10 minutes—more than enough to prevent crystal formation, were washed in distilled water. Some of them were then placed for ten minutes in pond water or 0.02 CaCl,, others not. Both groups were then plasmolyzed in 0.05M su- crose and deplasmolyzed in distilled water. The cells that had been treated with pond water or CaCl, formed crystals on deplasmolysis— the effect of the oxalate had been reversed. The other leaves showed no crystals, or many fewer crystals. Therefore, oxalate does indeed act by removing Ca from the surface, and this Ca may easily be replaced. One interesting observation made in this part of the work was that violent deplasmolysis after plasmolysis in oxalate-containing solutions did not destroy the tonoplast. Especially after. systrophy, completely isolated tonoplasts could be obtained. Substitution of Citrate for Oxalate In the experiments on the prevention of calcium release, oxalate was used because it had been used by the previou§ authors whose ideas we were testing. Similar experiments were later performed in which citrate was substituted for oxalate. Although citrate does not precipitate Ca, it binds it in a non-ionized complex. Furthermore, the pK of this CaCit is so large, that at high citrate concentrations the citrate should bind Ca at the expense of proteins (cf. McLean and Hastings, 1935). Our tests were made with electrical stimulation, and with plasmolysis- 318 DANIEL MAZIA AND JEAN M. CLARK deplasmolysis treatment. Leaves were washed in 0.04M sodium citrate for 10 minutes, then rinsed in distilled water and subjected to direct cur- rents of about 1 milliampere. No crystals were formed, although the usual deposits were obtained in cells treated with distilled water alone. The only visible effect of immersion in sodium citrate was a tendency of the chloroplasts to flow together, which was reversed on their return to pond water. In studying the effect of citrate on plasmolysis-deplasmolysis, the following procedure was used. Leaves were washed in 0.035M sodium citrate for 10 minutes, then rinsed in distilled water and plasmolyzed in 0.05M sucrose. On deplasmolysis with distilled water no crystals were formed. In control leaves, treated similarly except for the exposure to citrate, the usual precipitate of calcium oxalate was obtained. It must be remarked here that this experiment is much clearer than the corre- sponding one with oxalate; in the latter there are occasional doubtful cases and the presence of crystals on the outside of the cells makes judg- ment difficult. The reversibility of the citrate effect was also studied. Leaves im- mersed in citrate for 10 minutes were restored to pond water or to 0.02M CaCl, for 10 minutes. They were then plasmolyzed in 0.05M sucrose and deplasmolyzed in distilled water. Crystals were obtained in these leaves ; in greater abundance in the leaves that had been treated with the pure Ca salt solution. Tests with the electric current also showed that the effect of the citrate could be reversed by treatment with pond water or calcium solution. The Effect of Isotonic Salt Solutions Certain isotonic salt solutions are considered to be stimulating agents. Sodium salt solutions in particular cause muscles to twitch, and favor the action of other stimulants on Arbacia eggs. One of us (cf. Heil- brunn, Mazia, and Steinbach, 1934) has shown that in the latter case, isotonic NaCl caused an increase in the free Ca concentration of the cell interior. Elodea leaves were washed and allowed to stand in 0.13M NaCl, 0.13M KCl, and 0.08M CaCl,. They were examined at various times over 24 hours. The cells in NaCl stopped streaming first; at 18 hours the protoplasm was barely moving. At 18 hours, in the CaCl, solution, the protoplasm was streaming without carrying the chloroplasts, while in KCI the chloroplasts were in motion. No crystals were produced in any of these cells. The experiment was repeated a number of times; in no case was calcium oxalate precipitation observed. FREE CA AND STIMULATION OF ELODEA CELLS 319 Other Toxic Agents Having found effective so many agents which, although stimulants, are also injurious, it is necessary to show that injurious agents which are not considered stimulants do not have the same effects as stimulants. A number. of such agents were tried. All of them caused the death of the cells without the production of crystals. They will simply be . listed: —s . Ninety-five per cent ethyl alcohol. 2 Cold: a. Zero degrees centigrade. b. Temperature of solid CQ,. ounkleats a. Seventy degrees centigrade. b. One hundred degrees centigrade. . NH,OH—0.01M. . CuSO,—0.01M. iS) on Tue Errect oF ANESTHETICS If the release of Ca ion is a general concomitant of the action of stimulants on cells, it is important to know whether anesthetics can prevent this release. The anesthetics used were ethyl ether, chloral hydrate, and ethyl urethane. Their effect on the response to electrical stimulation and plasmolysis-deplasmolysis was investigated. Leaves were placed in stoppered test tubes containing 1, 2, 3, 4, and 5 per cent solutions of ethyl ether in distilled water. They were re- moved after various times of immersion, and their reaction to currents of 1 milliampere was observed. In 1 per cent ether, even after 40 min- utes immersion, the electric current caused the formation of crystals to the same extent as in the controls. The same result was obtained in 2 per cent and 3:per cent solutions. In 4 per cent ether the cells reacted like the controls for 20 minutes, then they appeared coagulated and no crystals were produced. Subsequent immersion in pond water for 30 minutes did not restore their responsiveness to the direct current. More severe treatment, such as 30 minutes in 5 per cent ether, itself caused the production of small tetragonal crystals in most of the cells. The cells were dead; they could not be plasmolyzed and the amount of calcium oxalate could not be increased by applying electric currents. The sequence of events could be more easily studied in the chloral hydrate experiments. Ten minutes in a 2 per cent solution did not pre- vent crystal formation. Longer treatment with this solution (13-14 minutes) caused the formation of large crystals, but passage of the 320 DANIEL MAZIA AND JEAN M. CLARK electric current at this stage resulted in the formation of more crystals, concentrated at the anodal end. Still longer treatment with the anes- thetic caused crystal formation accompanied by the death of the cells; subsequent treatment with the electric current produced no more crystals. A considerable number of experiments with chloral hydrate established the fact that severe treatments—use of solutions stronger than 2 per cent—cause the formation of large crystals, associated with the rapid death of the cell. In the study of the effects of anesthetics on plasmolysis-deplasmoly- sis, the anesthetic was dissolved in the plasmolyzing agent. For ex- ample, cells were plasmolyzed in 0.2M NaCl containing 2 per cent ethyl urethane, and deplasmolyzed in pond water. Crystals were produced to | the same extent as in the controls. The cells were exposed for as long as 30 minutes to the anesthetic, without effect. The theoretical implications of these results will be considered in the general discussion. DISCUSSION Our results seem to us to uphold the hypothesis that when cells are stimulated, the free Ca concentration within the cell increases. We have shown the applicability of this hypothesis to the Elodea cell in its reaction to the most important stimulating agents. More complete verification of the hypothesis would depend on similar studies on widely different types of plant and animal cells. However, the interpretation of the experimental production of cal- cium oxalate in plant cells can be examined more closely. The bulk of the free oxalate is in the vacuole; some may be in the protoplasm. Only a trace of Ca could be in a dissolved state in the vacuole unless the pH is 4 or less, at which reaction calcium oxalate is difficult to precipitate. Therefore, in our experiments, the Ca must come from the rest of the cell or from the surrounding medium. The latter possibility is elim- inated by experimental results, discussed on p. 316. Granted that the Ca comes from the protoplasm, it remains to be shown why calcium oxalate does not precipitate in the vacuole until the cell is subjected to the action of stimulating agents. There are two possibilities: (1) The vacuole membrane is impermeable to Ca, but becomes permeable at the moment of stimulation. (2) There is no con- siderable concentration of free Ca in the protoplasm, but bound Ca be- comes free when stimulating agents are applied. The first alternative faces two difficulties. In experiments in which the plasmalemma ruptures in Ca-containing solutions the vacuole imme- diately becomes packed with crystals. The vacuole membrane, there- FREE CA AND STIMULATION OF ELODEA CELLS 321 fore, seems to be very permeable to Ca, although this experiment does not prove certainly that this membrane is permeable when the plasma- lemma is intact. A second difficulty is that the cytoplasm seems to contain some free oxalate, which would bind any pre-existing free Ca. It could be shown experimentally that isolated tonoplasts are permeable to oxalate. If they are washed for an hour in distilled water, and Ca is added, no crystals are formed in the tonoplasts, although they can be formed before the washing. The observation recorded on p. 311, that on electrical dis- integration of the plasmalemma in sugar solution some crystals are formed outside the vacuole, also supports the idea that the protoplasm contains oxalate. Finally, the oxalate must originally form in the cyto- plasm; and would be precipitated as it formed if free Ca were present. These facts support the second alternative, that the Ca which pre- cipitates the oxalate in our experiments is released from a bound state on stimulation. Here the experiments with oxalate (and citrate) must be considered. The plasmalemma is relatively impermeable to sodium ox- alate; cells plasmolyze in hypertonic solutions of this salt and do not deplasmolyze in a short time. Therefore, whenever cells are washed in oxalate, only the outer layer is affected. This is better proved by the fact that the effect of oxalate and citrate can be reversed by solutions containing Ca; if there is doubt about the relative impermeability of the cell to these ions, there is at least less doubt (considering the observations recorded on p. 318) about its relative impermeability to Ca. If oxalate and citrate, in removing Ca from the outer layer of the cell, prevent the precipitation of calcium oxalate by agents which nor- mally cause it, it follows that it is just this Ca in the outer surface that becomes free and forms the crystals in the vacuole. Oxalate and citrate can remove Ca from its combination with proteins. In ordinary analysis all the Ca in blood serum is precipitated by oxalate, although 50 per cent of it is bound. McLean and Hastings, cited above, have shown the same to be true for. citrate. j In this discussion we may not have considered all the possibilities. The argument as it stands, however, indicates that when stimulants act on the Elodea cell, they act primarily on the non-diffusible calcium com- plexes in the cell exterior, releasing free Ca into the interior, some of which precipitates with the free oxalate that is present in the cell interior. This is the argument advanced by Heilbrunn and Daugherty (1933), as an interpretation of the viscosity changes in the exterior and interior protoplasm of Ameba under the influence of ultraviolet rays. Our direct results, therefore, support their contentions so far as this one reaction—Ca release—is concerned. 322 DANIEL MAZIA AND JEAN M. CLARK In considering our results in relation to the Heilbrunn theory, the results of the experiments with anesthetics must be explained. For- tunately, Daugherty (1936; in press) now presents evidence from vis- _cosity measurements that is in accord with our evidence, and yet pre- dictable from the Heilbrunn theory. She has found that the viscosity changes which she interprets on the basis of Ca changes are not affected by anesthetics. What is affected is the “clotting” reaction that nor- mally results from the Ca release; Heilbrunn (1934) had already shown that this reaction was prevented by anesthetics in the presence of an adequate amount of free Ca. SUMMARY We have attempted in a general way to test the hypothesis that a primary action of stimulants on cells is to cause an increase in the free Ca concentration within the cell. 1. The formation of crystals of calcium oxalate in the Elodea cell may be used as an intracellular method for detecting free Ca. 2. These crystals can be positively identified as calcium oxalate. 3. Condenser discharges and continuous direct currents cause the formation of crystals. The crystals are always formed at the anodal end of the cell. 4. Previous washing in solutions of oxalates and citrates prevents the formation of these crystals. 5. Ultraviolet irradiation causes the formation of calcium oxalate crystals. Previous washing in oxalate prevents this. 6. Mechanical stimulation causes the formation of crystals. 7. Heat treatment within the temperature range 40°—55° causes the formation of crystals. Oxalate prevents this. 8. Plasmolysis followed by deplasmolysis causes the formation of crystals. Oxalate and citrate prevent it. 9. The effect of the oxalate and citrate may be reversed by subse- quent immersion in Ca-containing solutions. 10. Isotonic salt solutions, and a series of toxic agents, do not cause the formation of crystals. 11. Anesthetics in non-lethal doses do not prevent the formation of crystals by electric currents or osmotic changes. 12. The results are shown to be consistent with the idea that when stimulants act on these cells, they cause a release of Ca from combina- tions located in the periphery of the cells. BIBLIOGRAPHY ANGERER, C. A., 1936. Effect of mechanical agitation on the relative viscosity of Ameeba protoplasm. Jour. Cell. and Comp. Physiol. (in press). FREE CA AND STIMULATION OF ELODEA CELLS 323 Betue, A., 1915. Kapillarchemische (kapillarelektrische) Vorgange als Grundlage einer allgemeinen Erregungstheorie. Pfliiger’s Arch., 163: 147. Brinks, L. R., 1932. Migration of anthocyan pigment in plant cells during the flow of electric current, and reversal by acids and alkalies. Proc. Soc. Exp. Biol. and Med., 29: 1186. Cuamot, E. M., anp C. W. Mason, 1930. Handbook of Chemical Microscopy. Vol. I. John Wiley and Sons. New York. DaucHERTYy, KatHryn, 1936. The action of anesthetics on Amceba protoplasm. (In press.) Heitprunn, L. V., 1928. The Colloid Chemistry of Protoplasm. G. Borntraeger. Berlin. HEILprunn, L. V., 1934. The effect of anesthetics on the surface precipitation reaction. Biol. Bull., 66: 264. HEILBRUNN, L. V., AND KatHRyN DaucGHeErTy, 1933. The action of ultraviolet rays on Ameeba protoplasm. Protoplasma, 18: 596. HEILBRUNN, L. V., D. Mazia, AND H. B. StTErnBAcH, 1934. Free and bound cal- cium in stimulation and injury. Anat. Rec., 60: (suppl.) 32. HEILBRUNN, L. V., anp R. A. Younc, 1930. The action of ultraviolet rays on _ Arbacia egg protoplasm. Physiol. Zool., 3: 330. Ktune, W., 1864. Untersuchungen ueber das Protoplasma und die Contractilitat. Leipzig. Kuster, E., 1929. Pathologie der Pflanzenzelle. Teil I. Pathologie des Proto- plasmas. G. Borntraeger. Berlin. Leg, A. B., 1900. The Microtomist’s Vade-mecum. P. Blakiston Sons and Co. Fifth Edition. Philadelphia. LepescHKIN, W. W., 1936. Fortschritte der Kolloidchemie des Protoplasmas in den letzten zehn Jahren. III. Protoplasm, 25: 301. McLean, F. C., anp A. B. Hastrnes, 1935. The state of calcium in the fluids of the body. I. The conditions affecting the ionization of calcium. Jour. Biol. Chem., 108: 285. Matpricui, M., 1675 and 1679. Die Anatomie der Pflanzen. (Ostwald’s Klassiker, No. 120. Emgelmann, Leipzig.) Mast, S. O., 1931. The nature of the action of electricity in producing response and injury in Ameceba proteus (Leidy) and the effect of electricity on the viscosity of protoplasm. Zettschr. f. vergl. Physiol., 15: 309. Napson, G. A., ET E. ROoCHLINE-GLEICHGEWICHT, 1927. De la Formation des Cristaux de |’Oxalate de Calcium dans les Cellules Végétales sous 1’In- fluence des Rayons Ultraviolettes (in Russian). Vestnik Rentgenologii i Radiologii, 5: 415. —, 1928. Apparition des Cristaux d’Oxalate de Calcium dans les Cellules Végétales sous l’Influence de la Radiation Ultra-violette. Compt. Rend. Soc. Biol., 98: 363. OsterHouT, W. J. V., 1910. On the penetration of inorganic salts into living protoplasm. Zeitschr. f. Physik. Chem., 70: 408. PatscHovsky, N., 1920. Studien ttber Nachweis und Lokalisierung, Verbreitung und Bedeutung der Oxalsaure im Pflanzenorganismus. Beih. gz. Bot. Zentralbl., 37 (Abt. 1) : 259. ROCHLINE-GLEICHGEWICHT, E. J., 1930. L’Influence de 1’Emanation de Radium sur la Cellule Chlorophyllifere. (In Russian.) Vestnik Rentgenologti i Radiologii, 8: 387. RUBINSTEIN, D. L., UND VERA UspensKaja, 1934. Uber den isoelektrischen Punkt der pflanzlichen Plasmahaut. Protoplasma, 21: 191. Weis, A., 1925. Beitrage zur Kenntnis der Plasmahaut. Planta, 1: 145. Wuerry, E. T., anp G. L. Keenan, 1923. Occurrence and forms of calcium oxalate crystals in official crude drugs. Jour. Am. Pharmaceut. Assoc., 122300. OBSERVATIONS ON THE, MUETIPEICATION OR BAG h- RIA IN DIFFERENT VOLUMES OF STORED SEA WATER AND THE INEEOENGEOn OXYGEN TENSION AND SOLID SURFACES CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON (From the Scripps Institution of Oceanography, University of California, La Jolla, California) INTRODUCTION The storage of raw sea water in the laboratory is usually accom- panied by increased bacterial activity. This may be manifested by the evolution of carbon dioxide, increase in the hydrogen-ion concentration, oxygen consumption, ammonia production, nitrate reduction, or by bac- terial multiplication. Waksman and Carey (1935a) report that the bacterial population of raw sea water increases from a few hundred bacteria per cc. to millions per cc. following a few days storage in the laboratory. They (19350) attribute this great increase primarily to the dying out of Protozoa and other marine animals which devour bacteria and to the modification of “ certain controlling factors injurious to free bacterial development.” Keys, Christensen, and Krogh (1935), who observed similar changes, believe that the activity of bacteria in the sea is limited by the “extreme stability of the ocean as a chemical and physical factor.” | In their preliminary studies on sampling methods, ZoBell and Felt- ham (1934) observed a great increase in the number of bacteria in samples of stored sea water even at refrigeration temperatures and also a diminution in the number of varieties. These changes were found to occur less rapidly in large samples than in small ones. ‘The effect of volume on the multiplication of bacteria in stored water was recognized by both Harvey (1925) and Foyn and Gran (1928), who mention the work of Whipple on stored fresh water. Whipple (1901) noted that the smaller the volume of fresh water, the greater is the rate of bacterial multiplication in it. After 24 hours of storage he found 300 bacteria per cc. in a gallon, 7,020 per cc. in a pint, and 41,400 per cc. in 2 ounces of a representative water sample having an initial count of 77 bacteria per cc. Pitas While it is impracticable either to collect or to store samples of sea 324 EFFECT OF VOLUME ON BACTERIAL ACTIVITY S25 water which are large enough to prevent for more than a few hours undesirable changes in its bacterial and chemical composition, the fact that the volume seems to influence the activity of bacteria may be of significance. For example, in one of their experiments on its organic- matter content Keys, Christensen, and Krogh (1935) ultrafiltered small amounts of sea water free of all particulate and colloidal organic matter. After its inoculation with normal flora there was just as much bacterial multiplication in it as in ordinary filtered sea water. Is it possible that the results were influenced by the smallness of the volume of ultra- filtered sea water? Other questions arise: Is the great increase in the bacterial population which accompanies the storage of sea water in any way related to its confinement to a small volume, and if so, to what extent and how? Also, if bacterial activity is influenced by volume, to what extent do controlled test-tube experiments indicate what is actually occurring in the ocean? In summarizing the factors which influence bacterial growth curves Rahn (1932) concludes that the volume of the media has no influence on the multiplication of bacteria. Woodruff (1911) and Greenleaf (1926) report that certain Protozoa multiply faster in large volumes than in small ones while Robertson (1922) found the antithesis of this. EXPERIMENTAL METHODS Sea water was collected in glass battery jars from the end of the Scripps Institution pier which extends 1,000 feet seaward where the mean depth of the water is about 20 feet. Despite its proximity to land, the water is chemically and biologically quite typical of marine conditions because of the oceanic circulation and the freedom from terrestrial drain- age. The water was siphoned into a Buchner 4G sinter-glass filter and filtered by gravity into a 10-gallon bottle. The filtration was merely to remove particulate matter and to permit the oxygen tension to come to equilibrium with the atmosphere. After thoroughly mixing the water by vigorous shaking to insure greater uniformity in its composition, it was siphoned into different sizes of Pyrex bottles or flasks. Some were filled to capacity and glass-stoppered and others were only partly filled, thus leaving the water in contact with air. Chemically clean, sterilized apparatus and aseptic technique were used throughout the experiments. The bacterial population of the water was determined by plating procedures immediately after its collection from the sea, again after it had been distributed to bottles of different sizes and at periodic intervals thereafter. Two or more dilutions of each sample were plated in dupli- cate. Sterile sea water was used as the diluent. The following nutrient 326 CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON agar has proved to yield larger plate counts than any of several other formule which have been tried : iProfeose | peptomey ssa ser eie eee ener 2.0 grams Bacto apeptomel es cane circle tener 3.0 grams Beefy exttacties cas doce serene tererret 2.0 grams IMIG ON Ashen sesbusgoedsesnJ conde 0.2 gram Bacto aga ten eerie ee ore omiceroeiortee 12.0 grams Seal) water ie ire a see re anys ohacererretnente 1,000.0 cc. After dissolving the agar by steaming, the reaction was adjusted to pH 7.8 with 1/N NaOH and the medium was clarified by filtering it through glass wool. The oxygen content of the sea water was estimated with Winkler’s reagents. RESULTS In the first series of experiments glass-stoppered Pyrex bottles hav- ing similar shapes, proportional dimensions and different capacities were 200 175 Bacteria per cc. X 104 0 2 4 6 8 10 12 14 16 17. — 20 22 Period of Storage in Days. Fic. 1. Influence of volume on the multiplication of bacteria in stored sea water. Growth curve in 10 cc. represented by solid line, 100 cc. by dashed line, 1,000 cc. by dot-dashed line and 10,000 cc. by dotted line. filled about five-sixths full with 10, 100, 1,000, and 10,000 cc. of raw sea water. The air in the upper one-sixth of the bottles provided for the saturation of the water with oxygen and it permitted more thorough mixing when the water was shaken at the time of sampling. The bottled water was stored in a dark constant temperature room at 16° C. Sam- ples were withdrawn daily for plate counts. Fig. 1 shows graphically EFFECT OF VOLUME ON BACTERIAL ACTIVITY 327 the number of bacteria which were found per cc. of water after different periods of storage in bottles of different sizes. This experiment has been repeated several times under a variety of conditions. Almost invariably the densest bacterial populations appear in the smallest volumes of sea water. Table I summarizes the maximum plate counts found in four. different volumes of three different samples of sea water. It is impossible to duplicate results in the usual sense of the term because each different sample of sea water varies in chemical and biological composition. The initial plate counts of the water col- lected for these experiments during the last four years has varied from less than 100 bacteria per cc. to nearly 30,000 per cc. Of greater im- portance is the organic-matter content of sea water which fluctuates in quality and quantity (Bond, 1933). Waksman and Carey (19355) have TABLE I Maximum number of bacteria found per cc. of sea water after storage at 16° C. in glass- stoppered Pyrex bottles of different sizes. SamipleINow. 64... hee cee | 86 149 165 Date collected........... 9/14/35 1/9/36 2/6/36 Initial count per cc....... 930 470 2,570 Volume of sea water Plate count per cc. following storage 1 Oo Ce Rp eae eee A ace 1,050,000 1,960,000 ‘1,810,000 MOORCen a 3 Sees ee 680,000 1,240,000 1,070,000 MROOOKEEES, : cn ee eae te 251,000 645,000 615,000 HOMOOOR CORRS... eee eee 164,000 255,000 340,000 shown that the amount and specific nature of organic substances control the abundance of the bacterial population developing in sea water. As indicated by Fig. 1, during the maximum stationary phase of the growth curves (Winslow, 1928) there may be twice as many bacteria per cc. in small volumes as in volumes ten times as large. It requires from three to six days storage at 16° C. for the maximum to be attained. The densest bacterial populations which appear in the smallest volumes of stored water require the longest periods of incubation. As a general rule, the growth curves in the smaller volumes are still in the phase of logarithmic. increase while the bacterial populations in the larger volumes are declining. Although there is evidence of multiplication within eight hours after the water is bottled, little or no difference is found in the density of the bacterial populations in the different volumes during the first two days. During the phase of decrease the bacterial populations in the smaller volumes continue to exceed those in the larger volumes 328 CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON until the phase of readjustment is reached 8 to 18 days after the sea water was first stored at 16° C. During the latter phase the effect of volume on the plate counts disappears entirely as the bacterial popula- tions fluctuate irregularly from a few thousand to over a hundred thou- sand bacteria per cc. At least a few thousand bacteria per cc. can be demonstrated in any volume of stored sea water for several weeks. After four years storage at 2 to 6° C. in a 100-cc. glass-stoppered bottle. there were 209,000 bacteria per cc. of sea water. In some experiments 1.0 cc. of sea water was stored in test tubes. Each successive day two of these were analyzed by washing them out several times with a 99 cc. dilution water blank. The wash water was appropriately diluted and plated. Also 1.0 cc. volumes were stored in 1.0 cc. pipettes which were filled to the mark, and the water retained by forcing the end of the pipette into a rubber stopper having a small hole bored halfway through it. These were similarly analyzed in duplicate by delivering the contents into 99-cc. water blanks and then washing the pipettes by filling and emptying several times. Table II shows the TABLE II Number of bacteria per cc. of sea water following storage at 16° C. of different volumes in various receptacles. Storage 1.0 ce. in 1.0 cc. in 10 ce. in 100 cc. in period pipettes test tubes test tubes bottles Stanciemrviy ers ae 280 280 280 280 DiGlepy Sees enieye tack petans em 131,000 128,000 115,000 275,000 GiGlayiSme mers seers ee 1,210,000 1,340,000 570,000 580,000 AAV Sor unc eeted sina 1,230,000 1,410,000 955,000 640,000 SHCA Siar rsa cistes eee 1,510,000 1,895,000 640,000 420,000 Opdaiy sei tee gota Pee 1,300,000 1,510,000 280,000 250,000 average number of bacteria found under both conditions of storage on each day, and for comparison the corresponding plate counts in 10 ce. and 100 cc. volumes of identical sea water are also given. It will be observed that again the smallest volume of sea water supported the largest bacterial populations. Following the storage of some 1.0 cc. samples of sea water, more than 4,000,000 bacteria have been demon- strated by plate counts whereas the maximum bacterial populations in 10-liter samples of the water have been only 2 to 5 per cent as much. The same sea water stored in the interstitial spaces of glass beads or sand which may be regarded as multiple minute volumes supports bacte- rial populations exceeding ten million per cc. Inasmuch as it is quite obvious that in small volumes of water, the temperature would change more rapidly than in large volumes, precau- EFFECT OF VOLUME ON BACTERIAL ACTIVITY 329 tions were taken to maintain the different volumes of water at the same temperature. First, all of the water in the 10-gallon reservoir was brought to the desired temperature. The change was usually slight because the temperature of the sea water which was collected for most of the experiments was 13 to 16° and it was stored in a constant tem- perature room at 16°. In order to avoid thermic disturbances, in some experiments the sea water was stored in a water bath at exactly the same temperature as when collected. Later, receptacles of different capacities were filled with filtered sea water. on the end of the pier and these were suspended in the sea, 4 to 5 meters below the surface. The bottles were hauled upon the pier each day to permit sampling. Again many more bacteria per cc. were found in the small bottles than in the larger bottles just as was found in the laboratory. Finally, bottles full of sea water were submerged in the sea stoppered with a diatomaceous earth filter candle in a rubber stopper. In other bottles the mouth was narrowed by stoppering with a one-hole rubber stopper in which was inserted a 5 mm. bore glass tube bent at a sharp angle. Under both conditions the bacterial populations in 100 cc. bottles increased to nearly a million per cc. and in the 10-liter carboy to over a hundred thousand bacteria per cc., whereas the sea water im situ in which the bottles were submerged con- tained less than a thousand bacteria per cc. When sea water is analyzed immediately after collection, from 25 to 35 different species of bacteria can be recognized by a careful scrutiniza- tion of their colonial characteristics. Actually 96 different species of marine bacteria have been isolated and culturally, morphologically and physiologically characterized. Almost before there is evidence of multt- plication in stored sea water, the number of species begins to decrease until at the time the maximum stationary phase in the growth curve is reached, only nine or ten different species are detectable on the plates. Quantitative data cannot be given, but it is our impression that there are usually fewer predominating species in the smaller volumes than in the larger volumes of stored sea water. That many species disappear dur- ing storage and do not merely escape detection by being outnumbered by others which multiply rapidly is indicated by the fact that during the readjustment phase when the populations have decreased to a few thou- sand bacteria per cc., only four or five species can be recognized on the plates regardless of the volume. EFFECT OF OXYGEN Rahn (1932) states that with the same organism in the same medium the final amount of growth “ will be proportional to the volume of the medium unless the volume has had an influence upon the rate of oxygen 330 CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON penetration into the medium.” Whipple (1901) attributed the more rapid multiplication of bacteria in small samples of fresh water. to the greater oxygen content of the water. However, in his experiments Whipple added 0.0005 per cent peptone to the water samples to hasten bacterial multiplication, and the additional peptone would increase the oxygen consumption. Likewise Rahn’s conclusions are based upon re- sults obtained in media rich in organic matter in which there would be appreciable oxygen consumption. ‘There is very little organic matter in sea water. According to Krogh (1931) it contains less than 10 mgm. of organic matter per liter, much of which cannot be utilized by bacteria. Waksman and Carey (1935a) as well as Keys, Christensen, and Krogh (1935) find that bacterial activity is arrested by the depletion of available organic matter in stored sea water before half of the dissolved oxygen is consumed. Tase III The initial and maximum plate counts per cc. of sea water stored in partly filled botiles of different sizes and the oxygen content of similar water in open botiles after 7 days quiescent storage at 16° C. Plate count per cc. Oxygen in cc. per liter Volume of sea water Initial Maximum Initial After 7 days CC. CC. CC. NOOR ete Mec he: 460 1,360,000 5.46 5.14 TOOO 2h. hy eect ok 460 670,000 5.46 5.39 LOOOO Ree eae 460 485,000 5.46 5.38 Those who have criticized our paradoxical observations that small volumes of stored sea water support larger bacterial populations than large volumes suggest that the explanation will most likely be found in the difference in the oxygen content of the water. Therefore this fac- tor has been carefully considered from various angles. In bottles which were only partly filled with sea water it was found that the contact with the air plus the daily shaking at the time the samples were taken kept the water virtually saturated with oxygen. The oxygen content of the water under these conditions was decreased only slightly even when not shaken. This is illustrated by the data in Table III, which give the maximum bacterial populations which were attained during the storage of different volumes of sea water and the oxygen content of similar water after 7 days quiescent storage. In the next series of experiments the bottles were filled full, glass- stoppered and submerged in a water bath at 16° C. Duplicate 100 cc. and one-liter bottles (actually somewhat more than 120 and 1,200 cc. Te eo en EFFECT OF VOLUME ON BACTERIAL ACTIVITY SO respectively) were analyzed each day for their bacterial and oxygen content. Samples were withdrawn from the 10,000 cc. bottle each day by means of a series of siphons and traps which permitted the with- drawn samples to be replaced by similar sea water from another recep- tacle without coming into contact with air. The pertinent data of a representative experiment are summarized in Table IV. In every case the oxygen consumption per unit of water was greatest in the smallest volumes in which were found the largest bacterial populations. This would seem to indicate that the oxygen content of the water does not account for the increased bacterial activity in small volumes of sea water. It may be of interest to note that oxygen consumption in any volume of stored sea water parallels the growth curves until the maximum sta- tionary phase is reached. However, during the period of logarithmic de- TABLE IV The maximum plate counts per cc. of sea water stored in completely filled glass-stoppered botiles of different capacities and the oxygen content of the water after 18 days storage at 16° C. Plate count per ce. Oxygen per liter Oxygen ee ee Initial Maximum Initial 18 days . CC. co. CC. cc. OO Rear: et ete ae 276 1,010,000 5.40 2.87 2.53 OOO ao Kerosene aie 276 580,000 5.40 4.09 1.31 NOKOOO Mees sites si bettas 276 370,006 5.40 4.43 0.97 crease, oxygen consumption is not retarded but continues in a straight line at the same rate as it does during the logarithmic phase of growth. This is illustrated by Fig. 2 which gives the growth and oxygen-consump- tion curves for 100 and 1,000 cc. of stored sea water. Although there were approximately the same number of viable bacteria per cc. in both volumes as demonstrated by plate counts after the tenth day, twice as much oxygen was consumed per unit of water in the 100-cc. receptacle as in the 1,000-cc. receptacle. This may indicate that there are many more bacteria which are respiring than the number which can be demon- strated by plating procedures. The oxygen-consumption curves do not continue as straight lines until all the oxygen is depleted, as may be inferred from Fig. 2. In- stead the curves start to flatten out between the fifteenth and twenty- first days, and oxygen is consumed only very slowly after 28 days storage at 16° C. Oxygen consumption in liter samples continues at a slower rate and for longer periods of time than in 100-cc. samples. Je CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON Oxygen consumption in 10-liter samples has not been followed for more than three weeks. Regardless of the volume of the samples, oxygen is not depleted from sea water which is initially saturated (10sto Ss: Sice! oxygen per liter) unless utilizable organic matter is added. Respiration as indicated by oxygen consumption virtually ceases while there is still between 2 and 3 cc. of oxygen per liter in solution. However, upon the addition of only small amounts of organic matter, the oxygen content is rapidly reduced to zero in any volume of sea water from which at- mospheric oxygen is excluded. Strips of filter paper and even agar-agar suffice. It has been shown by ZoBell and Anderson (1936) that bac- 100 80 60 Bacteria per cc. X 104 40 Oxygen in cc. per Liter 20 0 2 4 6 8 10 12 14 +16 iis un DON nen 20) Period of Storage in Days. Fic. 2. Influence of volume on oxygen consumption in stored sea water. Solid lines represent growth curves and dashed lines oxygen consumption in 100 and 1,000 cc. respectively. teria not only deplete the oxygen from marine bottom deposits but that they actually create an oxygen deficit. As will be discussed below, the volume effects disappear when more than a few milligrams of utilizable organic matter per liter is added to the sea water. As further evidence that the differential penetration of oxygen in sea water in receptacles of different capacities does not account for the increased bacterial activity in small volumes, sea water having different initial oxygen contents was stored. ‘Twenty liters of freshly collected sea water was deoxygenated by holding it for two hours at a pressure reduced to less than 1 cm. of mercury with a High-Vac pump during which time the water was agitated. Then pure nitrogen was used to EFFECT OF VOLUME ON BACTERIAL ACTIVITY 333 force the water into 120-cc. bottles which were immediately stoppered and sealed. Under these conditions sea water containing only 0.59 cc. oxygen per liter was obtained. Some of this deoxygenated water was mixed with oxygen-saturated sea water to give a series having an initial oxygen content of 3.08 cc. per liter. As a control some of the deoxy- genated water was saturated with oxygen by shaking it with air. It then contained 5.42 cc. of oxygen per liter. The tightly stoppered bottles were submerged in a water bath at 16° C. After different periods of incubation duplicate bottles were analyzed for bacteria and oxygen con- tent. Table V presents the findings. There was considerable bacterial activity in the deoxygenated water which contained only 0.59 cc. oxygen per liter as manifested by the rapid increase in the bacterial population and the depletion of the oxygen. As ADasiniay WE Bacterial and oxygen content of sea water initially containing different concentrations of dissolved oxygen after storage at 16° C. in 120-cc. bottles. Oxygen-saturated water | Partly saturated water Deoxygenated water siorake perio Baveeten Oxygen Boeteni Oxygen asian Oxygen ee ad a cece te celine 1) Or Startienvet: Seen sk 1,190 5.42 1,230 3.08 1,240 0.59 SGEA Sane Cemidaas 264,000 5.12 224,000 2.91 216,000 0.42 SidaySis.cs este: 1,458,000 4.68 1,170,000 2.48 660,000 0.22 TES i eee 792,000 | 4.35 | 1,226,000] 2.26 | 683,000 | 0.00 lO) GEWSeecdoous. 116,000 3.82 469,000 1.65 152,000 0.00 IS) GENS. 5505 odo 77,000 2.90 17,000 1.19 28,000 0.00 a matter of fact, the multiplication rate in the deoxygenated water was only slightly less than in the oxygen-saturated water until the oxygen was depleted. Likewise it will be observed that the amount of oxygen con- sumed during the first seven days was virtually the same regardless of the initial oxygen content of the water. Bacterial multiplication in sea water having a low oxygen tension was expected because most marine bacteria are microzrophiles and many of them are facultative anaerobes. However, the rate of multiplication and oxygen-consumption in the water of low oxygen tension proved to be surprisingly rapid and these results are reported only after having been repeated. Finally, similar volumes of water were placed in receptacles of dif- ferent shapes so that different surface areas were exposed to the air. For example, 500 cc. quantities of water were placed in liter Roux flasks, in 500 cc. graduate cylinders and in glass tubes 2.1 cm. in diameter by 334 CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON 140 cm. long. When a Roux flask was placed on its side, 220 sq. cm. of water surface was exposed to air. An identical volume of water in a Roux flask stood on end exposed only 40 sq. cm. of water surface to air. In the 500-cc. cylinder 16 sq. cm. was exposed to air and in the 2.1 cm. tube only 3.2 sq. cm. of the water surface was exposed to air. Although it is recognized that sampling difficulties introduced certain errors, the bacterial populations were found to be in no way related to the area of water surface exposed to air. However, a relationship was noted be- tween the area of glass surface exposed to water and the bacterial popu- lations in stored sea water. EFFECT OF SOLID SURFACE It is well known that the smaller the capacity of similarly shaped re- ceptacles the greater is the surface area per unit volume. This may be illustrated by the data in Table VI which gives the dimensions of the TasBle VI Volume of sea water stored in different receptacles, area of solid surface in contact with the water, ratio of volume in cc. to area in sq. cm. and the average * maxt- mum. bacterial populations per cc. after storage at 16° C. Volume of Solid surface Ratio of Average maximum * sea water in cc. in sq. cm. cc. : sq. cm. bacteria per cc. NA Sen Mae 37 1: 2.64 1,863,000 AU fete Brahe St en 148 1: 1.23 1,070,000 DD emi susinos 640 1: 0.52 553,000 TS 2208 ta seach as 3174 1: 0.24 261,000 * Average of three different experiments. receptacles which have been in use in the foregoing experiments. The average maximum number of bacteria which have been found in these receptacles in three different experiments is also given. It will be ob- served that the densest bacterial populations appear in the bottles of water which offer the largest area of glass surface per unit volume of water. As reported above, this was likewise found to be true when the volume was kept constant (500 cc.) and the shape of the receptacles varied to give different glass surface areas. Next receptacles were selected so that their volumes varied greatly but the ratios of glass surface to volume remained practically constant. This was accomplished by using different lengths of glass tubing 2.1 cm. in diameter, one end of which was sealed and the other end stoppered. From 28 to 925 cc. of sea water was stored in such cylinders and within EFFECT OF VOLUME ON BACTERIAL ACTIVITY 335 the limits of experimental error the growth curves ran practically paral- lel. Table VII summarizes typical bacteriological findings in experi- ments in which receptacles of assorted capacities and shapes giving a wide range of ratios of volume to solid surface were used for storing sea water. Prior to being used all the glassware was aged in sea water, cleaned with hot sulfuric acid dichromate solution, rinsed with distilled water and sterilized. While the expression of the relationship of glass surface to bacterial populations defies exact mathematical expression, it will be noted that: in general, the larger the ratio of volume to surface Wai WOU Volume of sea water stored in different receptacles, area of solid surface in contact with the water, ratio of volume in cc. to area in sq. cm. and the average maxt- mum. bacterial population following storage at 16° C. Vo ume Area of Ratio Bacteria Receptacle of solid of ce. : per sea water surface sq. cm. (oes I ORECMPIPEtiCman aris Lo See ee etek 1.0 cc.} 9.9sq.cm.|1: 9.9] 2,480,000 Blake bottle on side.............. 10.0 cc. 98.7 sq. cm. | 1 : 9.9] 3,120,000 1eSicma test tubes. ocsce. ce See ae 1.0 cc. 4.1sq.cm.|1:4.1} 1,895,000 2 5em test €UD@s shine seek odes 10.0 ce. 20.3 sq.cm.}1:2.0} 855,000 2.1 cm. glass tube, 8 cm. long...... 28.0 cc. 50.4sq.cm.|1:1.8] 970,000 2.1 cm. glass tube, 140 cm. long....} 550.0 cc.| 925.0sq.cm.|1:1.7] 720,000 Nest of glass tubes, 140 cm. long *..| 410.0 cc. | 2,556.0 sq. cm.| 1 : 6.2] 1,470,000 Witerstbottles othe teeta es 1,225.0 cc.| 640.0sq.cm.}1:0.5| 490,000 Liter bottle with 4,100 3 mm. glass beadse kn tc ere raion © 1,067.0 cc. | 1,690.0 sq.cm.}1:1.7] 642,000 Liter bottle with 1,350 cm. 3 mm. SlaASSpKOd Sr iessts4 ke aI ec creas 1,099.0 cc. | 1,840.0 sq.cm.}1:1.7| 598,000 * The nest of glass tubes consisted of a piece of 2.1 cm. glass tubing 140 cm. long in which was placed three glass cylinders nearly as long and 1.6, 1.0 and 0.6 cm. in diameter respectively. area, the larger the bacterial populations regardless of the total volume of stored sea water. In the liter bottles in which the ratios of volume to solid surface were increased by placing a shallow layer of 3-mm. glass beads or. glass rods in them, the plate counts were higher than in ordinary liter bottles but the increases were not proportional to the additional surface area nor were they as high as in other receptacles having the same ratios (see Table VII). This is considered as evidence that another factor. besides solid surface is involved which influences the activity of bacteria in stored sea water. Further proof was obtained by covering the bottoms of liter bottles with enough glass wool or fine silica sand to give ratios exceeding 1:1,000. Following the storage of sea water under these conditions, the 336 CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON plate counts were only slightly higher than in ordinary liter. bottles having ratios of only 1:0.5 and about the same as in liter bottles containing enough 3-mm. glass beads to give ratios of 1:1.7. The explanation 1s believed to be found in the proximity of the water to a solid surface or in the distribution of the solid surface throughout the water. For ex- ample, in a liter bottle whose diameter is 10.6 cm., part of the water is at least 5 cm. away from the nearest solid surface whereas in a 2.1 cm. tube, the greatest distance is only about 1 cm. and within the interstitial water between 3-mm. glass beads the greatest distance of any particles of water to the nearest solid surface is only a fraction of a millimeter. Thus while a layer of glass beads or sand in the bottom of the bottle increases the surface per unit volume, the additional solid surface affects only that part of the water which is in the proximity of the glass beads or sand. Sea water. stored within the interstices of small inert particles such as sand or glass beads supported larger bacterial populations than it did under any other condition of storage. Fine silica sand was treated with hot sulfuric acid dichromate cleaning solution and washed with distilled water until the reaction was neutral. The clean sand was then dried in an oven after which 25-gram portions were weighed into 100-cc. Pyrex bottles. These bottles containing sand were heated to 240° C. for 6 hours. When cool, the sand was moistened with 10 cc. of sea water which was just enough to barely cover the sand. After varying periods of storage the sea water and sand mixture was washed into a 900-cc. water blank with 90 cc. of sterile sea water. Plating dilutions of this revealed the presence of from six to twelve million bacteria per cc., whereas less than a million per cc. appeared in identical sea water when 10 cc. was stored in similar bottles without sand. Between four and six million bacteria per cc. were demonstrated in the sea water with which the small glass beads in a 100-cc. bottle were covered. Controls consisting of clean sand or glass beads covered with distilled water and inoculated with a mixed microflora showed no appreciable bacterial mul- tiplication, thereby indicating that neither the sand nor the beads sup- plied utilizable nutrients. The high counts which are obtained when sea water is stored in sand or glass beads are attributed to the proximity of the water to solid surfaces. There may be even more bacterial multi- plication under these conditions than is indicated by the plate counts because it is improbable that all of the viable bacteria are recovered from the sand or beads. Early in this work it was recognized that marine bacteria have a tend- ency to adhere to solid surfaces. Consequently in all of the experiments the water was vigorously shaken prior to sampling in order to recover EFFECT OF VOLUME ON BACTERIAL ACTIVITY 337 as many bacteria as possible. Usually the bottles were only partly filled with water to permit thorough mixing except in the experiments dealing with oxygen consumption in stored sea water. The 1- to 10-cc. volumes of stored sea water were poured directly into the dilution water and the receptacle in which the water had been stored was rinsed out several times. Receptacles which were filled to capacity were partly emptied into sterile vessels to permit shaking. The sea water stored in long glass tubes was poured out and in several times to wash the bacteria from the glass surfaces to which they adhere. In order to show quantitatively the tendency of bacteria to adhere to the glass, 100 cc. of sea water was stored in 120-cc. bottles, two of which were analyzed daily. First, all of the water was transferred to another vessel without any shaking and this was appropriately diluted and plated. TasleE VIII Number of bacteria per cc. in 100 cc. of sea water poured from bottles 1n which it had been stored without shaking and the number per cc. 1n the first and second 100 cc. of sea water with which the original botile was rinsed. Number of bacteria per cc. Storage period 2 Original water First rinse water Second rinse water Staieys cocletan creeks 2,800 0 0 DELAY Siaihs 2.2 ax. wate 65,000 48,000 2,600 SUG S\ca sie sina ae. 108,000 67,000 3,900 AAV Sie oko cls oe 220,000 90,000 10,600 EGE eeiais pereree os 278,000 122,000 6,450 OVday se Pls succes 270,000 136,000 10,200 WOday Ss wi sca ees 10,000 120,000 9,600 Then the original bottle in which the sea water had been stored was thoroughly rinsed with 100 cc. of sterile sea water and this was poured off and plated. The rinsing process was repeated with a second 100 cc. of sterile water which was plated. The resulting plate counts are given in Table VIII. The data show that on an average nearly half of the bacteria in the bottles of stored sea water were stuck firmly enough to the walls of the bottle to resist being detached by pouring off the water without shaking. The foregoing procedure fails to demonstrate all af the bacteria which are attached to the glass because many are not washed off by any amount of rinsing in water. After being rinsed with several 100-cc. portions of water, 20 cc. of melted nutrient agar cooled to 45° C. was poured into the original bottle in which the sea water had been stored and the agar was permitted to solidify on the inside periphery of the 338 CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON bottle Esmarch-tube fashion. Following incubation too many colonies developed to permit their enumeration. Direct microscopic examination (Henrici, 1933) of slides which were suspended in liter bottles of stored sea water for 24 hours revealed the presence of over 100,000 bacteria per sq. cm. If bacteria were equally abundant on all of the glass surface within the bottle, it is calculated that for each cc. of water there was the equivalent of 84,000 bacteria attached to the glass while the water itself contained only 31,000 bacteria per cc. including those which could be dislodged by shaking vigorously the bottle of water. The proponderance of attached bacteria increases progres- sively during storage until they become too thick to enumerate with precision. It is recognized that all of the bacteria observed micro- scopically may not be alive and also that plate counts do not demonstrate all of the viable bacteria in the water. Therefore the data obtained by the two procedures are not directly comparable. However, the data conclusively demonstrate that large numbers of bacteria in stored sea water attach to glass surfaces. The development of micro-colonies on the slides immersed in sea water indicates that many of the attached bac- teria are alive and multiplying. The respiration of these attached bac- teria is believed to account for the oxygen consumption which continues to occur in stored sea water after the bacterial population of the water has dropped to a low level (see Fig. 2). In these experiments the bacteria were not fixed to the slides by heat or otherwise. The criterion of periphytic bacteria is their ability to attach themselves so tenaciously to the slide that they are not dislodged by washing in water and the other staining procedures. According to ZoBell and Allen (1933) appreciable numbers of bacteria attach them- selves to glass slides within 2 to 4 hours after being submerged in sea water. They (1935) found as many as 720,000 bacteria per sq. cm. after the slides had been suspended in the sea for 24 hours. Nearly a billion bacteria per sq. cm. have been observed on glass slides submerged in the sea for a week. DISCUSSION The maximum bacterial population which sea water filtered free of particulate matter can support is probably between ten and one hundred million per cc. (including estimated periphytes). Bacterial multiplica- tion in stored sea water is undoubtedly limited by organic matter because over a billion bacteria per cc. have been found in sea water enriched with 100 mgm. of peptone per liter. However, when enriched there is no apparent difference in the activity of bacteria in different volumes of EFFECT OF VOLUME ON BACTERIAL ACTIVITY 339 sea water. The fact that the volume effects occur only in very dilute nutrient solutioris may explain why exhaustive growth curve studies of other investigators have not shown volume of the media or solid surfaces to be a factor which influences bacterial multiplication. Most growth curve studies on saprophytes have been conducted in media containing at least 1,000 mgm. of organic matter per liter (Rahn, 1932). Using dif- ° ferent carbon compounds Stephenson (1930) shows that 10 to 100 mgm. per liter is the minimum concentration in which Bact. coli and certain other. bacteria can grow. It will be recalled that sea water contains less than 10 mgm. of organic matter per liter. Multiplication and oxygen consumption are not the only bacterial activities in stored sea water which are influenced by volume or surface. It was found that the minimum concentration of organic matter which limits denitrification was considerably less in 10-cc. quantities of sea water than in 500-cc. quantities. This was not understood until the effect of solid surface on bacterial activities in dilute nutrient solutions was studied. Several Erlenmeyer flasks varying in size from 25 to 6,000 cc. were filled with raw sea water enriched with 10 mgm. peptone and 1.0 mgm. KNO, per liter. It was found that nitrite was depleted from the smallest flasks first and from the largest flasks last or not at all. When attempts were made to accelerate the reaction by adding 100 mgm. or more of peptone per liter, there was no difference between small and large receptacles in the rate of nitrite reduction. Changes in the pH and ammonia content likewise occur more rapidly in sea water stored in small receptacles than in identical sea water stored in large receptacles. The volume effects are attributed to the favorable influence of solid surfaces and to the proximity of these surfaces to the water particles. Solid surfaces are believed to favor bacterial activity in dilute nutrient solutions in several ways. First, they provide a resting place for peri- phytes. In discussing the advantages of semi-solid media, Hitchens (1921) suggests that bacteria may need objects upon which to cling while growing. It is difficult to estimate what proportion of marine bacteria are periphytes but ZoBell and Allen (1933) have demonstrated that many species have periphytic tendencies and that at least some of them are obligate periphytes. This is in agreement with field observa- tions. Lloyd (1930) calls attention to the fact that the saprophytic bac- teria of the sea are not planktonic but are attached to suspended particles. Waksman and co-workers (1933) find a definite parallelism between the bacteria and plankton in sea water and they conclude that “ bacteria exist only to a very limited extent in the free water of the sea, but are largely attached to the plankton organisms.” This conclusion is cor- 340 CLAUDE E. ZOBELL AND D. QUENTIN ANDERSON roborated by the literature cited by Benecke (1933) dealing with the distribution of bacteria in the sea. Secondly, solid surfaces favor bacterial activity in dilute nutrient solutions because the nutrients may become concentrated in a film upon the solid surface due to adsorption or other kinds of physical attraction. Studies on the micro-organic films which form on glass slides soon after being submerged in sea water give some evidence of such a concentration. The accumulation of a film of organic matter upon submerged slides can be demonstrated by differential stains as well as by micro-chemical analy- ses. This hypothesis is supported by the fact that the favorable effect of solid surfaces is most striking in dilute nutrient solutions and by the fact that a large percentage of the bacteria multiply on the glass surface although many of them are not firmly attached to it. Third, the solid surface, the capsular-like material which constitutes the holdfast of periphytic bacteria and the adsorbed organic matter may help retain the exo-enzymes near the bacteria until they can digest and ingest the nutrients. The interstices at the tangent of the bacterial cell and the glass surface may serve as concentration foci which retard the diffusion of the enzymes and the digestion end-products away from the cell, The inhibiting effect of continuous gentle agitation on the multi- plication of bacteria in stored sea water is attributed to an acceleration in the diffusion of either the exo-enzymes or the digestion end-products away from the bacterial cells. It is generally recognized that bacterial adsorption to particles exerts a favorable influence on bacterial enzymatic activity. Besides being of interest to students of bacterial physiology, this solid surface phenomenon is believed to be an important factor in ex- plaining the vertical distribution of bacteria in the sea. In the ocean, bacteria are most abundant in the zone 10 to 50 meters below the surface where there are usually a few hundred or less bacteria per cc. Bac- teria are only rarely encountered at depths exceeding 200 meters except on the sea floor where the bottom slime may contain millions of bacteria per cc. regardless of the depth of the overlying water. However, water collected from depths exceeding 200 meters contains sufficient nutrients to give rise to ten to a hundred million bacteria per cc. when the water is stored in small volumes. Why then is this deep water so sparsely populated? As ZoBell (1934) has shown, it cannot be attributed to the prevailing low temperature or high hydrostatic pressure. In view of the foregoing experimental observations, it is believed that the dearth of particulate matter to serve as solid surfaces is a factor which restricts the activity of bacteria in deep water. EFFECT OF VOLUME ON BACTERIAL ACTIVITY 341 SUMMARY The storage of sea water is accompanied by a great increase in the number of bacteria and a decrease in bacterial species. Appreciably more bacteria per cc. appear in small volumes of sea water than in large volumes of identical water. Although the oxygen content of sea water is a factor which influences the activity of bacteria, it does not account for the denser bacterial populations which appear in the smaller volumes. The favorable influence of small volumes is attributed to the contact of the water with the larger solid surface area in the small receptacles. Between ten and a hundred million bacteria per cc. have been demon- strated in sea water stored in sand which presents an enormous surface area whereas less than three hundred thousand bacteria per cc. appear in similar water stored in 10-liter bottles presenting relatively little surface area. The beneficial effect of small volumes and their large surface areas occurs only in dilute nutrient solutions. Upon the addition of 10 to 100 mgm. of organic matter to sea water, the volume effects disappear. Solid surfaces may serve as a resting place for periphytes, they may help concentrate the nutrients by adsorption or otherwise, or they may favor bacterial enzymatic activity and the absorption of metabolites. Appreciation is here expressed for the criticisms and suggestions of Drs. Blodwen Lloyd, Austin Phelps, and Roger Revelle. Acknowledg- ment is also made to Mr. Hiomi Nakamura who made the oxygen de- terminations and to Mr. Carl I. Johnson for technical assistance. LITERATURE BENECKE, W., 1933. Bakteriologie des Meeres. Abderhalden’s Handb. der biol. Arbeitsmethoden, 404: 717. Bonp, R. M., 1933. A contribution to the study of the natural food-cycle in aquatic environments. Bull. Bingham Oceangr. Col., 4: 1. Foyn, B., anp H. H. Gran, 1928. Uber Oxydation von organischen Stoffen im Meerwasser durch Bakterien. Avhand. Utgitt av det Norske Videnskaps- Akad., 1 Oslo, No. 3. GREENLEAF, W. E., 1926. The influence of volume of culture media and cell prox- imity on the rate of reproduction of Infusoria. Jour. Exper. Zool., 46: 143. Harvey, H. W., 1925. Oxidation in sea water. Jour. Mar. Biol. Assoc., 13: 953. Hewrici, A. T., 1933. Studies of freshwater bacteria. I. A direct microscopic technique. Jour. Bact., 25: 277. Hitcuens, A. P., 1921. Advantages of culture mediums containing small per- centages of agar. Jour. Infect. Dis., 29: 390. Keys, A., E. H. Curistensen, AND A. Krocu, 1935. The organic metabolism of sea-water with special reference to the ultimate food cycle in the sea. Jour. Mar. Biol. Assoc., 20: 181. 342 CLAUDE E. ZOBELL AND D, QUENTIN ANDERSON Krocu, A., 1931. Dissolved substances as food of aquatic organisms. Rapports et Proces-Verbaux des Reunions, 75: 7. Luoyp, B., 1930. Bacteria of the Clyde Sea area: a quantitative investigation. Jour. Mar. Biol. Assoc., 16: 879. Raun, O., 1932. Physiology of Bacteria. P. Blakiston’s and Son, Philadelphia. Ropertson, T. B., 1922. Reproduction in cell-communities. Jour. Physiol., 56: 404. SrePpHENSON, M., 1930. Bacterial Metabolism. Longmans, Green and Co., New York. Tanner, F. W., anp D. L. ScHNneEter, 1935. Effect of temperature of storage on bacteria in water samples. Proc. Soc. Exper. Biol. and Med., 32: 960. Waxsmavy, S. A., H. W. Reuszer, C. L. Carey, M. Hotcuxiss, anp C. E. RENN, 1933. Studies of the biology and chemistry of the Gulf of Maine. III. Bacteriological investigations of the sea water and marine bottoms. JBvzol. Bull., 64: 183. WaksmMan, S. A., anp C, L. Carey, 1935a. Decomposition of organic matter in sea water by bacteria. I. Bacterial multiplication in stored sea water. Jour. Bact., 29: 531. Waxsman, S. A., anp C. L. Carey, 1935b. Decomposition of organic matter in sea water by bacteria. II. Influence of addition of organic substances upon bacterial activities. Jour. Bact., 29: 545. Wuiprte, G. C., 1901. Changes that take place in the bacterial contents of waters during transportation. Tech. Quart., 14: 21. Winstow, C.-E. A., 1928. The Rise and Fall of Bacterial Populations. Jordan and Falk, “ Newer Knowledge of Bacteriology and Immunology, Univer- sity of Chicago Press, Chicago, p. 58. Wooprurr, L. L., 1911. The effect of excretion products of Paramecium on its rate of reproduction. Jour. Exper. Zo6l., 10: 557. ZoBELL, C. E., 1934. Microbiological activities at low temperatures with par- ticular reference to marine bacteria. Quart. Rev. Biol., 9: 460. ZoBEt., C. E., anp E. C. Atren, 1933. Attachment of marine bacteria to sub- merged slides. Proc. Soc. Exper. Biol. and Med., 30: 1409. ZoBELL, C. E., anp E. C. ALLEN, 1935. The significance of marine bacteria in the fouling of submerged surfaces. Jour. Bact., 29: 239. ZoBELL, C. E., and D. Q. ANnpeErson, 1936. Vertical distribution of bacteria in marine sediments. Bull. Am. Assoc. Petrol. Geol., 20: 258. ZoBett, C. E., anp C. B. FetrHam, 1934. Preliminary studies on the distribution and characteristics of marine bacteria. Scripps Inst. Oceanog., Bull., tech. ser., 3: 279. : Et Oi MeIZAATION OF SOUUMES BY MOS@OUITO) LAR VAs WILLIAM TRAGER (From the Department of Animal and Plant Pathology of the Rockcfeller Institute for Medical Research, Princeton, N. J.) INTRODUCTION One of the important factors conditioning the suitability of a habitat for the development of mosquito larve is the available food supply. Chiefly for this reason, a number of investigators (see Hinman, 1930, for a summary of the literature) have studied the food requirements of mosquito larve. Since these insects usually ingest large amounts of particulate matter, it has been commonly concluded that this material is the principal, if not the only, source of food. However, in 1907, Putter had shown that the chief food source for some aquatic organisms consists of organic colloids and solutes, and Matheson and Hinman (1930) suggested that such was probably the case with mosquito larve. Hinman (1930) presented evidence in support of this suggestion. He was able to rear small numbers of sterile Ades egypti in water very rich in organic matter that had been filtered through a Berkefeld N filter. Later (Hinman, 1932) he found that anopheline larve, in Berkefeld filtrates of the water of their natural habitat, could survive about 11 days and that some reached the second instar. If such water was sub- jected to dialysis against distilled water, and the dialysate sterilized by Berkefeld filtration, a medium was obtained containing only the solutes of the original water. In such a medium, sterile anopheline larvz died within 3 days without reaching the second instar. Shipitsina (1930) obtained results similar to those of Hinman. In Berkefeld filtered water, anopheline larve made a slight growth, some reaching the second instar. She concluded that while organic colloids play some part in the nutrition of the larve, by far the larger role must be attributed to particulate matter. Neither Hinman nor Shipitsina obtained any conclusive data con- cerning the ability of mosquito larve to utilize substances in true solu- tion. The work upon this problem to be reported in this paper was made possible by the development (Trager, 1935a, b) of a reliable method of rearing mosquito larve free from microorganisms. Three different lines of evidence were obtained all supporting the view that mosquito 343 344 WILLIAM TRAGER larve can utilize solutes. First it was found that a growth factor re- quired by the larve and present in Lilly liver extract No. 343 is water- soluble. Then a study of the effects of several substances of known chemical composition and known to exist in water in true solution showed that one of them, calcium chloride, is a substance essential for the normal development of the larve. Finally, it was possible to rear mosquito larve to the fourth instar in media ultrafiltered through collodion. METHODS Recently laid eggs of the yellow fever mosquito, 4des egypti (a stock colony of which was maintained), were freed from microorganisms by a 15-minute immersion in a fluid recommended by White (1931) and consisting of mercuric chloride 0.25 gram, sodium chloride 6.5 grams, hydrochloric acid 1.25 cc., ethyl alcohol 250 cc., and distilled water 750 cc. The eggs were handled in small glass “ boats” (MacGregor, 1929) made from coverslips, were rinsed in sterile tap water, and inoculated into tubes containing an autoclaved 0.5 per cent solution of Lilly liver extract No. 343. The bacteria-free larvee which subsequently hatched in this medium were washed in sterile distilled water and counted num- bers of them were inoculated into the experimental tubes. For most of the work 18 < 160 mm. test tubes holding 6 cc. of medium were used. Each tube received three larve. The tubes were incubated at 28 + 1° C. and were observed daily or every other day, the number of larve in each stage of development in each tube being noted. Figure 1 is a photograph of a typical culture tube in which one of the three larvee has already pupated while the other two are in the fourth instar. Asep- tic precautions were observed throughout, all the tubes were tested for sterility, and contaminated ones were discarded. EVIDENCE THAT THE GRowTH Factor (4) Exists as A SOLUTE The first evidence that the larve could utilize substances in true solu- tion was obtained in the course of the study of the properties of the accessory growth factor present in liver extract (factor 4). In the presence of whole killed yeast and at concentrations of Lilly liver extract No. 343 below 0.5 per cent, the growth of A. egypti larve is directly proportional to the concentration of liver extract and hence of factor A (Trager, 1936). There can thus be no doubt of the utilization of factor A by the larve. When a solution of liver extract 343 was dia- lyzed 24 hours in a collodion bag against an equal volume of distilled water, the concentration of A was as great in the outer as in the inner liquid. The factor thus dialyzes readily and must be present in true solution. SOLUTE UTILIZATION BY MOSQUITO LARVAL 345 Tue Errects oF CERTAIN SOLUTES OF KNOWN CHEMICAL COMPOSITION Tubes were prepared each containing about 3040 mg. powdered Harris vitamin B-free casein. They were sterilized by dry heat and to each tube was added the desired liquid medium, previously sterilized by autoclaving or by Berkefeld filtration. In tubes with casein and distilled water, first instar larvee died within 2 to 3 days. The presence Nie oO Ge Fic. 1. Photograph of a tube containing casein, yeast extract, and liver ex- tract. One of the three larve has pupated while the other two are in the fourth instar. Note the casein at the bottom of the tube. Photographed by J. A. Carlile. of 0.1 M glucose or sucrose did not prolong their life. However, with 0.1 M maltose several larve survived about a week. In the presence of a salt mixture containing NaCl, CaCl,, NaH,PO,:-H,O and K, HPO, in a total concentration of 0.05 M, several larve survived over a week, while one reached the second instar and lived for 22 days. ‘This would indicate an effect of the soluble salts on the longevity of small larve which were unable to grow because of the absence of accessory growth factors. 346 WILLIAM TRAGER Experiments were then performed to test the effect of solutions of salts and sugars on the growth of larve in the presence of accessory growth factors. It was previously shown (Trager, 1935b) that A. egypti larve could be reared to maturity on a diet of casein, yeast ex- tract (0.3 per cent yeast vitamin Harris) and liver extract 343. It was later found (Trager, 1936) that yeast extract contains the accessory growth factor present in liver extract (factor 4) as well as a second accessory factor. It was accordingly possible to rear larve to maturity on a diet of sterile casein and autoclaved yeast extract. The minimum concentration of yeast vitamin Harris needed to supply enough factor A was 0.8 per cent. In such a medium, half of the larve inoculated would usually reach the adult stage within about 3 weeks. This medium, consisting of casein and 0.8 per cent yeast vitamin Harris, was adopted as a basic medium and into it were incorporated various salts and sugars. Glucose, sucrose, and maltose in concentra- tions of 0.15, 0.10, and 0.05 M gave no better growth than that in the basic medium alone. The same was true of NaCl solutions in concen- trations of 0.075, 0.050, and 0.025 M, as well as of various NaCI-KCl and NaCl-MgSO, mixtures. In all these media, on the sixteenth day, none of the larve had pupated and only about half had reached the fourth instar. But in the presence of both NaCl and CaCl, a striking acceleration of growth was noted. In a medium containing 0.04 M NaCl and 0.01 M CaCl, all of the larve reached the fourth instar in an average time of 8.5 days, two-thirds pupated in an average time of 12.5 days and gave normal adults + 2 days later. In a medium containing 0.045 M NaCl and 0.005 M CaCl, all the larve reached the fourth instar in an average time of 6 days, all pupated in an average time of 10.5 days and gave normal adults+ 2 days later. The total time of development thus averaged 12.5 days, as compared to 9 days on a rich nutrient me- dium of whole dead yeast and liver extract. When other salts, or sugars, were substituted for part of the NaCl and CaCl, combination, growth was somewhat less than in the medium with 0.045 M NaCl and 0.005 M CaCl,. In another experiment a comparison was made of growth in the same basic medium in the absence of added salts and in the presence of con- centrations of CaCl, ranging from 0.05 to 0.005 M. Growth in the presence of CaCl, was much more rapid than in its absence and was as good as in the presence of both NaCl and CaCl,. These results show that CaCl,, known to be in true solution in water, is an additional factor conditioning the growth of A. egypti larve. 1 These adults were, however, smaller than those obtained in richer nutrient media. SOLULE UTILIZATION BY MOSQUITO LARV As 347 GrowTH IN Nutrient Mepra ContTAINING ONLY SOLUTES In previous work (19350) it was found that A. egypti larve grew quite normally in media consisting of yeast autolysate and liver extract, or yeast extract, casein digest and liver extract. Such media, sterilized by autoclaving, contained few particles but undoubtedly contained much material in colloidal suspension. From these results it is apparent that the insects can develop to the adult stage on a diet supplied either wholly in the form of colloids or partly as colloids and partly in true solution. Several attempts were then made to rear the mosquito larve in media containing nothing but solutes. It had been found that fair growth could be obtained in autoclaved 2.5 per cent sclutions of Vegex, a com- mercial yeast extract. These solutions after autoclaving contained some sediment. Three collodion bags were prepared each containing a fresh 5 per cent solution of Vegex and surrounded by an equal volume of distilled water. Both inner and outer liquids were covered with toluene and the dialysis was allowed to proceed 2 days at room temperature. The inner and outer liquids were then separated from the toluene, tubed, and autoclaved. No visible sediment formed in the dialysate after auto- claving. Each of two tubes of each solution was inoculated with three larve. The results follow. Preparation 1. Outer liquid. One reached 3rd in 13 days, 4th in 22 days, none pupated. Imimen hentide nee = ceacheday Sade in) a) days,» 2 reached 4th in 24 days, none pupated. Preparation 2. Outer liquid. Two reached 3rd in 8 days, none reached 4th. Inner liquid. Five reached 3rd in 10.5 days, 3 reached Ath in 23 days, 1 pupated in 35 days. Preparation 3. Outer liquid. One reached 3rd in 10 days, 4th in 32 days, none pupated. Inner liquid. Five reached 3rd in 9 days, 4 reached Ath in 29.5 days, none pupated. Untreated 2.5 per cent Vegex. Four reached 3rd in 6 days, 3 reached Ath in 12.5 days, 3 pupated in 19.5 days. It is evident that in all three dialysates there was significant growth of the larve. It should be remarked that all the larvee in these solutions reached the second instar and that the growth of some to the third and 348 WILLIAM TRAGER fourth instar did not take place as a result of cannibalism. In all cases there was more growth in the inner liquid than in the outer or dialysate. But the growth in the inner liquid was not as good as in untreated 2.5 per cent Vegex. Probably considerable amounts of nutrient material were adsorbed by the collodion membranes, which acquired a- brown color. A solution was prepared containing 1 per cent liver extract 343, 1 per cent Fairchild’s peptone, and 2 per cent Difco yeast extract. Part of this was sterilized by filtration through a Berkefeld N filter. An- other part was filtered under pressure through a disc-shaped collodion membrane held in a Seitz filter. The ultrafiltrate so obtained was sterilized by Berkefeld filtration. Each of the two liquids was tested alone and with casein, 6 larve being used for each test. The results were as follows. Not ultrafiltered. Six reached 3rd in 6.5 days, 2 reached 4th in 13 days, none pupated. Same with casein. Six reached 3rd in 4.5 days, 1 reached 4th in 12 days, none pupated. Ultrafiltrate. Three reached 3rd in 8.5 days, 2 reached 4th in 25 days, none pupated. Same with casein. Two reached 3rd in 9 days, 1 reached 4th in 13 days, none pupated. Even in the solution filtered only through a Berkefeld and in the presence of casein, growth was not as good as in similar solutions sterilized by autoclaving. The Berkefeld candle probably adsorbed part of the required nutrients. Growth was not markedly affected by ultra- filtration or by the presence of solid protein in the form of casein. In another similar experiment a solution was prepared containing a fuller’s earth extract (as a source of salts), 0.5 per cent liver extract 343, 1.2 per cent Vegex, and 2 per cent Fairchild’s peptone. This was filtered under pressure through a collodion disc and was sterilized by Berkefeld filtration. Of 6 larvze inoculated into this medium, 5 reached the third instar in 7.5 days and 2 reached the fourth in 14 days, but none pupated. The fourth instar larve in this, as in the preceding experiments, never attained the size of normal full-grown individuals. DISCUSSION The effect of calcium chloride on larval growth is, quite apart from | its bearing on the utilization of solutes, of sufficient interest to merit some discussion. A certain amount of this salt is necessary for the nor- mal development of Aides egypti larve. The effect is not an osmotic one. Sugars or other salts in equivalent concentrations do not show it. SOLUTE UTILIZATION BY MOSQUITO LARV/Z 349 Previous work on the effects of salts on mosquito larvee has been concerned mainly with the resistance of various species to high salt con- centrations. Wigglesworth (1933c) found that 4. egypti larve were killed by 1.1 per cent sodium chloride but, by very gradually evaporating the solution, could be made resistant ‘to this concentration and to arti- ficial sea water of a concentration osmotically equivalent to 1.75 per cent sodium chloride. He called attention to the fact that the larve can resist higher osmotic pressures in balanced salt solutions than in solu- tions of a single salt, and concluded that death of the larve in both cases was due, not to osmotic effects, but to the presence of directly toxic concentrations of certain ions. The possible stimulating effect of low concentrations of salts has received little attention. Roubaud, Colas- Belcour, and Treillard (1935), however, noted that while higher salt concentrations (in the form of sea water) were toxic to larve of three races of Anopheles maculipennis, lower concentrations, ranging from one-fifth to one-twentieth the strength of sea water, were distinctly favorable and resulted in better growth than that obtained in distilled water. They state that this effect may be due to a stimulation by the salts on bacterial growth or to the physico-chemical state of the water. In the light of my own results, it seems probable that the favorable ef- fect of low sea water concentrations observed by these workers was the result of the specific action of essential ions. It is, of course, well known that electrolytes have remarkable effects on a wide variety of animals and plants. One need only mention the pioneer work of Ringer (1880-82), Loeb (1910), and Osterhout (1922). Calcium ions seem to be of especial importance in many bio- logical phenomena (see Bayliss, 1915) such as the coagulation of milk and blood, the beat of the heart, the transmission of nerve impulses, etc. Recently Warén (1933) has shown that calcium is essential for cell division in the plant Micrasterias. It is, therefore, not surprising that calcium should be a growth requirement of mosquito larve. Whether the calcium is essential for its own sake, or because of its antagonism toward some other. ion, cannot be said at present. The effects on larval growth of calcium chloride, known to exist in water as a solute, and of factor A, shown to be also in true solution, demonstrate clearly the ability of the larve to utilize substances in solu- tion. Moreover, in view of the results described with ultrafiltrates, there can be no doubt that Aides egypti larve can grow to the last larval instar at the expense of nutrients existing wholly in true solution. It seems probable that the subnormal growth in such media containing only solutes is due, not to the inability of the larve to utilize nutrients in such a state, but to the experimental difficulty of providing, in the state of 350 WILLIAM TRAGER true solution, sufficiently high concentrations of some of the required substances. On the other hand, when solid food is available and the alimentary tract becomes packed with it, very high concentrations of soluble nutrients must be produced as a result of digestion. Such an explanation may very well apply to aquatic organisms other than mos- quito larve. For example, the finding of Gellis and Clarke (1935) that Daphnia magna could grow in media containing colloids but not in media containing only solutes may be explained in this manner. The difference between nutrients existing in the environment as particles, colloids, or in true solution becomes a quantitative one. Whatever. the initial state of the nutrient, it must be changed to a soluble form before it can be ab- ‘sorbed. When the nutrient exists as particles or even in colloidal form, the mosquito larve, or other aquatic organisms, are able to concentrate it by gathering up the particles. This they are not able to do with nutrients existing as solutes. These must enter the alimentary tract along with water and are present therein in the same concentration as in the sur- rounding fluid. . These considerations bring us to the problem of the method of in- gestion of food by mosquito larve. ‘The larval mouth brushes have been generally considered as an apparatus for gathering up fine particles. Shipitsina (1930) regarded them as filters capable of filtering out from the water even colloidal particles. She found that anopheline larvee could filter out particles of colloidal silver with a diameter of 20 my but could not filter out the particles of soluble starch with a diameter of 5 mu. Wigglesworth (19330), using larve of Aides egypti, also con- cluded that very little fluid was ingested by mouth. He observed that the larva accumulates bits of food in its buccal cavity and does not swal- low the food until a considerable bolus has accumulated. My own ob- servations indicate that in particle-free media the brushes are probably of little use, but the swallowing movements of the larva enable it to ingest fluid. Since larve can reach the fourth instar at the expense of nutrients in solution, these nutrients must somehow be absorbed by the larva. Howland (1930) has shown that dyes in solution in the external liquid can enter the alimentary tract of mosquito larve and be absorbed there. It seems most likely that soluble nutrients would follow the same path. However, Wigglesworth (1933b) found that, while the cuticle covering the body of mosquito larve is impermeable to water, water can and does regularly enter the body by way of the anal gills. The possibility of the entrance of nutrient solutes in this manner must be borne in mind, al- though Wigglesworth (1933a) has also shown that, while certain salts SOLUTE UTILIZATION BY MOSQUITO LARVA 351 can enter the anal gill cells from the external medium, they cannot pass from the cells into the hemolymph. SUMMARY The larvee of the yellow fever mosquito, ZEdes egypti, are able to utilize substances in true solution. They require for development at a normal rate a proper concentration of calcium chloride, known to exist in water as a solute. Their growth is likewise conditioned by an organic growth factor shown to be also a solute. Finally, the larvee can develop at least as far as the fourth instar in media the nutrients of which exist entirely in true solution. It gives me pleasure to acknowledge the helpful interest of Dr. R. W. Glaser. REFERENCES Bavtiss, W. M., 1915. Principles of General Physiology. Longmans, Green and Co. London. Geis, S. S., aNp G. L. CrarKe, 1935. Organic matter in dissolved and in col- loidal form as food for Daphnia magna. Physiol. Zo6l., 8: 127. Hinman, E. H., 1930. A study of the food of mosquito larvae (Culicide). Am. Jour. Hyg., 12: 238. Hinman, E. H., 1932. The role of solutes and colloids in the nutrition of anophe- line larve. Am. Jour. Trop. Med., 12: 263. Howtanp, L. J., 1930. The nutrition of mosquito larve, with special reference to their algal food. Bull. Ent. Res., 21: 431. Loes, J., 1910. Ueber physiologische Ionenwirkungen, insbesondere die Bedeutung der Na-, Ca und K-Ionen. Oppenheimer’s Handbuch der Biochemie des Menschen und der Tiere, Jena, 2: 104. MacGrecor, M. E., 1929. The significance of the pH in the development of mos- quito larve. Parasitol., 21: 132. MaruHeEson, R., anp E. H. HrnmAn, 1930. A seasonal study of the plancton ofa spring-fed Chara pool versus that of a temporary to semi-permanent wood- land pool in relation to mosquito breeding. Am. Jour. Hyg., 11: 174. OstERHOUT, W. J. V., 1922. Injury, Recovery and Death, in Relation to Conduc- tivity and Permeability. J. B. Lippincott Co. Philadelphia. Pirrer, A., 1907. Die Ernahrung der Wassertiere. Zeitschr. f. allg. Physiol., 7: 23S. \ Rrincer, SypNney, 1880-82. Concerning the influence exerted by each of the con- stituents of the blood on the contraction of the ventricle. Jour. Physiol., 3: 380. Rousavup, E., J. Coras-BeLcour, AND M. TreiLiarp, 1935. Influence de la con- centration en sel marin sur le développement larvaire d’Anopheles maculi- pennis (var. atroparvus, fallax et labranchie). Bull. Soc. Path. Exot., 28: 568. Suirrrstna, G. K., 1930. The réle of organic colloids in the nutrition of larve of Anopheles maculipennis. (In Russian.) Bull. de l’Inst. rech. biol. de Perm, 7: 171. Tracer, W., 1935a. The culture of mosquito larve free from living micro- organisms. Am. Jour. Hyg., 22: 18. 352 WILLIAM TRAGER Tracer, W., 1935b. On the nutritional requirements of mosquito larve (/%des egypti). Am. Jour. Hyg., 22: 475. Tracer, W., 1936. A growth factor required by mosquito larve. In press. Waren, H., 1933. Uber die Rolle des Calciums im Leben von Zelle auf Grund von Versuchen an Micrasterias. Zeitschr. wiss. Biol. Abt. E, Planta, 19: 1. Wuirte, G. F., 1931. Production of sterile maggots for surgical use. Jour. Para- sitol., 18: 133. WiIccLESworTtH, V. B., 1933a. The effect of salts on the anal gills of the mosquito larva. Jour. Exper. Biol., 10: 1. WIGGLESworTH, V. B., 1933b. The function of the anal gills of the mosquito larva. Jour. Exper. Biol., 10: 16. WiIGGLESwortH, V. B., 1933c. The adaptation of mosquito larve to salt water. Jour. Exper. Biol., 10: 27. ENVIRONMENT AND SEX IN DEE OVIPAROUS OYSTER OSTREAS VIRGINICA Wek COR, (From the Osborn Zodélogical Laboratory, Yale University) The Virginia oyster is a member of that group of oviparous pelecy- pods in which each individual, with relatively few exceptions, is sea- sonally functional as either male or female, with the possibility of re- versal to the opposite sex during the interval between two breeding seasons (Needler, 1932; Coe, 1932a, 1934; Burkenroad, 1936). All species of larviparous oysters so far. as known have a more or less regu- larly alternating rhythm of male and female phases. In the northern part of its range the Virginia oyster appears to be mainly or wholly protandric, while in somewhat warmer waters, at least as far southward as Virginia, a considerable proportion of the young oysters function as females during their first breeding season, which occurs at the age of one year. The more favorable the conditions are for the rapid growth of the young oyster, the more likely is the ini- tial male phase to be omitted or aborted. Hence those localities and seasons in which the yearling oysters reach the largest mean size usually have the largest proportion of females in the first breeding season (Coe, 1934). From North Carolina southward young oysters of both sexes reach functional maturity when only ten to twelve weeks of age. In the older age groups the two sexes occur in approximately equal proportions, with some tendency toward an excess of males at the sec- ond breeding season and an excess of females in groups that are more than five years of age. It is unknown whether every individual changes its sex during even a long lifetime nor is it known whether more than a single change ever occurs. The histological basis for these changes of sex depends upon the essential bisexuality of the gonads. In the young oyster before sexual maturity the gonad consists of a cortical layer of ovogonia surrounding a group of spermatogenic cells. In the typical case of protandry the spermatogonia proliferate rapidly, soon giving the gonad its dominantly male characteristics. Some ovogonia or ovocytes can nearly always be 353 354 Wioedie (COs, detected in the cortex, however. These, together with many sexually undifferentiated cells, remain as residual gonia after the functional male phase has been completed. The sexual phase for the ensuing year will then depend upon which of the two alternative types of cells are stimu- lated to renewed activity. In the sexually immature young oyster which is to function as a female at its first breeding season, on the other. hand, the proliferation of the spermatogonia in the primary bisexual gonad is soon checked and is succeeded by a differentiation and multiplication of ovogonia. The initial male phase is thus aborted, sometimes even after the appearance of a few spermatids. In approximately one per cent of the yearling population neither type of sexuality is fully dominant, so that partial or complete hermaphro- ditism results. Self-fertilization is successful experimentally and pre- sumably occurs occasionally in nature. Seasonal protandry. Occasionally the gonad of the yearling her- maphrodite shows a typical ovarian structure except for groups of spermatogenic cells or ripe spermatozoa next the lumens of some of the follicles. In some of these cases the sperm mature and are discharged early in the breeding season, leaving the ovocytes and a few residual spermatozoa to be discharged later during the same summer, as is the case with the young of viviparous species (Coe, 1932b). During the past few years evidence has been obtained which gives some indication as to the influence of environmental conditions in deter- mining which of the two alternative types of cells in the primary bi- sexual gonad shall be stimulated to proliferation; that is to say, which of the alternative sexual phases shall be functional at the breeding season. It has been previously shown (Coe, 1932a, 1934) that the mean size of the females in the yearling group from each locality considerably exceeds that of the males. In addition to the data previously presented, further evidence was obtained from the examination of the gonads of more than 3,000 individuals from the vicinity of New Haven, Connecti- cut, during their first breeding season, in 1935 and 1936. Of 670 young oysters which had been collected in the harbor and transferred to deep water off the New Haven Breakwater there was a ratio of 7.66 females to each hundred males. The mean size of the fe- males was 26.88 mm. as compared with 25.43 mm. for the males. Simi- lar collections of 391 yearlings made at the same season from the natural beds at Milford, only about eight miles distant, gave a ratio of 24 females for each hundred males, the females at this locality having a mean length ENVIRONMENT AND SEX IN THE OYSTER 355 of 30.52 mm. and the males 30.24 mm. Tables I and II indicate the sex ratios which were found at several localities along the coast in the breed- ing seasons of 1932-1935, the entries listing immature individuals repre- senting samples taken early in the season. ‘The size of the shell at date of collection will obviously depend upon the length and favorableness of the two growing seasons in the young oyster’s lifetime, but the size at the end of winter, at which time the sexual condition for the following season is presumably already determined, likewise shows an average larger size for females than for males. TABLE [| Ostrea virginica. Sex at first breeding season. Jm = sexually immature early in breeding season; F = female; H = hermaphroditic; M = male. Locality No. Im. | No. M. | No. H. | No. F. Total Perce ntaee New Haven (1932)......... 17 389 4 13 423 3.07 (Gireme Souda Ie Was. oe es 21 197 0 7 225 Salt New Haven (1933)......... 1 129 0 7 137 5.11 INews Haven’ (1935). 5-2...) 105 521 4 40 670 5.97 MulrondaGi935)) hy a oe ae 2 312 2 75 391 18.89 W. Sayville, L. I., ae 0 154 4 48 206 23.30 Beaufort, N. C.. ae ee 0 43 0 21 64 32.81 TABLE II Ostrea virginica. Correlation of sex ratio and size. Mean length at first breeding season. /# = female; M = male. Locality No. M. Mean No. F. Mean Bos Exo mm. mm. New Haven (1932)......... 389 31.28 13 38.54 3.34 Great South Bay, L.I....... 197 Ca. 29. 7 Ca. 34. 3.55 New Haven (1933)......... 129 39.90 7 44.14 5.42 New Haven (1935)......... 521 25.43 40 26.88 7.66 Milford (1935)............. 312 30.24 75 30.52 24.04 W. Sayville, L. I. ee 154 46.33 48 59.33 31.17 Beaufort, N. C.. ss 43 53.5 21 60.5 48.84 These results are in complete harmony with the evidence reported previously for other years and for other localities (Coe, 1932a, 1934). The data given in Tables I and II indicate that environmental conditions in the same locality may vary considerably from year to year, if it be 356 Woke COE assumed that samples of several hundred individuals are truly repre- sentative of the entire population. Date of setting and length of growing season. Since the sex ratio of the yearlings is evidently correlated with the size at the end of winter as well as with the size at the first breeding season, it seems probable . that the length as well as the favorableness of the growing season pre- ceding the first winter may influence the proportion of females. Years in which much of the set occurs in early August presumably show a higher ratio of females among the yearlings of the following year than is the case when most of the set takes place in September, providing the environmental conditions are similar. The time of sexual maturity of the yearlings also shows much varia- tion in different years and at various localities (Table I). In the north- ern portion of the species’ range a large proportion of the young remain sexually immature until their second year (Needler, 1932) but south of Cape Cod most of them become functional at some time during the sum- mer at the end of their first year. It was previously reported (Coe, 1932a, 1934) that in 1932 most of the yearlings in the Woods Hole area postponed spawning until their second summer, but in more recent years nearly all have become functional before the end of the first breeding season. Change of sex ratio during breeding season. ‘The progressively in- creasing ratio of mature females in the yearling population as the season advances is shown by recent collections from Milford, Connecticut, ob- tained through the courtesy of Dr. Victor L. Loosanoff of the Bureau of Fisheries station at Milford. At the beginning of the breeding season, July 3, 1936, a sample containing 911 yearlings was composed of 77 sexually immature individuals, 780 ripe males, 10 hermaphrodites and 44 females. A second sample, taken August 13, at the height of the breeding season, consisted of 7 individuals that were still sexually im- mature, 410 males, 4 hermaphrodites and 58 females. A third sample near the end of the breeding season contained 92 males and 21 females, besides 176 individuals which had almost completely spawned out. For each hundred males there were 5.64 functional females in the first sam- ple, 14.15 in the second and 22.83 in the third sample.. The increased ratio of females resulted in part from the maturity of females which had not become functional at the beginning of the season and in part from individuals which would have been classed as hermaphroditic males earlier but have later transformed to females by the maturation of the spermatogenic cells and discharge of the resulting spermatozoa. ENVIRONMENT AND SEX IN THE OYSTER Shy The breeding season occurs earlier on the oyster beds in shallow, warmer water but varies with the temperature from year to year. From one bed in Milford in July, 1935, the female ratio in a sample of 312 _ yearlings was 24.04, while from another bed in the same month of 1934 it was only 7.50. On July 28, 1936, 373 yearlings from New Haven harbor showed a female ratio of 12.54, as compared with 5.42 in 1933 and 7.66 in 1935. The distinctly larger mean size of the females during the first breed- ing season is well illustrated by random samples consisting of 202 indi- viduals taken from the culture floats at West Sayville. Of those having a shell length of more than 50 mm. there were 49 males and 42 females, while the group with a length of less than 50 mm. contained 105 males and only 6 females. If the selection had been made either by volume or by weight the superiority of the females would have been even more evident. Tables I and II offer substantial evidence that there is a rather close correlation between the mean size of the individuals in the yearling population and the percentage of functional females, or the ratio of females to males. The mean size of the females in all cases exceeds that of the males and with few exceptions those years and those locali- ties in which the metabolic conditions are most favorable, as indicated by the size attained at the end of the first year, have the largest propor- tions of females. With this additional evidence it now seems fair to conclude that favorable metabolic conditions increase not only the mean size of the individuals but also the proportion of females in the yearling population. Three assumptions may be suggested in explanation of this conclu- sion: (a) that inherent sex factors are associated with metabolic factors that lead to the more rapid growth of those individuals that are to become females; (b) that the more favorable conditions for growth stimulate the proliferation and growth of the ovogonia in the young bisexual gonad or inhibit the proliferation of the spermatogonia; (c) that there is in the immature oviparous oysters an inherent tendency toward an alternation of sexuality such as is usual in the viviparous species (Orton, 1926-27; Coe, 19320). The first two assumptions have been discussed in a previous paper (Coe, 1934). The last assumption would imply that those individuals that are to function as females in their first breeding season, when one year of age, may have already passed through an aborted protandric phase in the late autumn of the year of hatching, when only about four months old. There is some evidence that this assumption is justified, 358 W. R. COE for it has been observed that the proliferation of spermatogonia and the incipient stages of spermatogenesis are most pronounced in the largest of the young individuals. It seems quite possible that these are the same individuals as those which function as females the following summer. The observed ratios in the various localities are in harmony with this supposition. Protandry would thus dominate the sexuality of the oviparous species just as is the case in the viviparous forms (Coe, 1932b, 1934), the initial male phase being inaugurated, but aborted, in the autumn of the year of hatching for those individuals that function as females at the age of one year. If for any reason the protandric phase is delayed until the following spring, such individuals complete spermatogenesis and thus become functional males during the summer of the first breed- ing ‘season. In those localities from ‘Virginia northward which have the most favorable conditions for the inauguration of the aborted male phase in the autumn there will thus be the greatest proportion of fe- males at the first breeding season. It may be mentioned in conclusion that the primary gonads of both the oviparous and viviparous oysters show various grades of bisexuality, some individuals being more strongly masculine or more strongly fem- inine than others in appearance, as indicated by the relative abundance of the cells characteristic of the two sexes. This is likewise true of other pelecypods which normally experience one or more changes of sexual phase, as Teredo (Coe, 1935) and Venus (Loosanoff, 1936), as well as of certain gastropods, including Patella (Orton, 1928) and Crepidula (Coe, 1936). In all of these species the initial male phase is occasionally aborted, while other young individuals appear to be so strongly masculine that they have been designated as “true males.” In the latter the female phase may be long delayed. Protandry is thus dominant in all the species mentioned but is not exclusive in any of them. LITERATURE, Chirp BuRKENROAD, Martin D., 1936. Rate of reversal of sexual phase in oviparous oysters. (In press.) : Coz, Westey R., 1932a. Sexual phases in the American oyster (Ostrea virginica). Biol. Bull., 63: 419. Cor, WEsLEY R., 1932b. Development of the gonads and the sequence of the sexual phases in the California oyster (Ostrea lurida). Bull. Scripps Inst. Oceanog., Tech. Ser., 3: 119. Cor, WEsLEY R., 1934. Alternation of sexuality in oysters. Am. Nat., 68: 236. Cor, WESLEY R., 1935. Sequence of sexual phases in Teredo, Ostrea and Crepidula. Anat. Rec., 64: suppl., 81. ENVIRONMENT AND SEX IN THE OYSTER 359 Cor, WEsLEY R., 1936. Sexual phases in Crepidula. Jour. Exper. Zool., 72: 455. LoosANorrF, Victor, 1936. Sexual phases in the quohog. Science, 83: 287. NEEDLER, ALFREDA B., 1932. American Atlantic oysters change their sex. Prog. Rept. Atlantic Biol. Sta. and Fish. Exp. Sta., 5: 3. Orton, J. H., 1926-27. Observations and experiments on sex-change in the Euro- pean oyster (O. edulis). Jour. Mar. Biol. Ass., 14: 967. Orton, J. H., 1928. Observations on Patella vulgata. Part I. Sex-phenomena, breeding and shell-growth. Jour. Mar. Biol. Ass., 15: 851. RENAL FUNCTION IN MARINE TELEOSTS ITV. THe Excretion oF INoRGANIC PHOSPHATE IN THE SCULPIN ALLAN L. GRAFELIN 1 (From the Department of Anatomy, Harvard Medical School) Our knowledge of the nature of the water and salt cycle in the marine teleosts has been markedly clarified by Smith (1930, 1932), who has demonstrated quite conclusively (1) that marine teleosts normally ingest considerable quantities of sea water, which subsequently undergo ab- sorption in the gastro-intestinal tract; (2) that the greater part of the sodium, potassium, and chloride absorbed ftom the alimentary tract is excreted by some extra-renal route, presumably the gills; and (3) that essentially all of the absorbed magnesium, calcium, sulfate, and phos- phate which leaves the body is excreted by the kidneys alone. The renal excretion of inorganic phosphate in marine teleosts is particularly interesting. It has been shown by Marshall and Grafflin (1933) that the urine of the aglomerular goosefish (Lophius piscatorius ) may contain large amounts of inorganic phosphate, which is presumably formed in the kidney from some as yet unknown precursor and secreted by the renal tubules. The basis for this interpretation is that injected inorganic phosphate is not excreted either by this fish or by the aglomeru- lar toadfish (Opsanus tau). It is to be inferred from this fact that the renal tubules of the glomerular teleost, and perhaps of other animals, would be unable to secrete preformed inorganic phosphate, and that this substance would be excreted only by glomerular filtration. Observations upon frogs (Marshall and Grafflin, 1933) and dogs (Pitts, 1933) sup- port this suggestion. ‘There remains, nevertheless, the possibility that the tubules of the glomerular kidney might, like the tubules of the aglomerular kidney, be able to secrete inorganic phosphate from some occult precursor. Though no evidence in this direction has as yet been adduced for the frog or for mammals, the evidence presented in this paper clearly points to such an operation in the glomerular fish kidney. Our present experiments were made upon the common sculpin (My- 17 am indebted to Professor E. K. Marshall, Jr., for his constant advice and criticism during the present investigation, which was carried out at the Mt. Desert Island Biological Laboratory, Salsbury Cove, Maine. I am also indebted to Pro- fessor H. W. Smith for critical comments and for assistance with the manuscript. 360 EXCRETION {NORGANIC PHOSPHATE IN SCULPIN 361 oxocephalus octodecitmspinosus).* The study of phosphate excretion in this species has been found to be complicated by two factors: (1) the occurrence of a spontaneous increase in phosphate excretion under ex- perimental conditions; (2) the frequent presence of a precipitate of magnesium phosphate in the urine. Discussion of the experiments upon the mechanism of phosphate excretion, in terms of glomerular and tubu- lar activity, will be deferred until after the consideration of these two complicating factors. METHODS To determine the rate of urine flow, the bladder was first emptied, with the use of a fine glass catheter, by massage of the overlying abdominal wall. The urinary papilla was then tied off with fine silk thread; and subsequently, at the end of the collection period, the thread was cut and the urine removed by catheter. In this way repeated urine collections could be made upon the same animal. The re- gion of the papilla was always thoroughly cleansed with distilled water and dried before catheterization to avoid contamination. In sacrifice experiments blood was obtained by syringe from the exposed heart (bulbus of aorta) ; in the course of an experiment it was obtained by syringe from the caudal vessels without exposure (see Grafflin, 1935a). Phosphate was determined by the method of Fiske and Subbarow (1925), and is uniformly expressed in terms of millimols per liter. Potassium oxalate was used as anticoagulant. Whenever the urine showed tur- bidity the precipitate was dissolved with a trace of glacial acetic acid before the phosphate content was determined. The determination of xylose was carried out upon Somogyi (1931) filtrates of plasma and urine. After absorption of the glu- cose on yeast, xylose was determined by the Folin (1929) modification of the Folin-Wu method. Xylose values are expressed as xylose, not as glucose equiva- lents, appropriate corrections being made for the diminished reducing power of xylose and its dilution by the water content of packed yeast (see Marshall and Grafflin, 1932). OBSERVATIONS Normal Urinary and Plasma Phosphate Concentrations in the Sculpin Under normal conditions phosphate may appear in the urine of the sculpin in very high concentration, and it is not infrequent to observe a heavy urinary precipitate of magnesium acid phosphate (Pitts, 1934). This phosphate is entirely of metabolic origin, since phosphate is present in sea water in only slight traces. When oliguria is established by obstruction of the gastro-intestinal tract at the pyloric end of the stomach, phosphate continues to be excreted in high concentration (Grafflin and Ennis, 1934). (The oliguria is a consequence of the fact that the animal is dependent upon ingested sea water for its urinary 2 The nephron in this species consists of a glomerulus, neck segment, brush border segment (the homologue of the proximal convoluted tubule of higher forms) and initial collecting tubule; the brush border segment is bisegmental on cytological grounds (Defrise, 1932; Edwards, 1935). 362 ALLAN L. GRAFFLIN water.) In the urine obtained from six sculpins immediately after catching on hook and line, Pitts (1934) observed phosphate values ranging from 50 to 155 mM./liter. In six additional freshly caught specimens we have observed values of 7, 40, 113, 130, 140, and 190 mM./liter. Plasma phosphate determinations on freshly caught sculpins gave values of 3.2, 3.4, 3.5, 3.8, 4.4, 4.5, and 5.6 mM./liter. It follows from these observations that phosphate may be highly concentrated, rela- tively to the plasma, in the urine of normal fish. Course of Urinary Phosphate Concentration in Sculpins with a Single Daily Catheterization It was early observed, however, that when individual animals are followed from day to day, with a single catheterization and without tying of the urinary papilla, the urinary phosphate soon drops to low levels and remains there. For example, sculpin No. 14 was followed for twelve days: urinary PO, 7.0, 20.6, 4.0, 0, 0, 0, 1.5, 0.3, 1.1, 1.7, 1.1, and 1.9mM./liter. No. 20 was followed for six days: PO, 149.0, 10.8, 3.6, 2.5, 1.4, and 0.3 mM./liter. The urine flow on the seventh day was 51 ce. per kilogram per 24 hours, PO, 0.5 mM./liter. This fall in the urinary concentration of phosphate is in part due to diuresis, for it is known that, unless very special precautions are taken, sculpins readily become diuretic under experimental conditions (Grafflin, 1931, 19350; Smith, 1932; Pitts, 1934; Grafflin and Ennis, 1934; Clarke, 1934). Spontaneous Increase in Urinary Phosphate Concentration and Phosphate Excretion However, the present investigation has demonstrated a further very significant fact which has been missed in previous studies; namely, that when urine samples are taken at short intervals from a sculpin previously undisturbed (for exceptions see below), the urinary phosphate concen- tration and the rate of phosphate excretion very often show a moderate or marked, but transitory, increase.*. Data upon thirteen sculpins are given in Fig. 1. In eleven animals preliminary catheterization (open circles) was followed by a collection period (2 to 3.2 hours), with the urinary papilla tied. Since the phosphate values represent the average concentration for the period, the second points (solid circles) are plotted as of the mid-period. The other two animals (Nos. 70, 72) were 3 A similar rise was observed in the closely related daddy sculpin (M. scor- pius), from 0.5 to 44 mM./liter in a short collection period, with the urinary papilla tied. It is to be noted that this species shows marked evidences of glomerular de- generation, some specimens being functionally entirely aglomerular (Grafflin, 1933). EXCRETION INORGANIC PHOSPHATE IN SCULPIN 363 catheterized twice without tying of the papilla, and the second points are plotted at the time of catheterization. A further group of eleven sculpins which were more extensively studied are recorded in Fig. 2. All of the animals, except No. 37, were studied by catheterization alone, and the actual times of catheterization are plotted (open circles). In the case of No. 37, three collection periods were run successively, and the mid- 140 points of these periods are plotted. All sculpins record- ed in Fig. 2 were followed on the second day, and five of them still further on the third day (Fig. 3). The conditions under which the spontaneous in- crease in phosphate concen- tration was observed in these animals are varied. 160 ra i) ©) | 53 eS) © \ Fig. 1—In 11 specimens the rise occurred when the fish were first handled at varying intervals after catching: Nos. 70, 72—day after catching; No. 28—second day; No. 30-—fourth day; No. 65 —sixth day; Nos. 38, 41, 53, 54, 55, 93—interval indeterminate. (These latter fish were taken at Y random from the stock supply in yy a large live car. This stock was 6 added to as fish were procured, 0) 0.2 1 ip) 2 and records of individual speci- Tours mens were not kept.) No. 12 ine a was catheterized once on the Urinary POs - mMols/ liter oO fo) 20 ip iN je) w 26 wy Oo \N f, IS Spontaneous increase in urinary : phosphate concentration in the sculpin. Open eleventh and eighteenth days circles, catheterization alone. Solid circles, and the rise was observed on the —_yrine collection period with urinary papilla twenty-third day. No. 26 was tied. catheterized once on the third, fourth, and fifth days, and the rise was observed on the sixth day. Fig. 2—In 8 specimens the rise occurred when the fish were first handled at varying intervals after catching: Nos. 47, 48, 49—day after catching; Nos. 43, 44, 45, 46—third day; No. 37—interval indeterminate (stock supply). Nos. 34 and 35 were catheterized once on the third day and the rise was observed on the fifth day. No. 36 was catheterized once on the third and fifth days and the rise was observed on the eighth day. 364 ALLAN L. GRAFFLIN Figs. 1 to 3 deal only with the increase in urinary phosphate con- centration, and a more detailed analysis of the experiments is given in Table I. From the nature of the data given, this table can only include experimental periods in which the urine flow was accurately determined, with the urinary papilla tied. Necessary explanatory considerations for the interpretation of this table are given at the top of p. 365. 140 120 100 Urinary PO4- rn. Mols/liter Fic. 2. Spontaneous increase in urinary phosphate concentration in the sculpin. Open circles, catheterization alone. Solid circles, urine collection period with uri- nary papilla tied. Groups of animals handled identically are graphed similarly (e.g., Nos. 43, 44, 45, 46; Nos. 47, 48, 49). If, now, we proceed to analyze the data which we have obtained, we are led to the following considerations. (1) Sculpins very often show a spontaneous increase in urinary phosphate concentration which may be slight, moderate or marked. It must be emphasized that in some specimens it is not observed at all. The degree of consistency with which it occurs is demonstrated by con- sidering groups of sculpins with an identical history and handled iden- tically under experimental conditions (see groups of animals similarly rapid amma Z)): EXCRETION INORGANIC PHOSPHATE IN SCULPIN 365 (2) In nineteen of the twenty-four sculpins recorded in Figs. 1 and 2 the spontaneous increase was observed when the animals were first handled, irrespective of the number of days they had been in cap- tivity. The other five animals (Nos. 12, 26, 34, 35, 36) had been catheterized from one to three times previously (see above). (3) The increase may occur irrespective of the initial concentration (first catheterization) over a wide range (1 to 81 mM./liter—Figs. 1 and 2). (4) The increase is transitory, in general rising to a peak and then rather rapidly diminishing to low levels. (5) It seems quite certain from the data available that the increase in urinary phosphate concentration is not associated with an increase in plasma phosphate concentration. Thus, in No. 15 (Table III), in which the urinary concentration showed a spontaneous increase in the first period and just as marked a decrease in the second period, the plasma concentration was considerably higher in the second period (4.2) than in the first (3.1). The rate of phosphate excretion was identical in the two periods. There seems to be no direct relationship between (e) o> ine Ss | Urinary PO,= m.Mols/ liter | bi Si Fic. 3. Subsequent course of urinary phosphate concentration in sculpins re- corded in Fig. 2. Open and solid circles, as in Fig. 2. .the absolute level of the plasma concentration and the concentration in the urine (see Table III). Three freshly caught sculpins examined at once gave the following plasma and urinary (in parentheses) phos- phate values: 3.2 (190), 3.8 (130) and 4.4 (113) mM./liter. (6) As discussed above, the computed values for “ Increase in PO, excretion’ (Table I) are almost certainly minimal in most cases. A comparison of the figures for “ Increase in PO, excretion” with those for “ Increase in PO, concentration’ shows wide discrepancies in the different animals. Obviously, an increase in concentration of 20 mM./ liter, with a urine flow of 35 cc., would have exactly the same signif- 366 ALLAN L. GRAFFLIN dbmsien) Il Data upon phosphate excretion during urine collection periods in sculpins reported in Figs. 1, 2 and 3. Selon Increase in Additional Increase in No. Urine flow ae 4 PO, POs excretion® , concentration excreted” excretion? cc. per kg. . mM. per kg. mM. per kg. per 24 hours mM. /liter _ mM. per 24 hours per 24 hours 1. Sculpins reported in Figs. 1 and 2 (No. 37 only) 12 7.0 26 0.005 0.18 0.35 26 S25. 27 0.014 0.88 1.08 28 15.4 56 0.015 0.86 0.91 30 16.7 49 0.022 0.82 0.92 37 10.0 74 0.033 0.74 0.78 38 fo 79 0.043 1.36 1.39 41 26.6 61 0.024 1.62 1.84 53 8.3 64 0.020 0.53 1.20 54 6.8 108 0.027 0.73 1.05 38) 9.4 61 0.015 0.57 1.07 65 10.02 119 0.024 1.19 1.20 93 LeOe 26 | 0.008 0.44 1.58 2. Sculpins on second day of handling (Fig. 3—22 to 29 hours) 34 40.7 Sil > 0.0029 >0.07 0.11 35 49.8 >2.2 >0.0037 >0.11 0.16 36-1 33.2 0.3 0.0002 0.01 0.13 =2 113.6 0.1 0.0002 0.01 0.45 37 68.4 0 0 0 <0.07 43 72.0 all 0.0085 0.25 0.55 44 53.6 0 0 0 0.16 45 29.8: 0.8 0.0006 0.02 0.13 46-1 61.8 0.6 0.0007 0.04 0.20 =2 oat SED 0.0046 0.26 0.50 47 23.2 0.8 0.0005 0.02 0.18 48 41.5 0.3 0.0005 0.01 0.10 49 59.0 1.3 0.0022 0.08 0.16 34 93.4 3.3 0.0066 0.31 0.40 35 142.8 0.3 0.0008 0.04 0.20 43 83.5 1.6 0.0032 0.13 0.30 45 59.8 1.4 0.0015 0.08 0.16 47 11.4 es) 0.0006 0.03 0.06 = @ Actual volume (in liters) of urine formed in collection period X increase in POs concentration. > Increase in PO: concentration X rate of urine flow. © Actual PO: concentration of collection period urine (Figs. 1-3) X rate of urine flow. Approximate. EXCRETION INORGANIC PHOSPHATE IN SCULPIN 367 (1) “Increase in PO, excretion” is the difference between the observed ex- cretion rate (last column) and the rate at which PO, would have been excreted if the urinary PO, concentration had not increased, but remained the same. Actually, the computed values may be considered as minimal values. The calculation is based on the assumption that the urine flow did not increase during the collection period, while in many cases, from the evidence available (Grafflin, 1931, 1935), this paper ; Pitts, 1934), it probably did. (2) “ Additional PO, excreted” represents the amount of PO, which would have to be added to the actual volume of urine formed in the collection period, in order to raise the PO, concentration from the level observed at the preliminary catheterization to that observed at the end of the collection period. In order to make clear the significance of this calculation, we must proceed at once to a dis- cussion of the white precipitate frequently noted in sculpin urine. The following quotation is taken from Grafflin and Ennis (1934—p. 287): “It has been the ex- perience of the workers at the Mount Desert Island Biological Laboratory over a period of years that the urine often shows a white precipitate; this is particularly true of the urine in freshly caught fish and of urine accumulated within the first 24 hours after catching. It has further been observed that a clear bladder urine may suddenly precipitate in the catheter or syringe when attempts are made to collect it; or again, the urine may be perfectly clear when collected, but subse- quently develop a precipitate when allowed to stand at room temperature. The precipitate can always be dissolved by acidifying the urine with acetic acid.” 4 On analysis Pitts (1934) found the composition of this precipitate to corre- spond to MgHPO,-3H.O. He makes the following statement: “It is possible that some of the variability of phosphate and magnesium excretion in these fish results from the ejection from the bladder of solid magnesium phosphate, excreted by the kidneys at some earlier time, but remaining in a precipitated state till a later col- lection period.” If this were true, it might offer the entire explanation for the observed increase in urinary phosphate excretion. The significance of the calcula- tion, “ Additional PO, excreted,’ now becomes apparent. It represents simply the absolute magnitude of the hypothetical preformed store of phosphatic precipitate in the bladder (or elsewhere) which would be necessary to account for the increase in urinary phosphate concentration noted in the experimental periods. icance, in terms of “ Increase in PO, excretion,’ as an increase in con- centration of 100 mM./liter, with a urine flow of 7 cc. Under the cir- cumstances, it is evident that “ Increase in PO, excretion ” is really the fundamental consideration in a given experiment. Calculation from the data given in Section 1, Table I, shows that this quantity varies from 28 per cent to 99 per cent of the total “ PO, excretion,” with an average of 74 per cent. (7) Since sculpins usually become progressively diuretic when han- dled repeatedly under experimental conditions, the total phosphate ex- cretion may continue to increase tremendously when the concentration is steadily diminishing. This fact is strikingly demonstrated in the 4 This precipitate may occur in urines of widely varying PO, content. In the present study, on the one hand, cloudy urines (as withdrawn from the bladder) have been observed with a PO, concentration as low as 6 mM./liter. On the other hand, perfectly clear urines have been noted which contained as high as 100 mM./liter. 368 ALLAN L. GRAFFLIN following tabulation (Table II) of the results obtained upon sculpin No. 37 (see Figs. 2 and 3). TABLE II Sculpin No. 37; 245 grams : Urinary POs Urine flow POs, excretion 3 2 kg. M. kg. mM. |liter ber ay pee BG at hours ERGY ATO ye iag ecules ee ron ae 4.0 (10.0) * (0.04) O=225ehours.is'3 | Sika eae ee 78.0 10.0 0.78 DHO=4 OWNOUES 8. asec secon cece ease ease 46.5 ' 44.0 2.04 AKO—Os#NOUTS.-2 «ons ose cee cone 32.8 17.4 2.54 24.4 hours (pretying)........... 1.0 = = 24:4—26'4 hours. 3.3 8. ee 1.0 68.4 0.07 * Assumed maximum. Causal Factors in the Spontaneous Increase in Phosphate Excretion It will now be profitable to examine some of the factors inherent in the preceding experiments, in an attempt to determine whether one or more of them might be directly responsible for the observed spontaneous increase in phosphate excretion. Tying of the Urinary Papilla; Catheterization —Tying of the urinary papilla can be ruled out at once, since the increase occurs just as readily when sculpins are studied by catheterization alone as when the papilla is tied (Figs. 1 and 2). An evaluation of the factor of catheterization is more difficult. Sculpins are frequently very sensitive to this pro- cedure and struggle vigorously, particularly when the papilla is catheter- ized for the first time. Usually, on subsequent catheterizations, little or no response is elicited. In certain experiments (Nos. 12, 26, 34, 35, 36—Figs. 1 and 2) the papilla had been dilated by one or more catheter- izations previous to the day on which the increase was noted. The intensity of the stimulus in such animals is certainly much less than in animals catheterized for the first time, and one might be led to suppose that the procedure of catheterization is not of itself a significant factor in the observed rise. However, since this procedure is a constant one in the present experiments, it cannot be definitely ruled out.° Pressure; Asphyxia—That the pressure applied to the abdominal wall in milking the urine out of the bladder plays no role is demon- 5 The evidence is quite conclusive that in the sculpin catheterization is not fol- lowed by a shut-down in urine formation (Grafflin, 1935b). Even assuming that it did occur, calculation shows the impossibility of explaining the observed increase in PO, excretion solely on this basis. EXCRETION INORGANIC PHOSPHATE IN SCULPIN 369 strated by experiments in which the urine was drained into the catheter by gravity alone, without the application of any pressure whatsoever (e.g. Nos. 70 and 72, Fig. 1). For the manipulations of catheterization and tying the fish were usually taken out of water. However, that a short period of asphyxia is of no significance in the observed rise is demonstrated by experiments with the animal under water and breathing normally (e.g. Nos. 30 and 41, Fig. 1). Phosphatic Precipitate in Urine.—As previously discussed, sculpin urine frequently contains a white precipitate, which Pitts (1934) has shown to be MgHPO,:3H.,O. Pitts’ suggestion that this precipitate may be retained in the bladder, and so lead to confusion in experiments upon phosphate excretion, has been quoted above. It is essential to determine whether retention of this precipitate, or the presence of phos- phatic concretions, in the bladder or elsewhere might be responsible for the spontaneous increase in phosphate excretion which we have observed. With this end in view, the following experiments have been performed. Bladder Washings—In 2 sculpins the bladder was emptied by catheter and thoroughly washed out with isotonic salt solution acidified with acetic acid. The papilla was then tied and urine allowed to accumulate for 3 hours. Despite the bladder washing, the urinary PO, concentration showed an increase as compared with the pretying urine: No. 62, from 48 to 116 mM./liter; No. 63, from 75 to 100 mM./liter. The total PO, recovered from the bladder washings was 0.0015 mM. (No. 62) and 0.0008 mM. (No. 63). Evidence that even markedly turbid urine can be removed by catheter without leaving more than an insignificant trace of PO, in the bladder was obtained in 2 experiments on freshly caught sculpins. After preliminary emptying by catheter, the total PO, recovered from the bladder by washing as above was 0.0010 and 0.0015 mM. Ureteral Extractions—In 3 sculpins the large ureteral trunk visible on the ventral surface of the kidney was dissected out and extracted separately for each animal with acetic acid. Total PO, recovery from trichloracetic acid filtrates of the extracts was in each case 0.0010 mM. or less. Kidney Extractions—From 4 sculpins the combined posterior and middle kid- ney was removed and extracted separately for each animal by grinding with washed sand and dilute acetic acid. The anterior portion of the kidney was not used as it is entirely lymphoid in nature. The kidney samples weighed 0.17, 0.25, 0.27, and 0.28 gram, and the total PO, recovered from trichloracetic acid filtrates of the extracts was 0.0020, 0.0021, 0.0031, and 0.0026 mM. Urine obtained just before the tissue samples were taken showed PQ, concentrations of 64, 38, 82, and 93 mM../liter, respectively. Again, the combined posterior and middle kidney was removed from 4 freshly caught sculpins, and the entire sample (1.3 grams) ex- tracted as before. Total PO, recovery from the filtrate was 0.0128 mM., giving an average of 0.0032 mM., comparable with the average (0.0025 mM.) of the 4 previous extractions. 6 In spite of the frequency with which magnesium phosphate is precipitated in sculpin urine, bladder calculi have never been observed. In 456 sculpins (used for another study—Grafflin and Gould, 1936) in which the bladder was opened, calculi were never encountered. To this number can be added several hundred specimens observed in the course of previous studies. 370 ALAIN ie GRE ETN To evaluate these experiments, it is essential to know the order of magnitude of a preformed store of phosphate, either as the usual precipi- tate or in the form of concretions, which would be necessary to account for such increases in urinary phosphate excretion as we have observed. This calculation has been made for twelve of the animals in Figs. 1 and 2, and appears in Section 1, Table I, as “ Additional PO, excreted.” It varies from 0.005 to 0.043 mM., average 0.021 mM. This quantity is 17.5 times the average phosphate recovery from bladder washings (0.0012 mM.) in the above experiments. In the extreme case (No. 38) it is 36 times this latter quantity. The average hypothetical pre- formed store is at least 21 times the average recovery from ureteral extractions, and over 7 times that from 8 kidney samples (0.0029 mM.). This factor for the kidney extractions is deceptively low, since we must consider the phosphate contained in the blood, in the renal tissue cells, and in the urine contained in the lumina of the renal tubules and the collecting duct system. The above experiments justify the conclusion that in the sculpin no significant retention of phosphate, as precipitated magnesium phosphate or as phosphatic concretions, occurs in the bladder, ureter or collecting duct system of the kidney. And we are permitted to conclude, in turn, (1) that the phosphate concentration of urine as withdrawn from the bladder can be accepted as the concentration of the urine as it is elaborated by the kidney; and (2) that the spontaneous increase in phos- phate excretion reflects a real increase in renal activity. In seeking still further for an explanation for the observed spon- taneous increase in phosphate excretion, three possibilities were con- sidered. Muscular Activity—Sculpins are normally very sluggish. When handled they frequently struggle quite vigorously, to such an extent that to hold them reliably in routine experimental work requires con- siderable experience. Phosphate compounds are of paramount 1m- portance in muscular activity, and muscular exercise in mammals regu- larly leads to increased urinary phosphate excretion. Under the cir- cumstances, we feel that muscular activity offers the most likely expla- nation for the sharp increases which we have observed. The significant proportional increases in phosphate excretion observed on the second and third days of handling (Fig. 3, Table I) could be explained on the same basis, but at a much lower level of phosphate available for excretion. Increased Absorption of Phosphate from the Alimentary Tract.— The phosphate of the body is derived essentially only from the food. Under experimental conditions sculpins develop a progressive diuresis, EXCRETION INORGANIC PHOSPHATE IN SCULPIN Sot associated with a marked increase in the ingestion and gastro-intestinal absorption of sea water. A sudden increase in the absorption of sea water could well lead to rapid absorption of the phosphate contained in the partially digested food present, and so to a marked increase in phosphate excretion. That this mechanism is not fundamental in our experiments is indicated by the fact that some animals were starved for many days before the spontaneous increase was observed. How- ever, it could conceivably play a role in animals handled soon after catching. Reflex Effects upon the Kidney.—While tying of the papilla plays no role, reflex effects from the stimulus of catheterization cannot be definitely ruled out as a factor. It is impossible to rule out reflex effects from the handling of the skin. Many injection experiments have shown that liberation of adrenalin in the excitement of experimental handling would not of itself lead to increased phosphate excretion. The Question of Tubular Secretion of Phosphate in the Sculpin In approaching the question of tubular secretion, we were guided by the belief of Jolliffe, Shannon and Smith (1932) and Clarke and Smith (1932) that the xylose clearance could be used as a measurement of the rate of glomerular filtration. More recent experiments (Rich- ards, Westfall and Bott, 1934; Shannon, 1934, 1935a, 1935b; Shannon and Smith, 1935; Van Slyke, Hiller and Miller, 1935a, 1935b) have shown that there is some reabsorption of xylose by the renal tubules, _and that the rate of glomerular filtration can apparently be quite accu- rately measured in the dogfish, dog and man by the use of inulin. Clarke (1936) has found in the sculpin that the xylose clearance is about 20 per cent less than the simultaneous inulin clearance. But this discrepancy is not of such an order of magnitude as to invalidate the experiments which we had done with xylose in this animal. Our ex- periments consist of a series of simultaneous phosphate and xylose clearances, as given in Vable INI. (Clearance = (U/P)/, i-¢., urine/ plasma ratio times urine flow.) Data on the total phosphate excretion in these experiments are included in this table for comparison with the data in Table I. The tubular clearance of phosphate has been calculated by deducting the xylose clearance & 1.25 (in accordance with Clarke's figures) from the observed phosphate clearance. In a series of ob- servations on normal sculpins with low urine flow, the maximal U/P ratio for xylose observed by Clarke (1934) was 6.5. In our early ex- periments plasma and urine were removed from three sculpins im- mediately after catching, and the phosphate U/P ratios were 59, 34 and 26. These data would indicate that the U/P ratio for phosphate in SYD ALLAN L. GRAFFLIN the sculpin may normally be considerably greater than can be accounted for by the reabsorption of water from the glomerular filtrate. This indication is established as a fact by the data in Table III, where in four experiments direct xylose-phosphate comparisons were carried out. In only one instance did the phosphate clearance approximate the glomerular clearance, after allowing for 20 per cent reabsorption of xylose.’ A representative protocol for this series of experiments follows. Sculpin No. 15; 330 grams. July 17. From live car stock at noon; urine PO, 24 mM./liter. 5:30 P.M., 0.4 gram xylose intramusc. July 18. 8:17 A.M., urine PO, 13 mM./liter; bladder emptied, papilla tied. 10:13, bled 1.5 ce. 12:09 P.M., urine 0.35 cc.; papilla retied. 2:05, bled 1.5 cc. 4:01, urine 1.25 cc. IUAIBILIa, JD X ylose—inorganic phosphate data for the sculpin. U PLASMA ——V D - P ae Sout NON DEINE PHOSPHATE Cimaince ~ Ou ae No ae LOW er eee XYLOSE Kee tp | EXCRETION lose | phate | lose | phate cc. per kg. | meg. M. Bs kg. 5 kg. M. kg. Hours 24 eae oe Gee ae er nous iy uy Taos ter a4 hae 10 5.1 42.6 229 | 2.7 | 41.3} 2096 50.8 2044 5.67 11 5.2 15.8 261 | 4.0 | 15.6} 337 21.6 318 1.34 15 3.9 6.5 222) 3.1 | 37.7) 103 Dall 56 0.32 3.9 23.3 183 | 4.2 | 53.6) 75 1.4 8 0.33 17 4.1 8.6 ADA eri, 8.6] 409 46.4 398 1.10 4.1 21.5 102 | 2.7 | 32.3) 957 29.6 917 2.58 The frequency and the magnitude of the observed spontaneous in- crease in urinary phosphate concentration and excretion make it clear that experiments upon the excretion of exogenous phosphate would be uninterpretable until any significant increase could be adequately ruled out in a given experimental animal. It seems reasonable to suppose that in certain of the animals reported earlier (Nos. 12, 26, 34, 35, 36), catheterized at long intervals, the spontaneous rise may have occurred more than once. The question arose as to whether any significant in- crease could be suppressed, and the total rate of excretion of endogenous phosphate could be forced to low levels, if a sculpin were followed 7 It is to be noted that when the sculpin kidney is rendered functionally aglom- erular by heavy phlorizinization, phosphate continues to appear in the urine (Mar- shall and Grafflin, 1932). EXCRETION INORGANIC PHOSPHATE IN SCULPIN oS) closely (i.e., persistently, with repeated catheterizations not separated by long intervals) over a sufficiently long period. Lack of space pro- hibits a detailed discussion of this aspect of the problem, but the sum- marized data in Figs. 2 and 3 and Table I indicate that such a condition can be achieved in the sculpin, and that such an animal could be used with confidence in studying the behavior of the kidney toward exogenous phosphate. In summarizing our observations, we may contrast the aglomerular kidney of the goosefish and toadfish, as studied by Marshall and Graff- lin (1933), with the glomerular kidney of the sculpin, as examined here. Both types of kidney excrete inorganic phosphate; the one, where there is no glomerular filtration, the other at U/P ratios so much greater than the simultaneous U/P ratios of xylose that there is no doubt but that tubular activity is involved. The precursor from which the urinary inorganic phosphate is derived in the aglomerular goosefish and toad- fish is unidentified, but it is certainly not inorganic phosphate in the plasma. By inference, until evidence to the contrary is adduced, it may be assumed that the tubular secretion of phosphate in the glomerular kidney is an homologous process. Whether the tubules of the glomeru- lar fish kidney can secrete preformed inorganic phosphate cannot be answered from the present data. The demonstration that the tubules of the glomerular fish kidney can secrete phosphate, presumptively derived from some unidentified pre- cursor in the plasma other than inorganic phosphate, raises the question whether a similar process can occur in the frog or even in the mammal. We know of no evidence at the present time to enable us to answer this question. SUMMARY 1. When urine samples are taken at short intervals, the urinary phos- phate concentration and excretion in the sculpin very often shows a moderate or marked, but transitory, increase. It is demonstrated that this transient increase is not an artifact due to phosphate precipitates in the bladder, ureter. or collecting ducts. 2. By repeated catheterization the rate of endogenous phosphate ex- cretion in the sculpin can be reduced to a low level, and any significant spontaneous increase in excretion can be suppressed. 3. It is shown, by a comparison of simultaneous xylose and phos- phate clearances, that the glomerular kidney of the sculpin can under certain conditions excrete a urine with a phosphate concentration in excess of that which could be explained by glomerular filtration. 4. In line with Marshall and Grafflin’s observations that the aglom- 374 ALLAN L. GRAFFLIN erular kidney excretes phosphate that is derived from some unidentified precursor (other than inorganic phosphate in the plasma), the present observations establish the presumption that a similar process occurs in - the glomerular kidney of the sculpin. LITERATURE CITED CLARKE, R. W., 1934. Jour. Cell. and Comp. Physiol., 5: 73. CLARKE, R. W., 1936. Bull. Mt. Desert Island Biol. Lab., p. 25. CLARKE, R. W., ano H. W. Smiru, 1932. Jowr. Cell. and Comp. Physiol., 1: 131. Derrtse£, A., 1932. Anat. Rec., 54; 185. Epwarps, J. G., 1935. Anat. Rec., 63: 263. Fiske, C. H., anp Y. Suspparow, 1925. Jour. Biol. Chem., 66: 375. Foutn, O., 1929. Jour. Biol. Chem., 82: 83. GrRaAFFLIN, A. L., 1931. Am. Jour. Physiol., 97: 602. GraFFLin, A. L., 1933. Anat. Rec., 57: 59. GrarFiin, A. L., 1935a. Biol. Bull., 69: 245. GraFFLin, A. L., 19355. Biol. Bull., 69: 391. GraFFLin, A. L., AND D. Ennts, 1934. Jowr. Cell and Comp. Physiol., 4: 283. GRAFFLIN, A. L., AND R. G. Goutp, Jr., 1936. Biol. Bull., 70: 16. Jotiirre, N., J. A. SHANNON, AND H. W. Smiru, 1932. Am. Jour. Physiol., 100: 301. MarsHatt, E. K., Jr, anp A. L. Grarrvin, 1932. Jowr. Cell. and Comp. Physiol., 1: 161. MarsHatt, E. K., Jr., anp A. L. Grarriin, 1933. Proc. Soc. Exp. Biol. and Med., 31: 44. Prrts, R. F., 1933. Am. Jour. Physiol., 106: 1. Pitts, R. F., 1934. Jour. Cell. and Comp. Physiol., 4: 389. Ricuarps, A. N., B. B. WestTFALL, AND P. A. Bort, 1934. Proc. Soc. Exp. Biol. and Med., 32: 73. SHANNON, J. A., 1934. Jour. Cell. and Comp. Physiol., 5: 301. SHANNON, J. A., 1935a. Am. Jour. Physiol., 112: 405. SHannon, J. A., 19355. Jowr. Clin. Invest., 14: 403. SHANNON, J. A., AnD H. W. Smiru, 1935. Jour. Clin. Invest., 14: 393. SmitH, H. W., 1930. Am. Jour. Physiol., 93: 480. SmitH, H. W., 1932. Quart. Rev. Biol., 7: 1. Somoeyi, M., 1931. Jour. Biol. Chem., 90: 725. VAN SLYKE, D. D., A. Hitzer, anp B. F. Miter, 1935a. Am. Jour. Physiol., 113: 611. VAN StykgE, D. D., A. HILLer, anp B. F. Miiter, 1935b. Am. Jour. Physiol., 113: 629. SOME, UNUSUAL CYTOLOGICAL PHENOMENA IN THE SPERMATOGENESIS OF A HAPLOID PARTHENO- GENETIC HYMENOPTERAN, 7HNOPLEX SMITHII (PACKARD) CARL H. KOONZ DEPARTMENT OF ZOOLOGY, NORTHWESTERN UNIVERSITY Since the early part of the century numerous investigators have been concerned with the cytology of haploid animals. For extensive bibli- ographies pertinent to the cytology of haploid forms the reader is re- ferred to publications by A. Vandel (1931), Franz Schrader and Sally Hughes-Schrader (1931), and Ann R. Sanderson (1933). More recently Magnhild Torvik-Greb (1935) made a special study of the chromosomes of Habrobracon juglandis and also showed that the spermatogenesis of haploid males and diploid biparental males fol- lows the same course. Allan C. Scott (1936) reported haploidy in the males of a coleopteran, Micromalthus debilis, and especially emphasized aberrant phenomena associated with spermatogenesis. The development of the germ cells of the various haploid metazoan groups is much the same. Obviously a convergence or a parallelism is to be expected in view of the fact that no reduction of chromatin is nec- essary. The salient and conventional features identified with spermato- _ genesis have been pointed out by many authors. In the first spermato- cyte there may be an attempt to recapitulate the reduction division, which is always abortive as regards chromosomes, or the division may be omitted entirely. When division of the cell does take place it is in the form of the pinching off of a non-nucleated cytoplasmic bud. The sec- ond spermatocyte division ordinarily is equational, producing two func- tional spermatids. However, as in the case of Apis mellifica (Meves, 1907), the second spermatocyte division may be unequal, producing only one functional spermatid. Although haploid studies have been concerned chiefly with testing the validity of the concept that males are produced from unfertilized eggs and in confirming the usual development, recently a certain rather unique phenomenon has been reported. Allan E. Scott (1936) found, in haploid parthenogenetically produced males of Micromalthus debilis, that a unipolar spindle was regularly produced in the first spermatocyte division. During the first spermatocyte anaphase apparently the chro- 375 376 CARL H. KOONZ mosomes move away from the unipolar spindle, the movement being independent of the half spindle fibers: The popular idea that spindle fibers are responsible for chromosomal movement is thus shown to be deficient. It is the purpose of the present paper to describe the spermatogenesis of Atnoplex smithi. Emphasis will be placed on unusual cytological aspects which may contribute to our knowledge of the origin of half spindle components and to the morphological and functional reality of half spindle fibers. MATERIALS AND METHODS Ainoplex smuithu is the principal ichneumonid hyper-parasite of Spilocryptus extrematus, the latter living parasitically on the larve of Samia cecropia,a moth. The males develop only from unfertilized eggs and are thus produced parthenogenetically. The original material for this investigation was collected in the vicinity of Chicago, Illinois. The data presented in this paper, however. are based entirely on laboratory specimens reared from virgin females which lay only male-producing eggs. Male gonads of late pupze proved to be the most satisfactory for cytological study, showing few gonial and numerous spermatocyte cells. The gonads of females used to secure chromosome counts in the gonial stage were taken from early pupz. Gonads were dissected out in water and immediately transferred to a fixative, Bouin’s fluid being found most favorable. Gonads were dehydrated and cleared in the usual manner, imbedded in paraffin and sectioned at six microns in thickness. Heidenhain’s iron hematoxylin (long method) was employed anc destaining was done with a saturated aqueous solution of picric acid. When destained sufficiently the sections were exposed briefly to the fumes of concentrated ammonium hydroxide, dehydrated, cleared in xylol and mounted. The gonads of more than one hundred males treated according to the above method were used as a basis for this study. SPERMATOGONIA AND CHROMOSOME NUMBER OF HAPLOID MALES The paired testes lie dorsal to the intestine in the posterior half of the abdomen. Internally the testes are divided into many cysts, con- taining cells in the various stages of development. Testes of very early pupe show cysts containing only spermatogonial cells. The cells are arranged in the cyst so as to give a compact, rosette appearance. Divisions in gonial cells are few, but are synchronous in each cyst. SPERMATOGENESIS OF AHNOPLEX SMITHII 377 It was only from spermatogonial division figures that it was possible to count the chromosomes. The number was established as thirteen (Fig. 1), representing the haploid number of this species. CHROMOSOME NUMBER OF DIPLOoID FEMALES Polar views of the metaphase of gonial cells taken from late female pupz show twenty-six chromosomes (Fig. 2). Therefore it is estab- lished cytologically that females are diploid and that the males develop- ing from eggs laid by virgin females are haploid. SPERMATOCYTES AND THE MATURATION DIVISIONS The testes of late pupze show few gonial cells but show many first spermatocytes, second spermatocytes, spermatids, and spermatozoa. The first spermatocytes at the end of the growth period form a loose rosette or are scattered indifferently within the cyst. They are considerably larger than late gonial cells. The nucleus of the spherical first spermato- cyte (Fig. 3) is prominent and appears reticulated, having one or more poorly defined nucleoli. A prominent cytoplasmic inclusion, which has been referred to frequently as the “ Zellkoppel” or “interzonal body,” is present in these cells. In the late prophase the nuclear membrane becomes indistinct and the chromatin material becomes resolved into chromosomes (Fig. 4). At a slightly later stage the spherical spermatocyte has elongated (Fig. 5), the nuclear membrane has disappeared and the chromosomes have massed. A constriction has appeared, partially isolating that portion of the cytoplasm which subsequently will be pinched off as a non- nucleated bud. Indistinct fibers are seen extending from the vicinity of the chromosomal mass into the bud. This stage probably simulates the metaphase of an ordinary mitosis or meiosis. As a rule the bud contains the “interzonal body” (Fig. 9) which is not visible in some of the eosin counterstained preparations (Fig. 6). When the bud is almost completely detached, distinct fibers can be seen extending be- tween it and the nucleated mother cell (Fig. 6). Simultaneously with the pinching off of the bud there arises a monopolar spindle (unipolar or half spindle), the components of which would appear to be formed as outgrowths from the massed chromosomes. The monopolar spindle is cone-shaped with the apex pointing toward the bud; although at first small and inconspicuous, it later becomes progressively larger, elongates and becomes very prominent (Fig. 7). The massed chromosomes and their attached half spindle compo- nents next separate into two approximately equal halves (Figs. 8-9). 378 CARL H. KOONZ Simultaneously with this separation and with the spindle elongation there appears to be distinct anaphasic movement on the part of the chromosomes. At the completion of this anaphasic movement the chromosomes, with the attached spindle fibers trailing, have traversed the cell and occupy that part of the cell distal to the locus of the formerly attached bud (Fig. 10). There is no formation of a nuclear membrane between the first and second maturation divisions. The telophase chromosomes gather into a dense compact mass during an abbreviated interkinesis. In the formation of this interkinetic mass the chromosomes, preceding, appar- ently drag their attached fibers with them (Fig. 11), so that in early interkinetic masses the fibers of the monopolar spindle radiate from the mass (Fig. 12). In later interkinesis the spindle fibers completely disappear (Fig. 13). | New spindle fibers identified with the second spermatocyte division soon make their appearance. The fibers, a number of which constitute the bipolar spindle, from all appearances are formed as outgrowths from some material contained within the interkinetic chromosomal mass. The components of each half spindle comprising the bipolar spindle are at first darkly staining short extensions from the chromosomal mass (Fig. 14) but subsequently are less basophilic and become progressively longer (Fig. 15). The components of each half spindle very early assume a cone formation. The cones, however, do not grow out in opposite direc- EXPLANATION OF PLATES All figures are camera lucida drawings drawn to the same scale. They were originally drawn to represent a magnification of 4,807 and have been reduced one- third as represented here. EXPLANATION OF PLATE I Fic. 1. Side view of spermatogonial metaphase showing thirteen chromo- somes, the haploid number. Fic. 2. Polar view of female gonial cell showing twenty-six chromosomes, the diploid number. Fics. 3-4. First spermatocytes, early and late prophase respectively. Fics. 5-6. First spermatocytes showing the pinching off of the cytoplasmic bud and the formation of the monopolar spindle. Fic. 7. Second spermatocyte showing prominent monopolar spindle. Fics. 8-10. Second spermatocyte showing anaphasic movement of the chro- mosomes, with spindle fibers trailing. Fic. 11. Chromosomes of second spermatocyte, with spindle fibers trailing, gathering to form interkinetic mass. Fic. 12. Second spermatocyte with half spindle fibers radiating from inter- kinetic mass. Fic. 13. Interkinetic mass. Half spindle fibers have disappeared. SPERMATOGENESIS OF AANOPLEX SMITHII 379 380 CARL H. KOONZ tions in the same plane. On the contrary, both cones not infrequently appear to grow out in the same general direction and in forming the bipolar spindle in which the two halves are opposed in the same plane (Fig. 17), each half cone may be generally figured as describing an arc of about ninety degrees. In the course of such movement many crescentic figures may be seen (Figs. 14-16). THE SPERMATIDS The second spermatocyte division may be equal, producing two nor- mal spermatids (Figs. 18-19). Frequently, however, when the division has proceeded to late anaphase, it is suppressed (Fig. 22) and the cyto- plasm of the two potential spermatids, instead of separating, fuses and an aberrant spermatid is formed. ‘This spermatid is twice the size of the normal spermatid and is at first binucleate (Figs. 23-25). Later the two nuclei of the aberrant spermatid fuse forming a single large nucleus (Fig. 26). In early spermiogenesis the metamorphosis of the normal spermatids (Figs. 20-21) and the aberrant spermatids (Figs. 27-28), both types of which may be found within the same cyst, appears to be similar except for the distinct dimorphism in size. Dimorphism in size of sperm has not been observed and therefore no evidence is available that would suggest that these large spermatids metamorphose into functional sper- matozoa. It would seem likely that they are pathological in character. UNUSUAL CYTOLOGICAL PHENOMENA AND POSSIBLE SIGNIFICANCE Monopolar Spindle Spindles which normally have only one pole have been reported re- cently by several authors. Metz (1933) described a monocentric mitosis which normally effects a regular and precise segregation of chromo- somes in the first spermatocyte in Sciara. Sally Hughes-Schrader EXPLANATION OF PLATE II Fic. 14. Origin of bipolar spindle from interkinetic chromosomal mass of sec- ond spermatocyte. Fics. 15-16. Crescent figures of second spermatocyte prophase. Fic. 17. Metaphase of second spermatocyte division. Fics. 18-19. Normal spermatids. Fics. 20-21. Early spermiogenesis of normal spermatids. Fics. 22-25. Binucleated aberrant spermatids. Fic. 26. Large mononucleated spermatid, the nucleus being formed by the fusion of two nuclei. Fics. 27-28. Early spermiogenesis of aberrant spermatid. SPERMATOGENESIS OF ASNOPLEX SMITHII 381 Plate II 382 CARL H. KOONZ (1935) reports half spindle formation in connection with the condensed group of chromosomes in the germ cells of Phenacoccus acericola, re- duction being effected by the consequent passage of the condensed group to the single pole of the spindle. Scott (1936) found that a uni- polar spindle is regularly formed in the first spermatocyte division of a haploid parthenogenetically produced beetle, Micromalthus debilis. The monopolar spindle described in connection with the present work is similar in its general features to those already reported. They are all identified with the maturation divisions. However, the spindles reported by Metz and Sally Hughes-Schrader are found in diploid ani- mals. The monopolar spindle here described and the one reported by Scott are found in the first spermatocyte of haploid animals, and there- fore are in particular agreement as to time and place of formation. It is commonly assumed that spindle fibers are in some way asso- ciated with a bipolar tension. In the light of these monopolar spindles it would seem that spindle fibers may exist entirely independent of any bipolar condition. Functional Reality of Half Spindles Anaphasic movement of chromosomes such as that described here presents an interesting problem. It has generally been assumed that whatever the nature of the half spindle may be, it is concerned with, if not responsible for, this anaphasic movement. Quite obviously, the anaphasic movement here described in connection with the first sperma- tocyte is in no capacity identified with, and is entirely independent of, the half spindle fibers. The movement of these chromosomes is no doubt something like that described by Metz (1933) for Sciara during the first spermatocyte segre- gation of chromosomes or similar to the movement reported by’ Scott (1936) for the chromosomes during the anaphase of the first spermato- cyte division in Micromalthus debilis. Metz indicates that the move- ment in Sciara is an “ autonomous ”’ movement on the part of the chro- mosomes, and that the half spindle fibers could exert only a retarding influence. In view of these cases where half spindles play no part in the ana- phase movement of chromosomes, it is quite obvious that any theory which attempts to attribute the force responsible for the chromosomal movement to half spindles or their components is not correct in its entirety. SPERMATOGENESIS OF AANOPLEX SMITHII 383 Morphological Reality and Origin of Half Spindle Fibers Some evidence exists for the morphological reality of half spindle fibers. Perhaps the most convincing evidence, as pointed out by Schra- der (1934), pivots on the simple argument that if the half spindle or their components can be made to bend, it is evidence for their morpho- logical reality. Schrader noted that metaphase half spindles or their components undergo bending and distortion in centrifuged cells, and in view of this experimentally produced evidence concluded that the half spindle components have morphological reality. If bending and distortion of half spindles or their component fibers is evidence for their morphological identity, then perhaps the evidence afforded by the present study is most convincing. In the first place ex- treme bending of fibers is observed when the telophase chromosomes gather to form the interkinetic mass (Figs. 11-12). The chromosomes preceding drag their attached fibers with them so that in early inter- kinesis they radiate from the chromosomal interphase mass. In the second place the bipolar half spindles of the second spermatocyte divi- sion, in becoming oriented so that the two halves are opposed in the same plane, exhibit considerable bending (Figs. 14-16) as evidenced in the crescentic figures. The phenomena are even more convincing when it is realized that they are natural phenomena and not artificially pro- duced. Perhaps further evidence for the morphological reality of half spin- dle fibers is forthcoming in view of the apparent manner in which the spindle components arise. Carothers (1934) has especially emphasized that half spindle fibers are of chromosomal origin. In the present study it appears that the monopolar spindle of the first spermatocyte and the bipolar spindle of the second spermatocyte arise as outgrowths of some substance contained within the chromosomes. If the spindle fibers are outgrowths of the chromosomes, this would appear to be evidence per se that they do have morphological reality. SUMMARY The number of chromosomes as demonstrated by the cytological study of gonial cells is thirteen in the male and twenty-six in the female. Experimentally it was found that virgin females gave rise only to males. In view of the experimental and cytological evidence it is concluded that males having the haploid chromosome number develop from unfertilized eggs and are thus produced parthenogenetically. The pinching off of a non-nucleated cytoplasmic bud characterizes the first meiotic division. 384 CARL H. KOONZ A monopolar spindle is regularly formed simultaneously with the pinching off of the first spermatocyte bud. This spindle would appear to be of chromatin origin. | The chromosomes or chromosomal masses, with spindle fibers trail- ing, exhibit pronounced anaphasic movement identified with the first spermatocyte division. The force responsible for this anaphasic move- ment apparently is entirely independent of the half spindle fibers. There is no formation of a nuclear membrane between the first and second spermatocyte divisions. However, after their characteristic ana- phasic movement the chromosomes gather into a compact mass for an abbreviated interkinesis. The tendency for the half spindle components to be dragged about by the chromosomes during the anaphase of the first spermatocyte division and during the movement of the chromosomes in the formation of the interkinetic mass would seem to indicate that half spindle fibers have morphological reality. The bipolar spindle identified with the second spermatocyte division appears to originate as an outgrowth from the interkinetic chromosomal mass. This would seem to be evidence for morphological reality of the fibers. The second spermatocyte division may be equal, producing two nor- mal spermatids. Not infrequently aberrant spermatids are formed as a result of the secondary fusing of the parts of a partially divided second spermatocyte. They have a single large nucleus which is formed as a result of the fu- sion of two normal nuclei. In early spermiogenesis the development of the two types of sperma- tids appear to be similar except for differences in size. At present no evidence is available to indicate that the aberrant spermatids develop into functional spermatozoa. ACKNOWLEDGMENT The writer wishes to express his indebtedness to Dr. C. L. Turner for advice and assistance during the course of this work. BIBLIOGRAPHY CarorHers, E. E., 1934. The chromosomal origin of spindle fibers. Anat. Rec., 60: 81 (Supplement) (Abstract). HUuUGHES-ScHRADER, S., 1935. The chromosome cycle of Phenacoccus (Coccide). Biol. Bull., 69: 462. Merz, C. W., 1933. Monocentric mitosis with segregation of chromosomes in Sciara and its bearing on the mechanism of mitosis. Biol. Bull., 64: 333. SPERMATOGENESIS OF AZNOPLEX SMITHII 385 Meves, F., 1907. Die spermatocytenteilungen bei der Honigbiene (Apis mellifica L.), nebst bemerkungen tber Chromatinreduktion. Arch. f. mikr. Anat. und Entwicklungsgeschichte, 70: 414. Sanperson, ANN R., 1933. The cytology of parthenogenesis in Tenthredinide. St. Andrews Univ. Publ., No. 33: 321. ScurRADER, F., 1934. On the reality’of spindle fibers. Biol. Bull., 67: 519. ScHRADER, F., AND SALLY HUGHES-SCHRADER, 1931. Haploidy in Metazoa. Quart. Rev. Biol., 6: 411. Scott, A. C., 1936. Haploidy and aberrant spermatogenesis in a coleopteran, Micromalthus debilis LeConte. Wistar Institute Bibl. Ser., No. 277: 13. Torvik-Gres, M., 1935. The chromosomes of Habrobracon. Biol. Bull., 68: 25. VANDEL, A., 1931. La Parthénogenése. G. Doin et Cie, Paris. PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS PRESENTED AT THE MARINE BIOLOGICAL LABORATORY JuNE 30, 1936 The effect of methylene blue on the spectrophotometric picture of hemo- globin, CO-hemoglobin and CN-hemoglobin.. Matilda M. Brooks. The purpose of these experiments was to find out the effect of injections of methylene blue on the spectrophotometric picture of hemoglobin (hb) and of CO-hb and CN-hb in rabbits. The blood was taken by heart puncture, diluted to 1 per cent with .4 per cent NH.OH and then analyzed immediately in the spectrophotometer. The extinction coefficient at wave lengths 540/560 mu = R were determined. These values indicate the percentage of oxy-hb present. There were three main results found; the absorption maximum of the CN-hb curve and the curve for the blood containing methylene blue were identical with that for oxy-hb; the time curve for blood containing CO and methylene blue showed 100 per cent oxy-hb within a short time. There was no difference in the absorption of oxy-hb and CN-hb in the region of the u.v. spectrum but the absorption was greater in the case of oxy-hb than of CN-Hb when wave lengths from 7300 to 10,300 A were used. Eggert, using the infra-red region, found that blood containing methylene blue gave the same spectrum as oxy-hb while that for blood subjected to CO even after 45 minutes aeration, gave the spectrum of CO-hb. The conclusions are as follows: the evidence for the change of CO-hb to oxy-hb by the action of methylene blue, without invoking the methemoglobin formation theory seems unequivocal. This may be caused by a catalytic action on the part of methyl- ene blue, or it may be due to a poising effect on the oxidation-reduction potential of the system making it compatible with the re-formation of oxy-hb. The phosphatase content of the developing chick embryo. WUarry J. Lipman. The phosphatase activity of the white, of the yolk and of the embryo were de- termined in eggs of white Leghorn hens. The determinations were made in intervals of three days during the process of incubation. For the determinations the Brigg’s modification of the Bell Doisy colorimetric method for the determination of inorganic phosphorus was used. In the technique the amount of inorganic phosphate formed by the enzyme activity is measured and calculated in total milligrams (absolute values) and in milligrams per gram of wet body weight (relative values). The absolute values and the relative values for the whole egg at zero days incuba- tion are 0.487 mg. phosphorus and 0.082 mg. phosphorus respectively. Both of these values rise to two maxima, one at six days and the other at fifteen days of incu- bation, and in doing so run in a parallel fashion. The relative phosphatase values for white, yolk and embryo and the absolute values for white and yolk show these same two maxima. However, the absolute data for the embryo give a fairly good S-type curve. The steepest part of the curve coincides with the period of high ossification activity in the embryo. When the absolute and relative values for the embryo are plotted together on the same graph, the drop of the relative values after the fifteenth day stands out conspicuously against the continuous S-type rise of the absolute values. This drop can be interpreted by the fact that from this day on the 386 PRESENTED AT MARINE BIOLOGICAL LABORATORY 387 growth of the embryo proceeds at a very rapid rate with which the increase in phosphatase activity does not hold pace. It was found that the material used had a high coefficient of variation. This was also true for the experimental data. These facts make the results less significant than was expected. The description of an oxidative process maintaining the frequency of the heart beat. Kenneth C. Fisher and Laurence Irving. Oxidation-reduction potentials and potentiometric determination of ascor- bic acid. Eric G. Ball. JuLy 7 Nerve cells without central processes in the fourth spinal ganglion of the frog. Alfred M. Lucas. Serial sections of the fourth spinal nerve of the bullfrog which included the roots, spinal ganglion, dorsal rami, spinal nerve trunk, ventral ramus, communicating ramus, sympathetic trunk anterior to the fourth nerve and the celiac nerve which were stained in toto with osmic and with silver methods, furnished fiber counts in the regions mentioned. The number of myelinated fibers found in the two roots was 667 and about 259 of these passed to the dorsal rami leaving 408 fibers for the spinal nerve, but distal to the ganglion 950 myelinated fibers were counted. Total fiber counts from silver preparations revealed 1,372 in the dorsal root, 648 in the ventral root and 883 subtracted passing to the dorsal ramus left 1,137 for the spinal nerve. Actual count of the spinal nerve showed 5,277 fibers. The additional fibers arose from the dorsal root ganglion in which were located 5,220 cells. It is concluded, therefore, that some 3,500 cells are present in the spinal ganglion which lack central processes passing through the dorsal root to the cord. The distal processes pass through the communicating ramus to the celiac nerve. Other interpretations have been considered, namely, that dorsal root fibers might have branched in the ganglion, that during development the sympathetic ganglion failed to migrate peripherally as expected, and that the missing central processes had been overlooked when counts were made on the dorsal root. Receptor areas in the vene cave and the pulmonary veins and their relation to Bainbridge’s reflex. José F. Nonidez. Extinction of reflex responses tn the rat. C. Ladd Prosser. Factors influencing the electrical activity of the brain. R. W. Gerard. JuLy 14 The permeability of the erythrocytes of the ground-hog. A. K. Parpart. The red cells of the ground-hog have been found to be very much more permeable to a variety of dissolved substances than those of any other species thus far studied. The penetration of non-hemolytic non-electrolytes from isosmotic solutions was followed by the hemolysis method, using photoelectric recording. A few typical results are given below, the figures indicating the time in seconds required for 75 per cent hemolysis in 0.3 M solutions at 20° C.: ethylene glycol, 1.9; glycerol, 2.2; diethylene glycol, 3.0; triethylene glycol, 5.3; erythritol, 5.8; sorbitol, 59.0; mannitol, 80.0; d xylose, 133; d-l arabinose, 256; glucose, 1080; propyl alcohol, 1.4; a, 6 dioxy- propane, 1.9; a, y dioxypropane, 4.5; water, 0.9. The permeability of these cells to glycerol is only slightly reduced by NaHCOs;, Cut*, and CO:. The red cells of the 388 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS ground-hog combine the permeability characteristics of those of man and the mouse. Those of man are known to be permeable to sugars but not to polyhydric alcohols such as mannitol, the reverse is true in the mouse, while the ground-hog erythrocytes are readily permeable to both of these types of substances. Effects of salts on the injury potentials of frog’s muscle. H. Burr Steinbach. Variations in injury potential were noted as the salt content of the fluid bathing the uninjured portion of the frog sartorius was varied between the limits 0.1 molar to 0.0005 molar. When the observed potential differences are plotted against logarithms of concentration of salt applied, regular curves result. The shapes of these curves can be predicted by the equations for diffusion potentials for such salts as NaCl and - NaC;H;02. With high concentrations bathing the uninjured surface the slope of the curve is characteristic of the salt applied. With low concentrations the slope of the curve is determined by the electrolytes of the tissue and is nearly independent of the nature of the salt in the solution. With KCl solutions, the p.d/ log concentration curve is as calculated for concentrations less than 0.01 molar. With concentrations 0.01 molar or greater, the observed potentials become increasingly negative to the calculated as the KCl concentration increases. With isotonic (ca 0.1 molar) KCl, the difference between observed and calculated values is 40 to 50 millivolts, or nearly equal to the injury potential itself. A value of 40 to 50 millivolts must be added to calculated values in order to obtain a quantitative agreement with the observed values for NaCl, NaC.H3O. and the dilute range of KCl. It is concluded that the presence or concentration of such indifferent salts as NaCl and NaC,H;O2 (when they are applied to the uninjured surface) has little effect on the injury potential. Po- tassium, on the other hand, in concentrations 0.01 molar or greater, acts on the uninjured surface to lower the measured injury potential, probably by producing “injury” at the point of application. The concentration at which KCl first shows specific effects is about that at which KCl contracture sets in. Free calcium in the action of stimulating agents on Elodea cells. Daniel Mazia and Jean M. Clark. (The full paper appears in this issue of the BioLoGicAL BULLETIN.) ; A photoelectric method for recording past chemical reactions. Application to the study of catalyst-substrate compounds. Kurt G. Stern and Delafield DuBois. Juty 21 Observations on lens regeneration in Amblystoma. W. W. Ballard. In the literature, Amblystoma mexicanum is said to be capable of Wolffian lens regeneration at all stages in its life history. It is found that Amblystoma punctatum loses this ability very rapidly in late embryonic stages. At stages 38-39, lenses are formed and separated from the iris in 4 days in 100 per cent of cases. At stage 43, six days later, practically no lens regeneration is obtained. Similar results were gotten from embryos of A. tigrinum, and A. microstomum. During larval stages of A. punctatum, A. tigrinum, A. microstomum, A. jeffersonianum and A. opacum, no lens regeneration was obtained over periods up to four months after extirpation. Using uniform material, a greater percentage of A. punctatum embryos regener- ated lenses in light than in darkness, at 5° C. than at room temperature, in 6/8 Ringer’s solution than in 1/8 Ringer’s. Parabiotic twins were formed in Amblystoma punctatum at stages 38-41 in which one lens lay in the pupil between two eyeballs whose irises were superimposed. PRESENTED AT MARINE BIOLOGICAL LABORATORY 389 Of these two eyeballs, the one which had previously been deprived of its lens regener- ated a new one, in spite of the presence partly within it of the lens of the other eye. Fertile eggs from pheasants in January by night-lighting. T. H. Bissonnette. A quantitative analysis of the anterior pituttary-ovulation relation in the frog. Roberts Rugh. Since the original work by Dr. O. Wolf in 1929, ovulation has been induced in an increasing variety of amphibia. In 1934 it was first suggested that while-either male or female glands could be used to induce ovulation, there was a potency difference favoring the female anterior pituitary. Since the response to anterior pituitary injection was not uniformly dependable, it seemed desirable to study this relationship quantitatively. The work was divided into two phases: first, an attempt was made to determine any correlations between sex, body weight, body length, gonad weight, and anterior pituitary weight. The second phase was a study of the degree of ovulation elicited by quantitatively determined doses of the gland. Over 700 frogs were used, covering two periods: November and February. These months were chosen as representing the beginning and the end of the normal hibernation period. Since it was found that anterior pituitaries varied in weight from 0.6—-1.6 mgm., in order to lessen observational error the glands were removed from frogs of approximately the same body length and were weighed in groups of twenty. While the body lengths of males and females averaged almost the same for November and February, there was consistent reduction in average body weight and average gonad weight, over this hibernation period. Body length, as measured from nares to cloacal opening, was the standard by which other variables were considered. It was found that males and females with body lengths less than 71 mm. were imma- ture; that those between 71—74 mm. showed the beginning of sex differentiation and gonad growth; and males and females longer than 74 mm. showed sex differences in the relative weights of their anterior pituitaries. These differences favored the males, both in November and in February. Over this hibernation period there was con- siderable reduction in relative weight of the anterior pituitaries in both sexes, at some stages this was as much as 29 per cent. There is a similar reduction, over the hibernation period, of relative ovarian weights. If anterior pituitary weights are plotted against ovarian weights, it is demonstrated that there are correlating reduc- tions in the weights of these two organs, all points falling on a straight line for both November and for February. In respect to the induction of ovulation, it was found that in November 8 mgm. of male anterior pituitary induced about 42 per cent of the eggs to leave the ovaries while 5 mgm. of female gland tissue induced 85 per cent ovulation. In February 4 mgm. of either male or female gland tissue would induce 100 per cent ovulation, and in doses less than this there were indicated differences in potency favoring the female gland. These seasonal differences may be due to hibernation concentration of the donor’s anterior pituitary; to greater susceptibility of ovaries in February than in November to the anterior pituitary hormone; or even to the activation in this pre- breeding period of the recipient’s own anterior pituitary. The anterior pituitaries removed from females which had been induced to ovulate showed no decrease in potency in respect to inducing ovulation in other females. This supports the thesis that the injection of the hormone is comparable to the liberation of the host’s hormone, and that the host’s hormone is in no way affected. If frogs are selected at random, it will be found that the average male anterior pituitary is about 16 per cent heavier than that of the average female, but is only 390 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS 60 per cent as potent in respect to inducing ovulation. This supports the thesis that the female glands are approximately twice as potent as the male glands. Like many biological processes, which at first seem to be simple, this relationship between the anterior pituitary and ovulation resolves itself into a number of variables: (1) size and sexual maturity of the donor of the anterior pituitary; (2) concentration of the hormone through hibernation reduction in weight; (3) activity of the recipient’s own gland; (4) source of the hormone, i.e., from male or from female; (5) dose of the hormone (mgm. of gland tissue); (6) size and sexual maturity of the recipient; and (7) size and susceptibility (maturity) of the ovaries to ovulation induction. This quantitative study points toward seasonal as well as metamorphic changes in the anterior pituitary which must be studied from the cytological point of view. This study is at present being made on the bullfrog, Rana cateshiana, and will be reported subsequently. (This paper will appear in full in the January, 1937 issue of Physiological Zoology.) Experimentally induced coupling in the toad, Xenopus levis. (Moving picture.) H. A. Shapiro. joy 28 The relation between vitamins and the growth and survival of goldfishes in homotypically conditioned water. Gertrude Evans. A possible endocrine réle of the eosinophil leucocytes in the female rat. C. P. Kraatz. (The full paper will appear shortly in the American Journal of Physiology.) An interpretation of the secondary lymphoid nodules in the albino rat. J. E. Kindred. Quantitative data from a detailed study of serial sections of submaxillary, inguinal and mesenteric lymph nodes, and Peyer’s patches of eight 80-day old albino rats from the same litter, are presented in support of Rohlich’s view that the secondary lymphoid nodule is both a germinal center and a center of reaction to substances in the blood stream. The secondary nodules are prolate spheroidal or spherical in shape and each is roughly divided into a light and a dark zone which differ quantitatively from each other in respect to rates of mitosis, incidences of macrophages, and the several varieties of lymphocytes. The light zones are capped externally by a cup-shaped layer of small lymphocytes and reticulum. The blood capillary plexus of the nodule lies in the light zone. Because the dark zone has a statistically significant higher rate of mitosis than the light zone in all nodules studied (e.g. mitosis per 1,000 cells with standard error of mean, submaxillary nodes: dark zone 31 (2.3); light zone, 6.3 (0.3)), it is concluded that the dark zone is the germinal center of the nodule. On the other hand, the presence of a blood capillary plexus, approximately equal incidences of small lympho- cytes (e.g. 54 per cent (1.6) in mesenteric nodes; 60 per cent (0.6) in Peyer’s patches) and macrophages containing degenerating nuclei of small lymphocytes (e.g. 52 (6.3) per 1,000 cells in mesenteric nodes; and 42 (2.3) in Peyer’s patches) in the light zones of the nodules of both lymph nodes and Peyer’s patches, suggests that the functional activity of the light zone is related to the local reaction of small lymphocytes to substances in the blood. The quantitative data suggest that the minimum mitosis rate necessary to maintain this relation is the rate in the dark zones of Peyer’s patches (e.g. 15.5 (0.5) per 1,000 cells). The close relation between the nodules of the lymph nodes and the lymphatic channels, together with the higher rates of mitosis in the PRESENTED AT MARINE BIOLOGICAL LABORATORY 391 dark zones (e.g. submaxillary nodes, 31 (2.3); inguinals, 32 (2.4); mesenterics, 30 (2.5) per 1,000 cells) suggest that not only do the dark zones of the nodules of the lymph nodes function to maintain a local balance of small lymphocytes, but in addi- tion contribute small lymphocytes to the lymph stream. ' From the fact that medium-sized lymphocytes predominate in the dark zones (e.g. submaxillary nodes; dark zone 80 per cent; light zone, 40 per cent) and are the cells most frequently found in mitosis (e.g. submaxillary nodes; 90 per cent of cells in mitosis are medium-sized lymphocytes), it is concluded that the medium-sized lymphocytes are the lymphoblasts. Physiological adjustments to diving in the beaver. Laurence Irving. When a normal beaver ceases breathing during submersion, its heart is inhibited. During periodic breathing, anesthetized beaver showed slowing of the heart and a greater than fifty per cent fall in arterial blood pressure during the intervals between breathing movements. A similar fall in blood pressure, slowing of the heart and inhibition of breathing movements followed from gentle inflation of the lungs of the beaver. In addition it was observed that the blood flow through the hind leg was nearly arrested when the lungs were inflated. It was not obvious to what extent constriction of the vessels in the leg contributed to the retarded blood flow, but other observations indicated that the peripheral blood vessels were able to change blood flow rhythmically. Apparently the depressor effect of inflation of the lungs is the same in the beaver as it isin other animals. The difference in the beaver is in the accentuation of the response, which easily arrests breathing for a minute, slows the frequency and decreases the force of the heart and very nearly arrests blood flow through the muscles. These responses are part of the respiratory adjustment of diving animals for enduring asphyxia. It is suggested that the stimulus of inflation of the lungs occurs at the start of a dive and brings about this sequence of physiological adjustments which prepare for asphyxial conditions before the chemical stimuli are effective. The responses seen in the beaver following inflation of the lungs are like those of other mammals except that they are quantitatively greater. AvucustT 4 Development of Arbacia eggs without nuclei: parthenogenetic merogony. Ethel Browne Harvey. (The full article appeared in the August, 1936 issue of the BioLocicaAL BULLETIN, vol. 71, pp. 101-121.) Cortical changes in Arbacia eggs during fertilization—a moving picture. F. Moser. . Temperature effects on mitotic changes in Arbacia eggs. H. J. Fry. Sea-urchin larvae with cytoplasm of one species and nucleus of another. Sven Horstadius. (The full report has already appeared in Mem. Mus. d’ Hist. nat., Bruxelles, Ser. 2, vol. 3.) Aucust 11 A study of the cells of the adrenal gland of the ewe during estrus and pregnancy. Laura J. Nahm. Studies in calcification: III. The shell of the hen’s egg. E. Alfred Wolf and Grace Riethmiller. In two previous papers by McBride and Wolf it was shown that the Gomori modification of the silver nitrate method can be successfully applied to the study 392 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS of early stages of calcification of bone and teeth. Since these structures are com- posed mainly of calcium phosphate, it was desirable to test the method on calcified structures that are predominantly carbonates; the shell of hen’s egg was selected in which 89-97 per cent is calcium carbonate and only 0.5—5 per cent calcium phosphate. The eggshell shows three layers: a thin, hard, structureless outer membrane, the main calcareous body and an inner fibrous layer into which dip the calcified nipples of the main body. After treatment the shells were dark on both surfaces showing where the silver nitrate had reacted, but the main body of the shell was still white and hard; decalcification therefore was required. Cross sections through the main part of the shell showed no traces of silver, except for a brown color of the surface membrane. The fibrous membrane under low power magnification had the appearance of a leopard skin: a large number of black spots were more or less evenly scattered over the area of the membrane, the fibers of which were light brown in color. Under high magnification (oil) it could be seen that the spots were due toa precipitation of silver within the fibers themselves. This is taken as an indication of the location of the calcium phosphate deposits in the eggshell. The failure of the calcium carbonate to react in the case of the eggshell would then suggest that our description of the process of calcium deposition in bone and teeth of rat embryos, as it was given last year, was incomplete since it did not include the carbonate portion (6 per cent) of the calcifying structures. Diffraction paiterns of striated muscle and sarcomere behavior during contraction. Alexander Sandow. A method has been devised that permits photographically recording the diffrac- tion patterns formed by 1 mm. long segments (i.e., about 400 striations) at different positions along the length of the frog sartorius (Rana pipiens) over the course of single contractions. The simultaneously produced myograms are included on the records. The patterns produced by a set of different segments of the sartorius undergoing isometric twitch, supermaximally stimulated at 7m situ length, include the following: all records have 1st order spectra of practically unchanged intensity over the whole course of the twitch; the displacements of the 1st order spectra produced by the segments of the tibial two-thirds of the muscle indicate that, in relation to resting length, the corresponding sarcomeres are shortening, while the patterns of the pelvic- third segments show that here the sarcomeres are being stretched. The nature of the shortening and stretching of the sarcomeres at the various muscle levels indicates that the contraction wave is initiated at the nerve plexus and then spreads up and down the muscle away from this region. During supermaximally stimulated moderately loaded isotonic contractions the patterns show 1st order spectra similarly unchanged in intensity, but with very great displacements, indicating sarcomere shortening up to 20 per cent; contraction, however, takes place at all levels of the muscle. During both isometric and isotonic contractions the Oth order increases, and the 2nd order decreases in intensity. The above results, particularly the intensity variations, are interpreted to mean that Jordan’s description of contracting sarcomere behavior (Physiol. Rev., 13: 301, 1933) does not hold for the frog sartorius; but that in the contracting frog sartorius: (1) a sarcomere, in the sense of consisting of a Q band flanked at each end by one-half of each of the adjoining I bands, remains intact, although (2) it may contract or be stretched depending on its position in the muscle and the nature of the contraction; and (3) there occurs an unequal change in the refractive indices of the Q and I substances, or a variation in the ratio of the lengths of Q and I bands, or both of these modifications. And whatever the directions of these changes for contraction of sarcomeres, they are probably opposite in direction for stretch. PRESENTED AT MARINE BIOLOGICAL LABORATORY 393 The effect of some oxidation-reduction indicator dyes (phenol indophenol) on the eyes and pigmentation of normal and hypophysectomized amphibians. Frank H. J. Figge. . The discovery of the Lewises that the indophenol dyes produced a marked pallor of the skin and loss of the pigmented layer of the optic cup in tadpoles was used to study amphibian pigmentation. Phenol indophenol and ortho-cresol indophenol were tested on a series of six kinds of amphibian eggs: normal and hypophysectomized Amblysioma punctatum, black and white Mexican axolotl, Necturi, and bullfrog. The hypophysectomies were performed between the Harrison stages 30-37. The animals were placed in the dye solutions (1 : 500,000 and 1: 1,000,000) near the ’ hatching stages and kept there continuously for a period of three months. As the animals grew, it was necessary to increase the concentrations. Dye solutions were changed daily, and during the last half of the experiment, twice daily. From a study of the relative sensitivity of the salamanders to the dye, it was evident that the degree of sensitivity paralleled the metabolic rate. The higher the metabolic rate, the less sensitive the animal, and less effective the dye. The dyes eliminated the pigment from the tails of hypophysectomized Amblystoma punctatum larvae, proving conclusively that the dyes do not exert their influence on the melanophores through any intermediate effect on the hypophysis or its products. The white axolotls decolorized the dye as rapidly as the black. The black animals became as white as the white axolotls. The rate of decolorization of the dye was, therefore, not related to the rate of pigment destruction. The dyes were most effective in eliminating the pigment from Necturus, the form closely resembling cave salamanders that exhibit pigment and eye defects similar to those produced by the dyes. The animals raised in dye solutions exhibit eye defects that vary in degree and uniformity with the species, within the species, and even in the same individual. These defects involve not only the retina and pigmented layer of the eye, but also the scleral cartilage, lens and cornea. The elimination of neutral red by the kidney tubules. Robert Chambers. AucGust 18 Quantitative studies on blood clotting. H. P. Smith and E. D. Warner. The basis of the principle of the master reaction in biology. Alan C. Burton. The principle that the rate of a chain of processes is determined dominantly by the speed of the slowest process in the chain underlies the widely used method of “‘temperature analysis” of complex biological processes. Its origin in the analysis of a chain of two irreversible monomolecular reactions is traced. A table is calcu- lated showing how much one velocity constant must be slower than the other in this system in order that the slower reaction is within a given percentage of complete dominance of the chain. This requirement is so exacting that when the Arrhenius law of rate of change with temperature is followed, it is impossible that one reaction may be within 10 per cent of a true master reaction at 0° C. and the other similarly dominant at 40° C. unless the ‘“‘critical increments,” u, differ by at least 16,500 Calories. Smaller temperature ranges require proportionately greater differences in the critical increments. The straightness of a line through observed points on an Arrhenius plot may result from dominance shared by a number of reactions rather than of a single master reaction. 394 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS In longer, more complicated chains the requirements for a true master reaction are more exacting, and for systems in the “‘steady state” the principle has no appli- cation. It is concluded that the principle of the master reaction has a very restricted application even in these idealised chains and still less to the real systems of biology. Efficiency of photosynthesis in purple bacteria. C.S. French. Streptococcus varians assimilates carbon dioxide with hydrogen in the light but not in the dark. The effect of hydrogen pressure, pH, wave length, intensity, and temperature on the rate of the reaction is described. The curves of rate against intensity are S-shaped and influenced by the previous treatment of the bacteria. Using light of \ 852 with \ 894 from a caesium tube the quantum yield of a thin suspension was measured. About four quanta are needed per carbon dioxide mole- cule reduced, thus pointing to a similarity with green plant photosynthesis in spite of the fact that the net result of the assimilation does not involve an increase of bound energy. Some effects of ultraviolet radiation on bacteria. Alexander Hollaender. (Much of this material has already been published—see Jour. Gen. Physiol., May 20, 1936 and Proc. Soc. Exper. Biol. and Med., 1935, vol. 33, and a part of the data will be published in collabora- tion with Dr. Duggar shortly. A monograph containing some phases of the work is in print.) Myelination in the central nervous system of the albino rat, treated with thymus extract. (Microscopic demonstration.) Albert C. Buckley. AvucGustT 25 Cellular behavior in abnormal growths produced by irradiation of grass- hopper embryos. E. Eleanor Carothers. Embryos of Melanoplus differentialis, ranging in age from five to nine days, received either 250 of 3007 units in a single exposure to X-rays. The treated eggs and controls from the same lot were kept at 25° C. and embryos from each group were fixed daily until the eggs were well into diapause. Roughly 50 per cent of the embryos from the treated eggs show abnormal growths which are duplications and misplacements of normal parts. The majority of these structures occur in the head region. Sections of embryos 24 hours after irradiation show cells with pycnotic nuclei scattered in groups throughout the germ band. These pycnotic nuclei are formed by the coalescing of the chromatin at the end of the anaphase to form a homogeneous, basiphilic globule. Such cells are either phagocytized by other tissue cells, absorbed in situ or cast out of the embryo. Forty-eight hours after irradiation some metaphases show fragmented and lagging chromosomes. The most striking feature at this stage is another type of degeneration. The cytoplasm cytolyzes beginning at the periphery. The spongioplasm remains resistant for a time. The nuclear membrane disappears but the chromatin remains normal in appearance. Gradually it too goes into solution without clumping. By the fifth day after irradi- ation cytolytic lesions have been formed. The cytolysis may be due to (1) injury to the cytoplasm by the X-rays (2) failure or destruction of some other group of cells upon which the cytolyzed cells are normally dependent (3) the phagocytizing of pycnotic cells which transmit a toxic effect. Conclusion: The abnormal growths seem to be the product of groups of normal, healthy cells which either by cytolytic PRESENTED AT MARINE BIOLOGICAL LABORATORY 395 or pycnotic lesions are separated from their normal association and proceed to form as much of a new embryo as their cells retain the potentiality to develop. The original pycnosis is clearly traceable to definite stages of mitosis at the time of treatment. The behavior of the cell surface during cleavage. Katsuma Dan, T. Yanagita and M. Sugiyama. To find out whether the increase in cell surface which must accompany cleavage occurs uniformly over the egg surface or whether it is different in different regions, particles of kaolin were stuck to the naked eggs of the sand dollar Astriclypeus. During subsequent segmentation the positions of these particles were recorded by camera lucida outline drawings and photomicrographs. The data obtained by this method represent linear expansion. As elongation of the cleaving egg begins, the surface near the spindle poles expands, reaches a maximum (ca. 20 per cent increase) about halfway through cleavage, then shrinks somewhat, arriving at a stable value about 5 per cent greater than the initial reading. On the other hand, measurements taken simultaneously in the region of the cleavage furrow show first a slight shrinkage (ca. 20 per cent), then a very large increase (ca. 150 per cent) occurring later than that at the poles (at completion of cleavage furrow) and finally a shrinkage to a stable value which is again much higher than that at the poles (ca. 80 per cent increase over initial reading). The egg surface between these extremes shows intermediate behavior both in magni- tude of the maximum increase and retardation of its appearance. Modified sexual photoperiodicity in ferrets, raccoons and quai.* T. H. Bissonnette. In the work with raccoons (Procyon lotor (L.)), about fifty controls were used of both sexes, whose behavior and times of pseudo-hibernation and breeding were normal. Three groups (each of one o’, one breeding 9, previously successful, and one unsuccessful) were lighted by 60-watt bulbs from October 10, one hour per night for a week, increased by an hour each ten days to eight hours per night from December 10 onward. All had usual care and identical climatic conditions outdoors with temperature from above freezing to — 20° F., and deep snow occasionally. Some Q Q were removed from lighted cages when they became pugnacious toward the end of gestation or pseudo-pregnant periods; others were not; but without affecting pregnancies or suckling. Lighted animals did not decrease food consumption nor undergo pseudo-hiber- nation with onset of cold weather in December and early January as did controls. All females in lighted pens mated from December 16, 19, and 23, respectively, for ten or more days; controls only after February 1 or even April 15 (38-40 days later) with first litters on April 6 and 7 as in the previous year. The three previously sterile 9 2 again failed to have litters; but the productive ones gave litters of four on February 27, one on March 4, and three on March 10, respectively (exact numbers of preceding ‘‘normal”’ season), and suckled and weaned them through sub-zero weather. The former, returned to lighting and males on March 20, again mated without conceiving; of the latter, returned to lighting and males on May 18 and 19, after suckling, two mated from May 24 onward, the other probably did so unobserved; but no young resulted from these second matings. Second mating periods did not occur with controls nor in these @ @ in their previous ‘normal’ season. Male and female cycles synchronized. * The data on ferrets will appear in the American Naturalist shortly. The data on quail will appear in the October number of Bird Banding and need not be gone into here. 396 PROGRAM AND: ABSTRACTS OF SCIENTIFIC PAPERS Asexual reproduction in Dodecaceria fimbriatus. Earl A. Martin. Dodecaceria fimbriatus reproduces sexually by pelagic epitokes and asexually by autotomizing single segments in the middle region of the body. This worm lives in a flask-shaped burrow in which it is folded on itself so that the dorsal surfaces are apposed. The fold is located immediately posterior to the stomach in the intestinal region. When a worm is removed from its burrow in the calcareous materials in which it lives, and covered with loose sand or powdered limestone, it folds on itself and forms a burrow similar to those in the normal habitat. This shows that the folding is a fundamental characteristic which must be considered in relation to the morphology and physiology of the worm. When considered from this point of view, the anterior limb of the fold consists of the feeding and digestive mechanisms, while the folded region and the posterior limb, in addition to conveying the faeces to the anus, serves primarily for the storage of reserve food in the ccelome. Single segments are autotomized in this part of the body. : All autotomized units of the worm regenerate. The anterior part regenerates a new posterior end and the posterior part regenerates a new anterior end. A single segment regenerates from eight to twelve segments anteriorly and about thirteen posteriorly. From this stage growth may continue until a complete worm is formed or the anterior and posterior ends may autotomize at the junction with the original segment each of these autotomized parts then forming complete worms. The physiology of the stomatogastric system in Arenicola marina. G. P. Wells.* A series of lugworms (Arenicola marina) were watched in U-tubes of sea water. The majority showed an activity cycle, periods of rhythmic head movements or of rhythmic extrusion and withdrawal of the proboscis alternating with periods of rest. The duration of a single cycle is somewhat variable, but is commonly six or seven minutes. Experiments on worms dissected in various ways show that the pacemaker determining this phenomenon is a diffuse structure, probably a nerve-net, in the wall of the cesophagus. From this structure, periodic outbursts of excitation travel forward by way of the proboscis to the central nervous system, evoking outbursts of rhythmic activity in the muscles of the proboscis and anterior body wall. Ad- renaline excites the oesophageal pacemaker, the rhythmic discharge becoming continuous instead of intermittent. Injected into intact worms, adrenaline causes prolonged, continuous burrowing. The work was done in England, at Plymouth and London, and a detailed report will appear shortly. GENERAL SCIENTIFIC MEETING AvuGUST 27 The effect of lack of oxygen on the permeability of the egg of Arbacia punctulata. F. R. Hunter and E. N. Harvey. Fertilized and unfertilized eggs were studied in the presence and absence of oxygen under the following conditions: (1) when placed in hypotonic sea water, (2) in hypertonic sea water, (3) in a 0.475 molar ethylene glycol solution in 50 per cent sea water, and (4) in a 0.5 molar ethylene glycol solution in 100 per cent sea water. Eggs which were shrunk in 200 per cent sea water under anaerobic conditions showed a peculiar cortical change. Under the other conditions studied, lack of oxygen had no effect on the permeability of these eggs. These results are in conflict with the slight effect of anaerobic conditions on permeability which had previously * University College, London. PRESENTED AT MARINE BIOLOGICAL LABORATORY SH been reported by Keckwick and Harvey (Jour. Cell. and Comp. Phystol., 5: 43-51, 1934) and Hunter and Harvey (Biov. BULL., 69: 344, 1935). Jt is suggested that the previous conclusions were due to incorrect interpretation of the data. A photronic cell was used in measuring the volume changes in contrast to direct measurements of eggs as had been made during the preceding summer. In this way several hundred thousand eggs could be measured at one time, whereas in the previous experiments only a few cells were measured. Comparative permeability to water and certain solutes of the egg cells of three marine invertebrates (Arbacia, Cumingia and Chetopterus). Beleuckew ik. Riccavand) Ke Elarthine: Permeability of the cell to water is defined as the amount of water that passes per minute through one square micron of cell surface with a difference in osmotic pressure of one atmosphere. This quantity has been determined for the egg cells of the mollusk Cumingia tellenoides, the annelid Chaetopterus pergamentaceus and the echinoderm Arbacia punctulata by measuring the volume changes of these cells in hypotonic sea water by means of a diffraction method. The course of osmotic swelling of the three kinds of cells is found to be satis- factorily described by the equation dV ile KES PB — Pe) where dV/di is the rate of volume change in cubic micra per minute, S’ the cell surface, P — P., the difference in osmotic pressure between cell and medium at time, t, and K is defined as the permeability of the cell to water. For Arbacia eggs the values of permeability range around 0.1 at 22° C.; the eggs of Cumingia and of Chetopterus have much higher values which range from 0.4 to 0.5 cubic micra per minute, per square micron of cell surface, per atmosphere. Permeability to the rapidly penetrating solute ethylene glycol and to the slowly penetrating solute glycerol is investigated by means of a method devised by Jacobs and Stewart. The cells are placed in a medium made hypertonic by dissolving a quantity of the penetrating substance in sea water. In this hypertonic solution the cells at first shrink, but as the solute becomes distributed between the cells and the medium they gradually return to their original size. From the course of such volume changes permeability to the solute is computed. The values are expressed as the number of mols times 10~ which pass per minute through one square micron of cell surface with a concentration difference of one mol per liter. Values of permeability to ethylene glycol are found to average approximately 4.0 for Arbacia eggs; for Cumingia and for Chetopterus cells the values are approxi- mately 16.0at 22°C. Glycerol penetrates very slowly into Arbacia eggs; the average value for permeability is 0.03; Chetopterus eggs have a much higher value which ranges around 6.0. It is concluded that the egg cells of Cumingia and of Chetopterus are several times more permeable to water, and to ethylene glycol and glycerol than are the eggs of Arbacia. Permeability of Ameba proteus to ions. S. A. Corson. A kinetic method of studying surface forces in the egg of Arbacia. F. J. M. Sichel and A. C. Burton. Estimates of the surface forces of cells have in the past been made by static methods. The procedure used in the present study is a kinetic one and therefore of value although perhaps less accurate. 398 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS The measurements were made from a motion picture taken by Dr. Robert Chambers for the purpose of studying the nature of cleavage in the Arbacia eggs from which all extraneous membranes had been removed after fertilization. The pictures were made available to us through the courtesy of Dr. Chambers to investi- gate the part of the problem under discussion. If, just prior to the completion of the first cleavage, one of the blastomeres be ruptured with a microneedle, its contents are rapidly discharged into the surrounding sea water. If this be done at the proper stage of development, the contents of the remaining blastomere will be discharged through the connecting stalk. This dis- charge is due to an excess internal pressure which is in equilibrium with the surface forces. The rate of discharge follows a-law which can be explained by elastic forces rather than by surface tension. Assuming Poiseuille’s law and Heilbrunn’s value for the viscosity at the stage of development under consideration, we arrive at maximum values of 62 dynes per cm2. for the excess internal pressure and 0.09 dynes per cm. for the tangential surface forces. These are of the same order of magnitude as the determinations of Harvey and of Cole for the unfertilized egg. Experimental studies on the oil wetting property of the plasma membrane. R. Chambers. An oil droplet when applied to the surface of the protoplasm of various marine eggs in sea water either adheres to the egg, the surface of contact of the oil becoming flattened, or slips through the protoplasmic surface and occupies a position in the interior with a flattened surface of contact on the inner side of the protoplasmic surface. At least two conditions are necessary for this to occur: (1) the surface of the egg protoplasm must be freed of extraneous coatings which harden in the presence of calcium, (2) the surface of the oil on being brought into contact with the egg, must be expanding in order to be clean. ‘The animal material used were exovates of unfertilized starfish eggs and intact sea urchin (Arbacia) eggs carefully shaken and rinsed in calcium-free sea water and then returned to sea water. Olive oil was used. The results obtained indicate that an increase in acidity of the medium to pH 4.0 greatly enhances the slipping of the oil into the protoplasm while an increase in alkalinity to pH 9.5 decreases the effect. Interfacial films between oil and cytoplasm. M. J. Kopac. Minute drops (diameter, 10 to 20 ~) of percomorph liver oil injected into im- mature odcytes of Asterias.remain perfectly spherical. This condition holds for any animal, vegetal or mineral oil or non-toxic water-immiscible liquid. The injected oil drop may be readily deformed by compressing or stretching the cell, and on releasing the pressure quickly returns to a spherical shape. Upon cytolysis of the odcyte, the oil drop becomes heavily wrinkled and considerably deformed. This wrinkling occurs in odcytes immersed in acidified sea water (pH = 4), in sea water, or in isotonic CaCl, alkalinized to pH > 9.5. In isotonic CaCl, acidified to pH < 3.5, wrinkling occurs but never heavily enough to produce a violent deformation of the drop. The wrinkles on the surface and the collapse of the oil drop are due to the formation of a plastic solid film at the interface (cf. inter alia, Wilson and Fries. Coll. Symp. Monog., 1: 145 (1923)). In order to get wrinkling and deformation of the oil drop, the odcyte must be cytolyzed within 2 minutes after the oil is injected. If an oil drop is kept inside an intact odcyte for 5 to 10 minutes, no wrinkling or deformation will occur after the cell is cytolyzed. Apparently the surface of the oil has changed sufficiently to prevent the formation of a plastic solid film after exposure to the cytolytic products of the cell. The wrinkled oil drop may be readily impaled with a microneedle and removed from the cytolyzed area. The drop behaves as a plastic solid and may be cut into several fragments. Some of these fragments PRESENTED AT MARINE BIOLOGICAL LABORATORY 399 remain irregular while others become spherical as in ordinary oil drops. This behavior indicates that the interfacial film is a plastic solid which remains in place and does not affect the interior of the oil drop. The question of recovery from X-ray effects in Arbacia sperm. P.S. Henshaw. In our early experiments it was found that when mature Arbacia punctulata eggs are exposed to X-rays before fertilization the onset of cleavage following ferti- lization is delayed, the amount of delay varying with two known factors. The first of these is the quantity of radiation administered and the second is the amount of time allowed between treatment and the moment of insemination. The irradiation effect (i.e. cleavage delay) is increased as the amount of radiation applied is increased and reduced as the time between treatment and insemination is increased. The reduction or loss of effect may be regarded as a form of recovery, that is, as observed by determining cleavage time. The recovery process, whatever its nature, is in progress as soon as any irradiation effect is produced (even while treatment is going on), and the rate at which it takes place following treatment is practically exponential with time. More recently it has been shown that cleavage delay may be produced by irradiating sperm alone and that the effect is to some extent accumulative when both gametes are treated. In practically all cases, however, the effect for equal exposures was found to be greater on the sperm than on the eggs. During the present season, a series of preliminary experiments have been carried out to determine whether recovery takes place in sperm the same as in eggs, but without exception none has been observed. ‘Thus, since the sperm cell is composed largely of nuclear material, these experiments suggest that the X-ray effect was produced mainly in the nucleus and that the cytoplasm is involved in the recovery process. Lack of recovery in the sperm will probably account for its greater sus- ceptibility. The respiratory effects exerted by certain organic compounds in relation to their molecular structure. Anna K. Keltch, G. H. A. Clowes and M. E. Krahl. Previous work by the present authors has shown that nitro and dinitrophenols can, in extremely low concentrations, produce a large increase in oxygen consumption and a complete and reversible block to the division of marine eggs. Phenols con- taining nitroso or amino groups in place of the nitro lacked these effects. Nitro’ and dinitrobenzene derivatives having the OH replaced by NHz, Cl, Br, I, CH2OH, NR2, OR, COOH, or CONH: also lacked these effects. From such experiments it was tentatively concluded that the effects of the nitrophenols on cell respiration and cell division were not due to a reversible oxidation or reduction of the nitro groups. The present experiments provide an extension and confirmation of this view. Phenols containing two or three halogen substituents in the benzene ring produced respiratory stimulation and division block in fertilized Arbacia eggs in the same way and to approximately the same degree as paranitrophenol, dinitrocresol and 2, 4 dinitrophenol. Phenols having only one halogen substituent and no other groups were inactive. The isomeric hydroxy benzoic acids, the isomeric hydroxy benzaldehydes, and phenols having one aldehyde and one halogen substituent were inactive. To produce a substantial respiratory stimulation and an accompanying reversible division block in marine eggs it seems necessary to have a phenol hydroxyl group, and a suitable number and arrangement of nitro groups or halogen groups or both. Even replacement of the H of the hydroxyl group by either CH; or C:H; rendered the phenolic compounds inactive. 400 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS Action of metabolic stimulants and depressants on cell division at varying carbon dioxide tensions. M. E. Krahl, G. H. A. Clowes and J. F. Taylor. From previous experiments it was not known whether the division block produced in marine eggs by dinitrophenols was attributable to the phenols themselves or to secondary effects of metabolic intermediates or end products, such as an excessive amount of lactic acid produced aerobically. The following new evidence demonstrates that abnormally large amounts of acidic metabolic products cannot be responsible for the adverse division effects of phenol derivatives: (1) Phenosafranine, an inhibitor of the Pasteur reaction, produced no inhibition of cell division. (2) Penetration of high concentrations of butyric acid produced no division block comparable to that with the phenols. Likewise, concen- trations of CO, up to 2 per cent produced no marked inhibition of cell division. When either of these penetrating acids was employed in sublethal doses inhibition of division appeared to be equally intense at all points in the mitotic cycle whereas the nitrophenols were more effective before onset of prophase. (3) Division blocking concentrations of the phenol derivatives did not produce an abnormally high aerobic glycolysis in Arbacia eggs. The division block produced in Arbacia eggs by a fixed concentration of dinitro- cresol was dependent on egg concentration, a greater blocking being produced as the number of eggs per unit volume was increased. This same dependence on egg concentration was found for the effects of various nitro and halophenols on Arbacia and for the effect of dinitrocresol on Chaetopterus. These effects are just the reverse of those observed when anesthetics are employed to block division. The action of ‘high egg concentrations with nitro and halophenols was proved to be due to the accumulation of respiratory carbon dioxide. Experiments with added CO, showed: (1) A CO, tension, which alone had a relatively slight adverse effect on division, intensified the dinitrocresol division block and the magnitude of this effect was more than the sum of the separate effects for the individual reagents. (2) Concentrations of CO, from .5 per cent to 2 per cent intensified the respiratory stimulating effects of sub-optimum concentrations of dinitrocresol. (3) In a similar way, carbon dioxide intensified the division block and accentuated the respiratory stimulation obtainable with sub-optimum concen trations of paranitrophenol, 2, 4 dinitrophenol and 2, 4 dichlorophenol. Further studies on the effect of numbers present on the rate of See? mn Arbacia. W. C. Allee and Gertrude Evans. Paradoxical osmotic volume changes in erythrocytes. A. K. Parpart and M. H. Jacobs. Earlier work by the authors (BIOLOGICAL BULLETIN, 65: 513, 1933) indicated that erythrocytes exposed to solutions of non-penetrating non-electrolytes undergo a reversible decrease in their internal osmotic pressure through an exchange of anions from the cell for OH’ ions from the solution, with a consequent increase in the proportion of the total base bound by hemoglobin. Direct evidence, however, was lacking that under these conditions reversible volume changes of the cells occur. This evidence has now been supplied, partly by hematokrit measurements and more clearly by the photoelectric method described elsewhere by one of the authors (Parpart). Appropriate experiments show that in solutions of sucrose which from their freezing points would be expected to be isotonic or even slightly hypotonic to the cells there is a rapid decrease in the volume of beef and rat erythrocytes. The addition of small quantities of salts to suspensions of erythrocytes in such sucrose solutions at first causes a further shrinkage of the cells due to the increased osmotic PRESENTED AT MARINE BIOLOGICAL LABORATORY 401 pressure of the solution, but unless the electrolyte concentration be too great, this shrinkage within a few seconds is followed by a considerable swelling. The optimum concentration of NaCl for producing swelling is about 0.005 M, which is approxi- mately the same as that already found in unpublished experiments to cause the greatest increase in osmotic hemolysis in hypotonic sucrose solutions. Further studies on specific physiological properties of erythrocytes. M. H. Jacobs, H. N. Glassman and A. K. Parpart. It was previously observed by one of the authors (Parpart) that the erythrocytes of the ground-hog are remarkably permeable to certain non-electrolytes of relatively high molecular volume such as erythritol, mannitol, xylose, glucose, thiourea, etc. These observations have been confirmed by a further study of the blood of 3 more animals of the same species, and additional determinations have also been made of the permeability of the erythrocytes of the ground-hog and of 6 other species of mammals to a series of 15 ammonium salts. In general, the erythrocytes of the ground-hog have been found to be no more permeable to the ammonium salts studied than are those of the other species, except in the case of ammonium borate to which they seem to show a unique permeability. Experiments designed to give evidence by osmotic means of leakage of salts from the erythrocytes also failed to show a more rapid escape from those of the ground-hog than from those of several other species. Evidently, therefore, a very high permeability to non-electrolytes is not necessarily associated with a high permeability to electrolytes. In the case of the blood of the rat, two apparently specific characters, not found with the other species studied, have been investigated. They are (a) the complete failure of ammonium benzoate to retard hemolysis by ammonium chloride and (b) a very high resistance to hemolysis by distilled water after a previous exposure to a hypertonic salt solution. The blood of the skunk (3 individuals) has also been investigated. In addition to certain other less striking specific properties, it apparently differs from that of the other species studied in undergoing rapid hemolysis in a 4: 1 mixture of M/6 NaCl and M/6 NaHCO. AUGUST 28 Experiments on the contractile substance of muscle fibers. C. C. Speidel. Retraction caps of injury may be induced readily in the striated muscle fibers of living frog tadpoles, if these are orientated properly and subjected to electrical stimulation of suitable strength. A primary retraction cap appears first at the muscle-tendon junction. It is a coagulation product, or muscle clot, which involves a few or many sarcomeres of the contractile substance. It quickly stabilizes the injured end of the fiber, prevents loss of tension, and thus makes for the preservation of the remainder of the fiber. Further electrical treatment usually causes the contractile substance to separate from the retraction cap. As it draws back a second cap forms. [If the fiber is long enough, third and fourth caps may arise in similar fashion until ultimately the entire fiber has undergone complete retraction. In the case of long fibers, such as those teased from the gastrocnemius muscle of the adult frog, multiple retraction caps (a dozen or more) may be induced successively. If, however, muscle fibers in the tadpole’s tail are orientated with their long axes at right angles to the direction of the current, retraction caps do not form. Retraction caps may be induced on young regenerating fibers any time after the differentiation of cross strie, but not before. Retraction caps do not form in either smooth muscle or heart muscle. A wide variety of methods (and reagents) has been employed in investigating the response of the contractile material, which supplements the data from electrical 402 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS stimulation. Many of the phenomena have been recorded by means of ciné-photo- micrography. The injection of aqueous solutions, including acetylcholine, into the isolated muscle fiber. Elsa M. Keil and F. J. M. Sichel. Very small amounts (5,000 u*) of distilled water, Ringer’s solution, isotonic sucrose or sodium chloride solutions, when injected into, or blown upon the surface of the isolated single muscle fiber (adductor muscles of frog’s leg), produce no visible effects apart from a purely mechanical passive disturbance. This has been found to be true only if a clean, previously unused, or freshly washed micropipette is employed. If a larger quantity is injected or if the injection is done too rapidly, or if a previously used pipette is employed, a localized contraction of the interior of the fiber occurs, which is immediately reversible on stopping the injection. Co-: incident with the contraction the fibrillar structure becomes sharply visible at the tip of the micropipette. The contraction is n eeeo miami by the diminishing of the inter-striz distances. The application or injection of CaCl, as pointed out by Chambers and Hale, causes a violent shortening of the fibrillar material. This shortening is reversible and occurs even when minimal amounts of M/200 solution are injected. The application or injection of acetylcholine in concentrations ranging from 1 in 103 to 1 in 10" has no effect on the isolated fiber, even when eserinized, beyond the effects described above attributable to distilled water, Ringer’s solution, etc. The fiber preparation is capable of giving a twitch response to electrical stimu- lation although it is incapable of giving a conducted response, and is apparently denervated. Effects of acetylcholine in concentrations within the range here used have been reported in perfused striated muscles. We conclude that such action is not upon the contractile mechanism directly. The double refraction of smooth muscle. E. Bozler. The double refraction of smooth muscle and its changes during contraction are studied by a new method. By means of a mica wedge in front of the analyser, fringes are produced, the position of which indicates the double refraction of any part of the muscle. In some preliminary experiments it was found that, on stretch- ing, the double refraction of the muscle (retractor of the snail Helix pomatia) increases linearly with the square root of the length. This change cannot be regarded as a photo-elastic effect because no tension is present within the range of lengths used. The effect produced by increasing the length is probably due to the orientation of molecules previously unoriented. It was furthermore found that during an iso- metric contraction there is no change of double refraction within the limits of the accuracy of the method (1-2 per cent). The contrary result obtained for smooth muscle may be explained by the fact that the contractile segments of the myofibrils are able to shorten even in an isometric contraction. There is no evidence that production of tension in muscle is associated with a change of double refraction. Some physical and chemical properties of the axis cylinder of the giant axons of the squid, Loligo pealiit. F. O. Schmitt, R. S. Bear and eZ Vounise Because of their large size the giant axons of the squid are well suited for the investigation of the physical and chemical properties of typical axis cylinder. The birefringence is, as usual, positive with respect to the length of the fiber, in this case being 5 X 107%. In alcohol and certain other reagents, the axis cylinder shrinks from the surrounding sheaths to less than 50 per cent of its original diameter and PRESENTED AT MARINE BIOLOGICAL LABORATORY 403 may be withdrawn as a single strand retaining considerable birefringence. By using the immersion technique this double refraction was shown to be largely of form character, though a considerable amount of intrinsic micellar contributions is present. The axis cylinder proteins exhibit striking shrinkage and swelling changes of an osmotic type. These properties may be observed even after fixation for hours in formalin and must be considered in connection with histological methods. By extruding the axis cylinder material into small volumes of water or salt solution it is possible to investigate the protein components of pure axoplasm. It was found that the principal protein of this material corresponds to that extracted by neutral solutions from lobster claw nerves by Schmitt and Bear. The extrudate dissolves almost completely in distilled water, but the solubility is markedly increased by the presence of salts. Though in this respect the protein resembles the globulins, the fact that addition of acetic acid produces a flocculent precipitate, which is again soluble in dilute alkali, suggests instead that it is a nucleoprotein. Certain other chemical evidence obtained in connection with attempts to prepare nucleotides from this protein offer confirmation, but complete proof awaits experiments to be carried out with the more abundant lobster nerve material. The structure of the eye of Pecten. GG. Schoepfle and J. Z. Young. To assist in Dr. Hartline’s investigation of the action potentials in the optic nerve of Pecten irradians we have re-examined the structure of the eye, using various general histological and special neurological techniques. As has been shown by Dakin, Kiipfer and other workers the retina contains two distinct layers of sensory cells, each giving rise to a nerve. The two nerves run separately for some distance before uniting behind the eye. In Golgi preparations it is easy to see that the processes of the elongated cells of the proximal layer pass into the nerve without the intervention of any synapse. In the case of the more rounded cells in the distal layer the connections are less easy to make out, but in favorable cases it can be seen that axons arising from the cells pass directly into the nerve which arises from this part of the retina. While it is possible that there are other types of cell present which have not been stained in our Golgi preparations, yet since we have never seen any indication of them it seems probable that both the proximal and the distal layers consist only of primary sense cells and interstitial cells, and that there are no other nerve cells, and no synapses in the eye. A further point of interest is that the fibers of the common optic nerve, formed by fusion of the branches arising from the two layers, run across the ring nerve which runs around the edge of the mantle, and pass directly into the radial pallial nerves. The optic nerve bundle sometimes runs with the ring nerve for a short distance but certainly the great majority, and probably all of its fibers pass on into one of the pallial nerves without passing around the mantle for any great distance. The discharge of impulses 1n the optic nerve fibers of the eye of Pecten irradians. H. K. Hartline. The discharge of impulses in the optic nerve fibres of the eye of the scallop, Pecten irradians, has been studied by recording their amplified action potentials. In the whole optic nerve the response to illuminating the eye is strongest at the beginning of illumination, diminishes distinctly after several seconds, but nevertheless continues as long as the light shines. When the light is turned off there is another strong outburst of nerve impulses, lasting several seconds. It has been possible to study separately the responses of the two different layers of sensory cells in the Pecten eye, since the fibres from each layer form distinct branches of the optic nerve. 404 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS Impulses are discharged in the fibres from the proximal sensory cells only when the eye is illuminated. They cease when the light is turned off. Records from single fibres show a regular series of impulses beginning at a high frequency and adapting fairly rapidly to a level which is maintained as long as the light shines. The level of frequency is higher the greater the intensity of illumination. Impulses are discharged in the fibres from the distal sensory cells in response to cessation of illumination. This discharge may last several seconds, or even minutes; it is abruptly stopped if the eye be re-illuminated. Records from single fibres show that the frequency of the discharge is initially high, but diminishes rapidly. The level of frequency and the duration of the discharge are greater the more intense and more prolonged the preceding illumination. In some cases, following strong illumina- tion, it has been observed that the discharge in these fibres may break up into - rhythmic bursts of impulses, the bursts occurring synchronously in all the fibres of this branch of the optic nerve. The effect of light on the CO-poisoned embryonic Fundulus heart. K. C. Fisher and J. A. Cameron. When embryos of trout, salmon, or Fundulus are subjected to mixtures of CO and QO, dissolved in water the frequency of the heart falls below the normal value by an amount depending on the ratio of the concentrations of O, and CO, and the inhibition is relieved by visible light to an extent depending on the intensity of the illumination. The behavior found suggests that in the presence of CO the frequency of the heart depends on the ratio of the CO-free and CO-combined fractions of Warburg’s “‘atmungsferment.” Light causes the dissociation of the CO-ferment complex, more Op, is then able to pass through the system and the beat frequency recovers. If this interpretation is correct, it should be possible to replace ‘‘O» consumption” in Warburg’s theoretical development by “‘heart beat frequency.”’ uninhibited frequency [O?] ; ; a = Kk = / h tiall sa edrcq uence [CO] we have partially tested this possibility using Fundulus embryos, and, as required by the equation, find that at [Oz | uninhibited . : : : 5 CO} the cahibited 1°? function of J, the light intensity. The slopes of Using the formula “uninhibited ‘ } a ' Z : the graphs of aed against J lines’”’ vary in the expected direction with KO) Cay oe changes in a , though the data are not yet complete enough to be quantitative in this respect. The time course of the frequency change upon illumination of a CO-poisoned preparation is that of a first order process, which process is probably not the decomposition of the CO-ferment compound. We conclude that there is a reactant interposed between the ‘‘atmungsferment”’ and the rhythmic production of stimuli to the heart muscle. Change in the light intensity changes the concentration of this reactant as indicated by the frequency of the heart and it is this reactant whose change in concentration is an apparent unimolecular chemical reaction. Prevention of edema in frog perfusions in the absence of serum proteins. G. Saslow. Drinker has observed that leakage through frog capillaries can be prevented for over 3 hours by the addition of a minimum of 15% of horse serum to a perfusion liquid such as 3% acacia in 0.65% NaCl. The acacia alone is ineffective, even if enough be used to make the oncotic pressure three times that of frog blood. Drinker concludes that serum possesses a peculiar power of restraining capillary leakage. PRESENTED AT MARINE BIOLOGICAL LABORATORY 405 In preliminary experiments on this subject, it was observed that the web of the foot of perfused frogs developed no edema in 5 hours (when the experiments were terminated) when either of the following solutions was used: 1. 5% gum acacia 0.014% KCl 0.75% NaCl P (NaH2PO., Na2HPO,), 10 mg./100 ce. 23% beef red cells (washed twice) pH 7.2 2. The same as 1, except that the acacia concentration was 3%. During perfusion the solutions were continuously stirred with oxygen from a tank. It is thus clear that if oxygenation be adequate, the serum constituents can be replaced completely by a colloidal polysaccharide in such concentration as to give approximately the same viscosity and oncotic pressure to the fluid. That serum possesses any special power of restraining capillary leakage, under normal circulatory conditions, appears quite improbable. These conclusions are in complete accordance with those already published by Amberson, Stanbury and co-workers for mammals. Behavior of frog tadpole epidermal cells during seven successive regenera- tion periods. J. A. Cameron and K. O. Mills. Bullfrog tadpoles about 5 cm. long were kept at high temperature and exposed to alternate twelve-hour periods of darkness and intense visible light. Every twenty-four hours a 5 mm. piece was cut from the tail of each experimental animal. The plane of the cuts was at right angles to the tail-axis. The seventh cut was usually fatal, but a few lived until the ninth day. Histological study showed that after each injury the epidermis covered the exposed wound surface by migration from the region anterior to the cut. Since on each successive day a new and more extensive injury than that of the day preceding stimulated new migration, it might have been expected that mitosis would increase with the severity and duration of the demand for new cells. In no case did the rate of mitosis in the skin of the regenerating animal equal that of the controls or of the first specimen from the experimental animal concerned. The counts included all the skin of the parts cut off and almost all the mitoses found were in lateral, non-migrating cells. The sample for each day showed a definite decline in the mitotic rate as compared with the rate for the pre- ceding day. Two explanations are suggested: one, that each injury induced or contributed to the production of an inhibition of mitosis in the skin which extended at least as far as 5 mm. anterior to the cut and lasted for at least twenty-four hours; the other, that cells lost through migration are not in all cases replaced by mitosis of epidermal cells but by addition and differentiation of traveling cells from the connective tissue of the dermis. Observations on conditions affecting growth of cells and tissues, from microscopic studies on the living animal. E. R. Clark and Eleanor I Clarke A careful microscopic study of the growth and differentiation of new tissue in many double-walled transparent chambers installed in rabbits’ ears, in which the tissue has been subjected to a variety of experimental conditions, has revealed the following. The inflammatory exudate which fills the space within the first 24 hours after installation, consisting chiefly of blood plasma, erythrocytes singly or in clumps, 406 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS fibrin and leucocytes, all in varying proportions, promotes rapid growth of cells from surrounding tissues. If the chamber is subjected to continued insults, so that there are repeated renewals of the inflammatory exudate, new growth continues until the space is completely occupied by new tissue. If, however, the chamber is protected from insults so that no appreciable renewal of the inflammatory exudate occurs but only the normal interchange between blood and outside intervascular substance, the rate of growth gradually diminishes. Thus in a chamber demonstrated at this session, in which the maximum success has been achieved in eliminating insults, such a slowing-down of growth occurred that, in the central part of the chamber, a lake of semi-fluid material filling more than half the thickness of the space has remained for a month uninvaded by fibroblasts or blood-capillaries. It seems obvious that, to induce growth of cells there must be present in their immediate environment growth-promoting conditions which involve both physical and chemical properties. The mere presence of a space filled with fluid or semi-fluid material and undoubtedly containing proteins, is not sufficient to induce growth of fibrous or vascular tissue. Incidentally, the lymphatics which vary in different specimens have been carefully studied in many chambers, and it has been found that the rate of growth of fibroblasts, blood capillaries, etc., shows not the slightest relation to the presence or absence of functional lymphatics. Some nuclear phenomena im the T: richodina (Protozoa, Ciliata, Peri- trichida) from Thyone briareus (Holothuroidea). Laura N. Hunter. The Trichodina occurring in great numbers in the gut of Thyone exists in two forms. Type I inhabits the anterior secretory portion of the gut, type II the posterior absorptive region. Type II individuals are elongated, flattened, possess few pellicular wrinklings and contract seldom. Type I individuals are bag-shaped, possess pellicular annulations as do certain vorticellids and contract frequently. One hundred type I individuals averaged 112 microns long by 75 microns wide as contrasted to 78 by 50 microns in 100 type II individuals. The square root of the average length times width of the corona, a skeletal ring in the adhesive disc, was 51 for 100 type I and 37 for 100 type II individuals. The vegetative macronucleus is elongated and somewhat E-shaped in type I and bean-shaped and compact in type II. In both types it appears vacuolated and heterogeneous,—becoming homogeneous as fission approaches. The resting micronuclei of both forms show striations suggestive of chromosomes. Binary fission occurs similarly in both forms but the early post- fission macronucleus in type I is elongated and bent in contrast to the compact bean-shaped post-fission macronucleus in type II. The micronuclei of both forms divide mitotically, with chromosomes and spindle fibers clearly visible. Association between a large and a small individual also occurs in both types. It is accompanied by ‘macronuclear fragmentation and disintegration and several mitotic divisions of the micronucleus. The exact process is not yet worked out but during the ensuing reorganization the larger individual—in type II at least—contains successively 4 micronuclei and 4 partly granular macronuclear anlagen, then 1 micronucleus and 4 macronuclear anlagen. Cells containing 2 macronuclear anlagen and 1 micro- nucleus presumably result from a cell division which separates the macronuclei while the micronuclei divide mitotically. The 2 macronucleate stage shows macro- nuclei free from chromatic granules. Investigations on determination in the early development of Cerebratulus. S. Hérstadius. In the 16-cell stage the egg of the nemertine Cerebratulus lacteus, which undergoes a spiral cleavage, consists of four layers of each four cells, which layers we may PRESENTED AT MARINE BIOLOGICAL LABORATORY 407 designate as am, Am, vegi, and vego. Vital staining of the animal half shows that this material gives rise to the ectoderm of the Pilidium-larva, inclusive of a great part of the ciliated band. The vegetative half forms the stomach, oesophagus, the inside of the two lappets, and a part of the ciliated band. veg. constitutes the material for the stomach, am; corresponds to the greater, most animal part of the ectoderm probably including a small piece of the ciliated band. Several authors, E. B. Wilson, Zeleny, and Yatsu have found that fragments cleave as fragments, and also, to some extent, shown that they differentiate as frag- ments. This was confirmed in greater detail. Animal halves have apical organ and ciliated band but no archenteron; vegetative halves have no apical organ, but a large archenteron and a small ectoderm with a ciliated field. an + veg; + veg» has no apical organ and the ectoderm is too small. In am + an, + veg, the stomach is missing, but we find under the ciliated band a wall, corresponding to the oesophagus and the inside of the lappets. am isolated develops into a blastula with an apical organ and a minute piece of ciliated tissue (compare above!), whereas vegs isolated dies at a late cleavage stage. amo + veg, differentiates as the middle part of the larva, with a broad ciliated band. Also when fusing am; with veg. the fragments develop as they would have done normally: the larva lacks a large ciliated band and oesophagus. The same holds for an animal half transplanted to a meridional half: each component differentiates in accordance with its prospective significance; they seem not to exert any influence upon each other, as is so markedly the case in the sea-urchin egg. The early cleavage stage of Cerebratulus thus belongs to the mosaic type of development. But in the uncleaved egg both the animal and the vegetative halves can form complete larvae. Preliminary evidence as to a source of the growth and the sex-stimulating hormones in the bullfrog. R. Rugh. Mechanism of hatching in Fundulus heterochtus. P. B. Armstrong. Two factors operate to induce hatching in Fundulus eggs: (1) the lashing move- ments of the tail of the embryo, and (2) an enzyme liberated by the developing embryos. The operation of the first of these can be shown by abolishing body movements with chlorotine or KCI at the time when hatching normally should occur. In the absence of the body movements there is no hatching. The operation of the second factor can be shown by treating the eggs immediately after fertilization with 4 per cent to 5 per cent alcohol for the first two days of development. In eggs so treated, some embryos are found with a symmetric menophthalmia and a long, snout- like and sometimes non-patent mouth. In such embryos the hatching enzyme which is produced within the mouth and pharyngeal cavities does not get out into the perivitelline space where it can act on the membrane, and the body movements alone, which appear unimpaired, cannot rupture the chorionic membrane. The enzyme-bearing cells are seen within both the mouth and pharyngeal cavities. They are unicellular and are full of a granular eosinophilic substance which crowds the nucleus to the periphery of the cell. After hatching, these cells disappear completely. Hermaphroditism in Mollusca. B.H. Grave and J. Smith. A study of fifteen species of mollusks has been made during the past summer by the method of paraffin sections, with the object of ascertaining whether hermaphro- ditism is the rule or the exception among the commonest local species. The selection was made to represent as many families as possible, but those chosen were chiefly lamellibranchs. The Amphineura are represented by Chaetopleura apiculata, and the Gastropoda by two marine and two fresh-water species, Urosalpinx cinereus, Nassa obsoleta, Planorbis trivolvis, and Succinea ovalis, respectively. Our findings in 408 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS the main confirm the generalization by Pelsener (1910) that hermaphroditism reaches its highest expression among the most highly specialized forms, such as the oyster, the scallop, and the shipworm. However, some degree of bisexuality is common, varying from that scarcely detectable, to pronounced hermaphroditism. Some lamellibranchs in which the sexes are quite separate in the adult, have hermaphroditic gonads during the early stages of sex differentiation. Some of the latter are pro- tandric, and some develop from the beginning into males and females, the gonads showing some degree of hermaphroditism. This is especially true of the males. In fact, the testes of young individuals of many species of lamellibranchs contain small and sometimes well developed o6gonia as well as mature spermatozoa. The gonads of Venus mercenaria, for instance, contain well developed o6gonia and sperma- tozoa, at least during the first year’s growth. Loosanoff (1936) describes it as pro- tandric, but the sexes are separate in the adult. Young males of Chaetopleura apiculata contain small odgonia which later disappear. Other individuals of this species are females from the beginning of sex differentiation. The sexes appear to be entirely separate in the adult and practically so in the young. During the breeding season the sexes appear to be separate in the following species, and there is no sign of sex inversion: Yoldia limatula, Anomia simplex, Ensis directus, Mytilus edulis, Modiolus demissus, Mya arenaria, Cumingia tellinoides, Venus mercenaria, and Chaetopleura apiculata. ‘The young stages of these species have not been studied, with the exception of Venus mercenaria and Chetopleura apiculata, as described above. In all these forms the sexes are separate in the adults if not throughout life. Pecten trradians and Levicardium mortont are truly hermaphroditic, having a functional testis and a functional ovary at the same time. Self-fertilization takes place in this species of Pecten as was shown by a series of experiments which resulted in normal development when cross-fertilization was prevented by isolation. Self- fertilization may take place also in Lewcardium, although experiments with this animal were not attempted. There is a change in sex in Teredo navalis from male to female, and vice versa. This species is not truly hermaphroditic in that the gonad does not function as an Ovary and a testis at the same time but alternately. In Planorbis trivoluis and Succinea ovalis an ovatestis is present. The alveoli give rise to both eggs and sperm at the same time. These snails appear to be perma- nently hermaphroditic. Hickman (1931) gives the breeding season of Succinea ovalis as April to September and states that after the hibernating period spermatozoa and eggs develop at the same time, an example of ‘simultaneous hermaphroditism.” In conclusion it may be said that many species of mollusks appear to be dioecious, but when sexual cycles have been studied adequately, especially during early sex differentiation, some classed as dioecious may be shown to be hermaphroditic or protandric. Sex in many species is not fixed but may be altered both by external and internal changes. The indications are that it may be possible to induce changes in sex in mollusks experimentally. PAPERS READ BY TITLE Separation of the conducting and contractile elements in the retractor muscle of Thyone briareus. H. G. duBuy. — The retractors which are attached to the calcareous ring of the lantern of Thyone briareus present a type of muscle which is intermediate between vertebrate striated and vertebrate smooth muscle. The histological characteristics of these retractor muscles and their sensitivity to choline-esters and adrenalin can be compared with properties of vertebrate smooth muscle; their sensitivity to curare and the effect of this drug on the contractile elements can be compared with properties of vertebrate striated muscle. PRESENTED AT MARINE BIOLOGICAL LABORATORY 409 It can be shown that the action potentials obtained from the nerve-muscle preparation of Thyone occur only when the conducting elements are not blocked (for example by curare). After curarisation no action potentials occur while local contractions are still obtainable. These phenomena allow a further differentiation between the properties of the conducting and the contractile elements. After curarisation of a part of the muscle, the interesting phenomenon of a distal contraction of the non-curarised part can be observed. This is similar to the phe- nomenon which was first described by Steiman for the single muscle fibers of the frog’s tongue. When an electric stimulus is applied to the curarised part of the muscle of Thyone, only the non-curarised distal part of the muscle shows a contraction which can be more or less local. After an increase of the intensity of the electric stimulus both the local contraction under the stimulating electrodes in the curarised part, and the distal contraction of the non-curarised part can be obtained. Therefore the threshold for electric stimulation of the conducting elements is markedly lower than the threshold of the contractile elements. The application of curare prevents the occurrence of spontaneous contractures, etc., which often have hampered investigation of the plastic properties of smooth muscles. Curare does not seem to change these properties. Some differences, however, occur in the extension-curve obtained by means of weights. These differ- ences can be explained by the absence of the activity of the conducting elements. Hyperparasitism: A species of Hexamita (Protozoa, Mastigophora) found in the reproductive systems of Deropristis inflata (Trematoda) of marine eels. A. V. Hunninen and R. Wichterman. A species of Hexamita was discovered parasitic in the reproductive organs of the trematode, Deropristis inflata, found in the marine eel, Anguslla chryspa. Of 154 eels examined, 41 were infected with the trematode; 15 of these infected eels harbored 80 trematodes which contained the protozoan parasites. Measurements showed the length of the protozoén to be 10.2 u and the width 5.2 u; the animal being rounded anteriorly and pointed posteriorly. Six flagella originate from the anterior end, being approximately of the same length as the body and directed posteriorly. The two posterior flagella which are approximately twice the length of the body, continue from the axostyles. Two compact club-shaped nuclei are located in the anterior end of the flagellate and extend slightly more than one-third the body length. The flagellates were limited exclusively to the reproductive organs of the trema- tode, especially to the female system. They were found in large numbers in the eggs, uterus, oviduct and seminal receptacle. In a few cases they were observed in the vitelline glands and testes. What seemed to be noteworthy was the continued occurrence of the parasites in varying numbers in the eggs of the trematode. In several instances as many as 20 were found in one egg. The protozoa live at the expense of the protoplasmic material of the egg which would normally give rise to a miracidium. In some cases all the eggs in the uterus were parasitized by Hexamuta, completely preventing miracidial development. The conspicuous absence of Hexa- mita from the intestine of the eels strongly points to a rigid host-parasite specificity between the protozodn and the trematode since flagellates were found in only 2 of the 141 eels examined. Members of the genus Hexamita are of interest because of their widely diversified associations in nature. Some species of the genus are free-living and others associated with vertebrate and invertebrate hosts. This appears to be the first observation of trematodes being parasitized by these flagellates. Further details with figures will be presented in the complete paper. 410 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS Cytological studies of the genus Chilodonella. III. The conjugation of Chilodonella labiata, variation? Mary S. MacDougall. Results of studies of the effects of ultraviolet variation on Chilodonella made an investigation of the cytology of this genus desirable. Previous work has shown that the six species whose conjugation and division phenomena have been studied fall naturally into three groups: 1. The Cucullulus group. 2. The Labiata group. 3. The Uncinata group. Each of these groups has certain definite characteristics peculiar to it, especially as to behavior in conjugation and division. In Stuart’s Pond, a brackish body of water in Woods Hole, Mass., a Chilodonella was recently found which looked more like Stokes’ figures of C. caudata than the fresh-water form of this species. The so-called tail, however, was not constant, and the living material was often confused with C. uncinata. The macronucleus is centrally located, however, as it is in /abiata, and the details of conjugation are identical with that form, i.e. the macronucleus as well as the micronucleus is ex- changed during conjugation. There are several points of difference between the new form and Jabiata which seem to be constant. While it is definitely not the typical /abiata recently described by the author, it is evidently a variation of this type. It has been found that in nature there is a series of closely intergrading forms in this genus, and further genetic work must wait on more complete knowledge of the cytology of this group. Comparative hypotonic cytolysis of several types of invertebrate egg cells and the influence of age. Victor Schechter. The resistance to cytolysis in distilled water was studied for several species of unfertilized invertebrate egg cells. The figures below give the time required for fifty per cent of the cells to cytolyze. ; The egg cells of Mactra solidissima showed a drop from approximately 83 minutes when fresh to 3 minutes when 20 hours old. For Cumingia tellinoides the corre- sponding figures were 8 and 34 minutes. The egg cells of Heteronereis limbata showed a resistance of 63 minutes when fresh and a drop to 4% minutes when 28 hours old. For Echinarachnius and Asterias sufficient quantitative data are not yet available but a similar trend is indicated. Freshly shed Arbacia egg cells cytolyze in 34 minutes. What significance, if any, lies in the initial difference in resistance of the fresh eggs of these various forms is not yet apparent. In 1932 (Goldforb and Schechter, Proc. Soc. Exp. Biol. and Med.) a progressive decrease in resistance to hypotonicity by the ageing egg cells of Arbacia was demonstrated. The extension of this finding to other species weights rate of cytolysis as a factor of some importance in the process of ageing. In view of the fact that hypotonic cytolysis is a complicated phenomenon involving the interior of the egg as an osmotic system and also the physical and chemical properties of the cell surface, it is of special interest that all eggs so far studied show a similar trend with age. Preliminary experimental analysis has been made and will be presented in an early paper. It may be of interest to report at this time that, for Arbacia egg cells, pretreatment with excess calcium increases resistance to hypotonic cytolysis, decreased calcium lowers resistance and similar solutions also affect the duration of life. PRESENTED AT MARINE BIOLOGICAL LABORATORY 411 Notes on life cycles of digenetic trematodes. H.W. Stunkard. Experiments, conducted under carefully controlled conditions, have given the sexually mature stages of two species of larval trematodes from Nassa obsoleta. Both species were studied and named by Miller and Northup (1926). One species, Cercaria quissetensis, occurs in about one per cent of the snails examined. The cercariae swim about near the bottom and encyst in various species of mollusks. Experimental infection of Mya arenaria, Mytilus edulis, Modiola modiolus, Cumingia tellinoides, Pecten irradians, Ensis americana and Crepidula fornicata was obtained. Metacercarize recovered three days after encystment were infective for gulls and developed into adults of the species identified by Linton (1928) as Himasthla elongata (Mehlis). Constant differences between these specimens and the description of H. elongata were observed by Linton and it appears that the worms belong to a distinct species, Himasthla quissetensis, first named in the larval stage by Miller and Northup. The specimens found in natural infections are identical with those obtained in experimental ones. Cercarieum lintont of Miller and Northup is identical with Distomum lastum Leidy, 1891. These larvae encyst in the parapodia of Nerets virens and other annelids. After three weeks in the second intermediate host, the metacercariz were fed to various fishes and developed to sexual maturity in the intestine of the eel, Anguilla chrysypa, and the toadfish, Opsanus tau. The worms persisted for two to three weeks in the intestine of Paralichthys dentatus and Tautoga onitis, and it is probable that they reach maturity in these species also. Natural infections were found in the eel. The adult worms are specifically identical with Zoogonus rubellus (Olsson, 1868), (synonym, Zoogonus mirus Looss, 1901). Inhibition of gastrulation in Arbacia with Nickel Chloride. A. J. Waterman. For several summers a study has been made of the effect of various physical and chemical agents upon gastrulation in the sea-urchin, Arbacta punctulata. Gastrula- tion in this form is much more resistant to experimental manipulation than in Paracentrotus, while the embryos of the sand-dollar and starfish are particularly sensitive. The chlorides of Na, K, Mg, and Ca alone or in combination, hypotonic and hypertonic sea water, variation in temperature, X-ray and alcohols have been tried. Recently the blastula and young gastrula stages have been exposed for many hours to various concentrations of some of the metallic chlorides. Of these, certain concentrations of a 1 per cent solution of NiCl, in sea water have given numerous, excellent exogastrule which grew and differentiated a fairly typical body form, skeleton, external gut and stomodeum. The types are not quite as good as those obtained previously with Paracentrotus but are the best that have thus far been secured with Arbacia. Nickel chloride seems to have a differential effect upon gastrulation in some individuals but the embryos of a culture vary in their sus- ceptibility. Some show total inhibition of any endoderm formation and grow as large, oval vesicles with differentiated ectoderm, apical plate, stomodeum and irregular apron-like growths around the edge of the former vegetal pole region. Nickel chloride also produces skeletal modifications, as well as inhibition and retarda- tion of development depending upon the concentration. A comparison of the effects upon the blastula and young gastrula stages seems to show little or no differential susceptibility previous to or during gastrulation. Gastrulation may be temporarily inhibited for many hours. The general effects upon development, except for exo- gastrulation, are very similar to those observed with the alcohols and other agents listed above. 412 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS Division and conjugation in Nyctotherus cordiformis (Ehr.) Stein . (Protozoa, Ciliata) with special reference to the nuclear phenomena. R. Wichterman. A cytological study of the complete life cycle of Nyctotherus cordiformis (Ehr.) Stein, a heterotrichous ciliate from the large intestine of tadpole and adult tree toads, Hyla versicolor, was studied. In division, a partial dedifferentiation of parental ingestatory structures occurs, then redifferentiation to be retained by the anterior daughter individual. For the posterior daughter, the ingestatory apparatus arises de novo. The macronucleus divides amitotically while the micronucleus divides mitotically, demonstrating small bead-like chromosomes. Dividing animals may be found in tadpole and adult hosts. However, during metamorphosis of the tadpole into adulthood, divisions of the ciliates result in small forms, the pre-con- jugants, which are the only ones capable of conjugation. Therefore conjugation takes place only during the metamorphosis of the host. In conjugation, two ciliates fuse along their peristomes. The macronucleus undergoes complete fragmentation while the micronucleus of each conjugant divides three times. The first pregamic division results in two micronuclear products; the second division, in four micro- nuclear products, three of which degenerate. The remaining product enters into the third pregamic division to produce the two functional migratory and stationary pronuclei. Interchange of migratory pronuclei follows at the fused anterior ends of the conjugants to form the amphinucleus which divides once to produce the micro- nucleus and the macronuclear anlage. Development and behavior of the unusual macronuclear anlage (“‘spireme ball” of Stein and Schneider) was observed for the first time. Ex-conjugants were found nearly exclusively in recently transformed hosts. While the animals remain fused in conjugation for less than a day, the reorganization process of the ex-conjugants takes several weeks. Toward the end of the conjugation process ingestatory structures of each conjugant completely dedifferentiate while a new set arises de novo posterior to the old ones. Conjugation here is interpreted as an effort made by the ciliates to overcome a physiological crisis during the drastic host transformation changes. Note: The complete paper will appear in the Journal of Morphology. BIOLOGICAL MATERIALS The Supply Department of the Marine Biological Laboratory has a com- plete stock of preserved and living material for the coming school season. We would be pleased to, quote prices and furnish our current catalogue upon request. Our materials are put up by skilled preparators of long experience, and they are guaranteed to give absolute satisfaction. From November to March we will be in a position to furnish Living Marine Aquaria Sets. For the past six years these sets have proven very popular, especially in the schools located away from the seashore, giving the students an opportunity to study.and observe the various living forms. This living material is collected and assembled here at Woods Hole, and we guarantee to deliver the sets in good condition to any part of the country east of the Mississippi if they are shipped during the above- mentioned period. | SUPPLY DEPARTMENT Est. 1890 MARINE BIOLOGICAL LABORATORY WOODS HOLE, MASS. CONTENTS , Page PARKER, G. H. | The Reactivation by Cutting of Severed Melanophore Nerves in the Dogfish |Mustelus. ig i ii CNA nace aie pes) ABRAMOWITZ, A. A. | Physiology of the Seinnotiote System in the Catfish, VANS TITLES ee PA ANAM ree Chg NE mae Sey ee Crk eee 259 SMITH, G. M., AND C. W. COATES Cutaneous Melanosis in Lungfishes (Lepidosirenidae). ye _ 282 COWLES, R. P., AND C. E. BRAMBEL A Study of the Environmental Conditions in a Bog Pond with © Special Reference to the Diurnal Vertical Distribution of Gonyostomum semen......... Ratha Bh, Ae ME ee OO HEILBRUNN, L. V. | | Protein Lipid Binding in Protoplasm...... contain Lash 299 - MAZIA, DANIEL, AND JEAN M. CLARK Free Calcium in the oes of Stimulating Agents on 1 Elodea ZOBELL, CLAUDE E., AND D. QUENTIN ANDERSON Observations on the Multiplication of Bacteria in Different Volumes of Stored Sea Water and the Influence of Oxygen Pension .and. Solid’ Surfaces 7/45 4 aay Be ae el he 324 TRAGER, WILLIAM The Utilization of Solutes by Most Larvae. (iain oe oS Cor, W. R. Environment and Sex in the Oviparous Oyster Ostrea virginica se ae PLEAS MEAN Gi lca MOR elie aha ah ele Doe eae eyes 353 GRAFFLIN, ALLAN L. Renal Function in Marine Teleosts. IV. The excretion of inorganic phosphate in the sculpin..... Per les As Sa Se OO KOONZ, CARL H. Some Unusual Cytological Phenomena in the Boe uatieenecn of a Haploid Parthenogenetic Hymenopteran, peusul smithii (Packard) Pear R APPR at seo el TNO ak Uarrrgceiu ree a ied lee 375 PROGRAM AND ABSTRACTS OF SCIENTIFIC ‘Mertines, MARINE BIOLOGICAL LABORATORY, 1936......:........ Rae Slee ik 386 Number 3 ao i BIOLOGICAL BULLETIN PUBLISHED BY ‘THE MARINE BIOLOGICAL LABORATORY Editorial Board GARY N. CALKINS, Columbia University E. G. CONKLIN, Princeton University FRANK R. LILLIE, University of Chicago E. N. HARVEY, Princeton University CARL R. MOORE, University of Chicago SELIG HECHT, Columbia University GEORGE T. MOORE, Missouri Botanical Garden ' LEIGH HOADLEY, Harvard University T. H. MORGAN, California Institute of Technology M. H. JACOBS, University of Pennsylvania G. H. PARKER, Harvard University H. S. JENNINGS, Johns Hopkins University W. M. WHEELER, Harvard University E. E. Just, Howard University EDMUND B. WILSON, Columbia University ; “ALFRED C. REDFIELD, Harvard University Managing Editor DECEMBER, 1936 Printed and Issued by LANCASTER. PRESS, Inc. PRINCE & LEMON STS. LANCASTER, PA: THE BIOLOGICAL BULLETIN THE BioLoGicaL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50. Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain: Wheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2. Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Mass., between June 1 and October 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the remainder of the year. | INSTRUCTIONS TO AUTHORS Preparation of Manuscript. In addition to the text matter, manuscripts should include a running page head of not. more than thirty-five letters. Footnotes, tables, and legends for figures should be typed on separate sheets. Preparation of Figures. The dimensions of the printed page (496x7 inches) should be borne in mind in preparing figures for publication. Draw- ings and. photographs, as well as any lettering upon them, should be large enough to remain clear and legible upon reduction to page size. Illustrations should be planned. for sufficient reduction to permit legends to be set below them. In so far as possible, explanatory matter should be included in the legends, not lettered on the figures. Statements of magnification should take into account the amount of reduction necessary. Figures will be reproduced as line cuts or halftones. Figures intended for reproduction as line cuts should be drawn in India ink on white paper or blue-lined coordinate paper. Blue ink will not show in reproduction, so that all guide lines, letters, etc. ‘must be in India ink. Figures intended for reproduction as halftone plates should be grouped with as little waste space as possible. Drawings and lettering for halftone plates should be made directly on heavy Bristol board, not pasted on, as the outlines of pasted letters or drawings appear in the reproduction unless removed by an expensive process. Methods of repro- duction not regularly employed by the Biological Bulletin will be used only at the author’s expense. The originals of illustrations will not be returned except by special request. Directions for Mailing. Manuscripts and illustrations should be packed flat between stiff cardboards. Large charts and graphs may be rolled and sent in a mailing tube. Reprints. Authors will be furnished, free of charge, one hundred re- prints without covers. Additional copies may be obtained at cost. Proof. Page proof will be furnished only upon special request. When cross-references are made in the text, the material referred to shouid be marked clearly on the galley proof in order that the proper page numbers may be supplied. Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894. Vol. LXXI, No. 3 December, 1936 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY iit BEER CT OF TALCOHOE,ON (GASTRULATION IN ARBACIA A. J. WATERMAN (From the Thompson Biological Laboratory, Williams College, and the Marine Biological Laboratory, Woods Hole, Mass.) I Following certain changes in the environment, the endoderm of echinoderm embryos (Arbacia, Paracentrotus, Astertas, Echinarachnius) can be made to develop externally as an exogastrula. Other effects include the development of the endoderm at the expense of the ecto- derm and the irregular formation of skeleton and body form. First discovered by Herbst as a lithium effect, exogastrulation has since been shown by several investigators to be caused by a large variety of agents, both chemical and physical (cf. Waterman, 1934). The result is given by fertilized eggs, cleavage stages or blastula stages. Re- cently exogastrulation has been provoked in another group of animals, ‘the Amphibia. Holtfreter (1933) has found that exposure of axolotl eggs to NaCl solutions caused exogastrulation in some of the embryos, while Curtis et al (1936) have provoked exogastrulation by exposing blastulae of Amblystoma, Rana and Bufo to X-ray. As alcohols are said to act as depressive narcotic agents, several have been tried to see if gastrulation could be permanently suppressed by them without interfering markedly with subsequent growth and differentiation. While this was not found to be the case, except in instances of fatal injury to the more susceptible individuals, certain effects were observed which are described below. Blastula and young gastrula stages were exposed for about eighteen hours to various concentrations of methyl, ethyl, propyl, isopropyl, butyl, amyl and capryl alcohols. Since individuals of a culture vary widely in their susceptibility to the alcohols, the results are based on the condition of approximately fifty per cent or more of the embryos as determined by the study of at least three random samples from each culture. In general alcohols act as depressive narcotic agents although they may exhibit exactly opposite effects in different concentrations (Lillie, 413 414 A. J. WATERMAN 1912a). In very low concentrations they may stimulate development (Lillie, 1912a; Mast and Ibara, 1922, on tadpoles; Elhardt, 1930, on chick) while the higher concentrations are toxic. The higher alcohols are more toxic than the lower ones. Alcohol produces cyclopean eyes in Fundulus and also tends to suppress the development and differentiation of the auditory vesicles (Stockard, 1910). Certain concentrations will cause complete but reversible arrest of cleavage (Lillie, 1914). Eggs show a rhythmical susceptibility. The action of alcohol, and other anesthetics, has been variously interpreted (Lillie, 1916; Henderson, 1930; Winterstein, 1926). II Reference to Tables I and II will show the approximate concentra- tions which produced certain effects. The concentrations are ex- TABLE I Influence of alcohols upon gastrulation and subsequent development during eighteen hours of exposure. Blastula stage. Concentration of alcohol expressed in cc. per 100 cc. of sea water Approximate limits of Com- Simple Alcohol Toleration.| Retarda- complete but | plete in- exogas- No effect | tion and Gastrula- reversible hibition All trulz on devel- | inhibitory tion only inhibition of of any dead which did opment effects gastrulation gastru- not dif- in some or all lation ferentiate individuals ; Methyl 0.2-1.5 | 1.5-3.5 | 4 -5.5 5.0-5.9 5.5 6.0 2.5 Ethyl 0.5-0.9 | 1.0-2.0 |* 2.0-3.6 2.7-4.4 3.8 5.0 0.9-1.3 Propyl 0.1-0.5 | 0.5-1.0 | 0.6-1.3 1.4-2.0 1.8 2.0 0.2-0.8 Isopropyl 0.1-1.5 | 1.6-2.0 | 2.0—2.4 2.1-3.1 3.2 3.5 1.0-1.7 Butyl 0.1-0.2 | 0.3 0.7 0.7-1.3 1.0 1.6 0.3-0.4 Amyl 0.1 0.1-0.2 |: 0.3-0.4 ().4-0.6 0.5 0.7 0.1-0.3 Capryl 0.01 0.01—0.02 0.02-0.05| 0.04 0.06 | 0.01—0.02 pressed as cubic centimeters of the alcohol which were added to 100 cc. of sea water containing the embryos. It will be noted that very little if any significant difference exists between the results for the two stages, indicating that they are probably equally sensitive. To avoid repetition only the results secured from the exposure to methyl alcohol will be described. The data for the others are included in the tables. Methyl alcohol—The embryos tolerate long exposure to concen- trations from 0.2 cc.—1.5 cc. with no appreciable effect on gastrula- tion and development. In a 2 cc. concentration, development is slightly retarded but plutei develop. In a 2.5 cc. concentration, EFFECT OF ALCOHOL ON GASTRULATION 415 gastrulation and subsequent development are further retarded, and after 18 hours fewer top-swimming individuals are present. These are small plutei with frequent abnormal skeletal structures and inhibited arm development. A few undifferentiated cases show slight external outpushing of the vegetal pole cells or rupture in this region. Movement is slower. In a concentration of 3 cc. gastrulation is further retarded or even inhibited and fewer stunted young plutei (especially with reference to arm development) are lethargically swimming through the medium. Most of the culture is on the bottom of the container. Individuals in which the vegetal pole cells are unevenly extruded or in which rup- ture, followed by some disintegration of the vegetal pole region, has TABLE II Influence of alcohols upon gastrulation during eighteen hours of exposure Concentration of alcohol expressed in cc. per 100 cc. of sea water Alcohol Approximate limits of i complete but reversible Simple exogastrule | Complete inhibition inhibition of further All dead which did not of any gastrulation in some differentiate gastrulation or all individuals Young Gastrula Methyl 3.9-5.5 5.9 0.4-2.5 a2 Ethyl 2.0-4.0 4.7 0.5-1.5 3.5 Propyl 1.6-1.9 2.1 0.5-1.5 1.8 Isopropyl 2.2-2.6 3.0 0.3-1.4 2.3 Butyl 0.7-0.9 1.3 0.1-0.7 1.0 Amyl 0.4-0.6 (O)s7/ 0.1-0.2 0.5 Capryl 0.02—0.03 0.05 none 0.04 occurred are quite common. The ectoderm of the embryos is ragged but the apical plate is present, even in embryos which do not go beyond the late gastrula stage. In the same length of time the control formed medium-sized plutei. In a 3.5 cc. concentration the embryos develop as far as the tri- angular stage with differentiation of the apical plate, but lack skeleton. In a 4 cc. concentration the embryos only go as far as the late gastrula in 18 hours. A 4.5 cc. concentration gives shallow or early gastrule and their movement is very slow and tends to be rotational, while 5 ce. gives only shallow gastrule at the most, and many blastule. Gastrulation is inhibited for the most part at a concentration of 5.5 cc. although a few show initial gastrulation. There is very little movement. On return to fresh sea water after 18 hours in the solution, 416 A. J. WATERMAN these blastule gastrulate and continue development. In concen- trations from 5.6 cc.—5.9 cc. fewer and fewer blastule survive and these gastrulate in fresh sea water, although frequently in-an abnormal manner. Subsequent differentiation is abnormal and many do not go beyond a late gastrula stage. In a concentration of 6 cc. the embryos do not survive the experimental period. Above a minimal concentration of the alcohols the general effect is a depression of gastrulation and subsequent development which is proportional to the kind of alcohol, the concentration and the length of exposure. Methyl is least depressive and capryl the most, while isopropyl is less depressive than the normal alcohol. Lillie (1914) found the latter to be the case in his study of the action of alcohols in suppressing cleavage in sea-urchin eggs. The different alcohols give quite similar results but within progressively narrower limits for the higher ones. The present series was found by Pantin (1930) to hold for the action of alcohols on the ameboid movement of Ameba and by Lillie (1914) in the suppression of cleavage. Gastrulation may be inhibited by alcohol for at least as long as 18 hours without loss on the part of some embryos of the ability to gastrulate and without appreciably detrimental effect on whatever processes are involved. When returned to fresh sea water, the surviving blastule gastrulate with apparently undiminished vigor and subsequent development may be quite typical. When develop- mental abnormalities appear, they are associated with differentiation, i.e., formation of the skeleton, size of the body, and position and length of the arms. These plutei are less resistant and die sooner than do the control. This temporary inhibition of gastrulation is shown by the following observations of embryos exposed to ethyl alcohol which is only slightly more toxic than methyl. After 5 hours in a 3 cc. concentration the embryos were still in the oval blastula stage. Lethargic movement was visible but all were on the bottom of the container. At this time the controls were at the late gastrula or triangular gastrula. When changed to fresh sea water these developed after 17 hours to the late triangular and early pluteus stages. A few, consisting of the weaker individuals, were at the blastula stage or showed irregular gastrulation with some slight external development of the vegetal pole cells or rupture of the vegetal pole region. After the same length of time in a 5 cc. concentration, no movement was visible and there was considerable evidence of cellular swelling and disintegration especially at the vegetal pole region. Seventeen hours after changing to fresh sea water the em- bryos were at the late triangular gastrula stage with large skeletal HEREC OR ALCOHOL ON GASTRULATION 417 spicules. At this time the control was at the young pluteus stage. Twenty-four hours later the embryos had developed into normally appearing individuals with no skeletal peculiarities as far as were seen. A 5 cc. concentration was fatal during an exposure of 18 hours. If inhibition occurs during gastrulation, it is resumed at the point where it left off and goes on to completion with the removal of the alcohol. During the interim nothing vital to the process has been removed, passed over or destroyed. Under the influence of the alcohol the stage of gastrulation does not express itself as determining the capacity to continue gastrulation in the modified environment. Concentrations which inhibit gastrulation at a late cleavage or blastula stage likewise stop it after it has begun or during the process. Thus no difference in sensitivity was demonstrated. Only death or fatal injury completely inhibited typical gastrulation as tested by the ability of the embryos to undergo further development. Gastrulation is not apparently more susceptible to an anzsthetic like alcohol than other developmental processes. Differentiation without gastrulation was not seen, nor gastrulation without differentia- tion except in fatally injured embryos. Such embryos live for only a short time after transfer to fresh sea water and gastrulation is not typical. These may show an irregular proliferation of cells at the vegetal pole or the blastoccel becomes packed with cells. The few cases of apparent initial exogastrulation which appeared in lower con- centrations and after shorter exposure periods were of this type or showed a very shallow or irregular outpushing of the vegetal pole cells. Only some of the more susceptible individuals in the lower concentra- tions showed this condition. They did not differentiate beyond form- ing an apical plate and a small, triangular body form. Frequently the blastoceel was more or less packed with cells. Abnormal embryos, no matter how produced, always tend to swim in asmall circular path. The more retarded or abnormal the structure, the shorter is the diameter of the circle. Modification of shape would account for this in part, but, where embryos are very retarded though otherwise quite symmetrical, there does occur a modification of ciliary distribution and size. The cilia of the blastula and gastrula stages are least sensitive to the narcotic action of alcohol. In concentrations which inhibited development beyond the stage at which it was put into the solution, the cilia continued to beat but at a slower rate. In concentrations immediately above this value, movement was much reduced but it was stopped completely only with complete death or disintegration of the embryo. In lower concentrations, the rate is unaffected but becomes gradually slower above a minimal concentra- 418 A. J. WATERMAN tion. The change does not parallel developmental inhibition. This difference in the narcotizing concentration for different tissues is well known. Lillie (1912a) found that the cilia of Avenicola larve con- tinued their activity for hours in isotonic sugar or magnesium chloride solutions, ether and chloroform, which completely inhibited any muscular movement. In the present study the coérdination of the ciliary action was not apparently disturbed although the pathways of the embryos tended towards small circles. Dying embryos did show a fluttering effect attributable to disintegration. One of the first of the differentiating tissues of the embryo to be affected by the treatment with alcohol is the mesenchyme and this becomes apparent in the modification of the skeleton. This effect has been observed by Runnstrém (1928) and others (Waterman, 1932, 1934) following the use of LiCl, various other salts, and also physical agents. The modifications first appear in concentrations just above the limit of toleration and are evident in the irregularity and loss of symmetry of the skeletal parts. The ends of the body rods are smooth and unfused. In higher concentrations there appear suppression and reduction of the skeletal rods on one side or both, difference in thickness and amount, accessory parts or projections, and abnormal positions. Alcohol thus markedly affects the distribution and activity of the - skeleton-forming mesenchyme, even when the remainder of the embryo may be quite typical in general appearance for the particular stage. Another modification is that of the fusion of embryos. This varied from pairs up to large aggregations. Only those fusions involving two embryos underwent any considerable differentiation although the members of a large aggregation might gastrulate. The former are comparable to those secured by other means and by other investigators (cf. Waterman, 1932, 1934). Fusion seemed to occur at regions of initial disintegration, and, since this occurred generally in the region of the vegetal pole, many cases failed to gastrulate or did so in a very irregular manner. This region is more susceptible and is the one which seems to be most active physiologically and developmentally at this time (Child, 1928). The swelling and partial separation of the blastomeres in this region point to an injurious permeability change. The final result is an irregular mass of SHEEN spherical cells which gradually disintegrates. That weak solutions of the alcohols may give a slight acceleration of development seems to have been found in the course of this study. This was seen to occur with methyl, ethyl and isopropyl alcohols in the lowest concentrations employed. In the case of the earliest stage EREPECL OF ALCOHOL ON GASTRULATION 419 the embryos for the most part gastrulated sooner, and in all cases the plutei matured more quickly than the control. They did not survive as long as the control. There was, however, no noticeable difference in final size. Others have likewise found that weak solutions of certain alcohols accelerate development. In a study of the effect of ethyl alcohol on tadpoles, Mast and Ibara (1922) found that, although there was a reduction in activity of the tadpoles, there occurred an increase in the rate of growth and a decrease in the rate of mortality. While it is not clear as to how these effects were produced, they suggest that the reduction in activity might account for the longer life and that the alcohol served as food which conserved the reserve food material in the tadpole’s body. The animals were not fed during the experiment. Alcohol may also increase irritability when used in low concentrations, while exhibiting typical anesthetic action in higher concentrations. Reference should be made to Lillie (1912a), page 374, for a discussion of this point and additional examples. More recently Elhardt (1930) has found that young chicks respond by an increased growth rate to doses of ethyl alcohol. Methyl, propyl and butyl alcohols (1932) had a stimulating effect on growth for a time. SUMMARY The general effect of immersing the blastula and young gastrula stages in various concentrations of methyl, ethyl, isopropyl, propyl, butyl, amyl and capryl alcohols for 5-18 hours is a depression of gastrulation and development which is conditioned by the kind of alcohol, the concentration and the length of exposure. They all give quite similar effects but within narrower limits for the higher ones. Apparently the process of gastrulation is no more sensitive to the alcohols than other developmental processes. It may be inhibited for 5-18 hours without the ability or vigor of gastrulation being diminished. The stage of gastrulation does not express itself as determining the capacity to continue gastrulation in this modified environment. Differentiation without gastrulation was not found nor gastrulation without differentiation except in fatally injured embryos. Information is given on the concentrations which caused complete but reversible inhibition of gastrulation and on the limits of alcohol toleration. 1,500. Fic. 7. Higher magnification of segment c, Fig. 1. Beginning of pallisade formation. X 1,500. Fic. 8. Higher magnification of segment d, Fig. 1. Bacteroid layer buckling at the end of .5 mm. odcyte. X 1,500. Fic. 9. Higher magnification of segment e, Fig. 1. Ridges of bacteroids at Boe of a 1 mm. oocyte, yet limited between o6cyte membrane and follicle cells. X 1,500. Fic. 10. Higher magnification of segment f, Fig. 1. Bacteroids dispersing through the polar cytoplasm of a mature egg. X 1,500. Fic. 11. Posterior end of 1 mm. oocyte and anterior end of mature egg of Blatta orientalis. X 150. Fig. 12. Bacteroid mass, with one giant nucleus and three smaller ones, at the posterior end of an egg after 60 hours incubation. X 650. 442 HERSCHEL i Gini wall is very distinct both in well-fixed tissues and in shrunken material. The oocyte membrane is very thin and is visible only as a definite limit to the cytoplasm. Shortly before the egg is oviposited and soon after the first appear- ance of the chorion, the odcyte membrane disappears and a new mem- brane, interpreted as the vitelline membrane, is formed between the bacteroids and the chorion (Fig. 10). The bacteroids, formerly lim- ited to definite ridges, now disperse throughout the cytoplasm between the vitelline membrane and the yolk so that by 36 hours after oviposition there is a homogeneous disc-shaped mass of bacteroids at each pole, and numerous isolated clumps around the periphery of the egg. No bacteroids have ever been seen within yolk granules, as reported by Buchner (1912, his Fig. 6). Buchner (1912) explained the polar concentrations by assuming an active migration of the bacteroids toward the poles. Such an as- sumption is not necessary if the differential rate of surface increase is considered. For example, an odcyte 90 pw in diameter and 100 » in length is surrounded by a uniform layer of bacteroids, while one 160 by 600 » has very noticeable concentrations at the poles. Considering the odcytes as cylinders, the ratio of end area to side area of the smaller is 6360 :28,- 300 and of the larger is 20,100:301,600. This is an increase in end area of 201 per cent and in side area of 966 per cent; or, the side has expanded 4.8 times as much as the end. The original end of the odcyte, however, enlarges at most to 140 y, the remainder of the increase in diameter being accounted for by curving of the side, thus adding to the calculated growth of side surface. Consideration of the changing pro- portions during the final growth period is an even more convincing argument for this point. In enlarging from .35 by 1 mm. to 1 by 3 mm., the area of the sides increases approximately ten times while the original end of the egg expands less than 50 per cent. Without calculating the differential increases accurately, it is obvious that these differences are sufficient to account for the polar concentrations and the later thinning of the lateral layers, 1f the bacteroids multiply at the same rate all around the oocyte. In all species of roaches studied, the ovarian history of the bacteroids is remarkably uniform, the only differences being in the number of the bacteroids present and in the structure of the ovary itself. Periplaneta australasie is like Periplaneta americana in all respects noted. Eurycotis floridana differs from these only in having smaller polar masses of bacteroids. _ Blatta orientalis shows the same bacteroid relations around the | smaller oocytes, but the layer becomes much thicker, buckling distinctly BACTEROIDS OF ROACHES 443 in .3 mm. eggs. The resulting ridges increase to such an extent that at each pole of a .5 mm. egg there is a nearly solid hemispherical mass which enlarges to a maximum of 60 by 75» ina .7 mm. egg, then spreads out in the final growth period into a vacuolated disc about 40 » thick and 160, in diameter (Fig. 11). The vacuolated appearance is due to a partial restitution of the original ridges of the bacteroids. Cryp- tocercus punctulatus, although it has the same type of ovary as Peri- planeta americana, ordinarily does not have bacteroids around the first four to ten distal odcytes. The bacteroid layer becomes 4 p to 5 p thick, with no tendency to form a pallisade arrangement. The layer buckles uniformly around the .6 mm. egg, with only slight polar concentrations until the side layers are thinned in the final rapid growth period. No mature eggs of this species were studied. Consistently in the Cryp- tocercus punctulatus specimens examined, about one-tenth of the young oocytes were infected and apparently destroyed by the bacteroids. They penetrated the oocyte membrane, usually in very early stages, and in- creased within the yolk until what had been an oocyte was a mass of bacteroids. _ A similar condition is occasionally found in Blatta orientalis and rarely in Periplaneta americana. In these latter, however, the bacteroids never increase as much as in Cryptocercus. In Blattella germanica and the species of Parcoblatta studied, the ovarioles are very slender near the germarium, each containing only one oocyte in cross- section in contrast to the two, three or four in a similar section of a Periplaneta ovariole. Accompanied with this difference is the absence of any bacteroids around the first two to four distal oocytes. The most anterior one to have bacteroids always has the greater number at the posterior end. There are always two or more oocytes surrounded by bacteroids anterior to a completed follicle, which here, as in Peri- planeta, splits the layer between oocytes. The bacteroids gradually become more numerous, making a double layer or an incomplete palli- sade around the oocyte. There may be a slight buckling at the corners and at most a triple layer at the poles. In the final enlargement of the egg, the bacteroids become scattered around the sides while the end con- centrations remain as thin layers. In all species of roaches studied, bacteroids invariably occupy the space between the follicle cells and the oocyte until the vitelline mem- brane begins to form, at which time they enter the egg cytoplasm. After oviposition, while the early cleavage nuclei are dividing within the yolk, the bacteroids retain their position against the vitelline mem- brane. Many of the first nuclei to reach the periphery (two days in- cubation), enter portions of cytoplasm containing bacteroids, and within a few hours sink again into the yolk accompanied by the bacteroids. 444 HERSCHEL ain iGiik These nuclei, with their cytoplasm and bacteroids, I shall call “ primary mycetocytes.” Others, indistinguishable from the above, come to the periphery and return into the yolk without bacteroids. These are the yolk nuclei or vitellophags. The remainder of the nuclei stay at the surface and form the blastoderm, which is continuous by the third day, except at the regions of the polar masses of bacteroids. The course of events at the poles is not so simple as that for the rest of the egg surface. Each polar mass (Fig. 12) receives five to ten nuclei toward the end of the second day of incubation. The number increases to about forty by the fourth day. Most of them resemble the primary mycetocyte nuclei but a few (one to four) of the original ones become very large and have a considerable volume of clear cyto- plasm. Apparently these giant nuclei do not divide and their later history is unknown except that they become separated from the bacter- oids early in the migration of the polar masses. After this, they are not readily distinguishable from the ordinary vitellophags. On the third day, the anterior bacteroid mass breaks into several smaller clumps which within two days sink into the yolk, reunite and proceed toward the center of the egg. The posterior bacteroid mass at PrLate JI Explanation of Figures Fic. 13. Posterior bacteroid mass passing ventrally and anteriorly between the yolk and the embryonic rudiment. One giant nucleus has been left behind. Posterior ventral segment of a median sagittal section of an egg after 314 days incubation. X 150. Fic. 14. Bacteroid masses in the yolk. The outlines of the yolk granules are shown only around the bacteroid clusters. This and the following eggs have yolk granule and oil globule arrangement similar to that of Fig. 13. Median sagittal section of a six-day embryo. X 55. Fic. 15. Bacteroid masses uniting near the center of the yolk to form the primary mycetome. Median sagittal section of a nine-day embryo. X55. Fie. 16. Eleven-day embryo with the primary mycetome still intact. 55. Fig. 17. Longitudinal section of the last thoracic segment and the anterior part of the abdomen of a 12-day embryo, nearly full length of the ovary. Approxi- mately at the location of guide line b on Fig. 18. Many cells of the fat body con- tain bacteroids. X 110. Fic. 18. Cross-section through the third abdominal segment of 12-day embryo (from the level of a on Fig. 17). X 110. Fic. 19. Parcoblatta pennsylvanica embryo at approximately the same stage of development as that in Fig. 14. Bacteroids moving in small streamers between the yolk granules. X 70. Fie. 20. P. pennsylvanica embryo intermediate in development between those of Figs. 15 and 16. Bacteroids in the primary mycetome. X 70. Fic. 21. Portion of an ovariole of a four-month-old Periplaneta americana. Bacteroids are located in the mycetocyte outside the tunica propria and as a small clump between two odcytes around which they are dispersing. X 1,000. Fig. 22. Median section through the mycetome of a ten-day embryo. X 300. 445 BACTEROIDS OF ROACHES PLATE II 446 HERSCHEL T. GIER this time is nearly spherical and lies at the posterior end of the newly differentiated embryonic rudiment. Near the end of the third day this mass begins to migrate anteriorly and ventrally, passing between the germ band and the yolk to about the middle of the embryo (Fig. 13), then it swings directly into the yolk (fifth day), either as a solid mass or as several clumps which reunite after proceeding a short dis- tance. Both polar masses now move slowly toward the center of the yolk (Fig. 14) being joined on the way by the small lateral masses. All bacteroid masses join just anterior to the center of the egg about the end of the eighth day (Fig. 15) to form a compact mass approx- imately .1 mm. across, which mass Buchner (1912) called the “ primary mycetome.” The bacteroids within this mycetome are arranged in pyramidal segments, fitted together to form a sphere (Fig. 22). Each segment has one or two nuclei. Around the bacteroid sphere there is a single to double layer of vitellophag nuclei. No cell walls have been distinguished in this mass. Heymons (1895) believed that the nuclei associated with the bac- teroids increase by amitotic division. Although mitotic figures are rarely found and the nuclei are often lobulated as if pinching in two, I am not convinced that there are not enough mitotic divisions to ac- count for the very slow increase in the number of these nuclei. There is no question that the bacteroids do exhibit definite migra- tions. The posterior bacteroid mass in some eggs moves as much as 2.5 mm. in five days. This mass is made up of a loose clump of bac- teroids and mycetocyte nuclei, surrounded by a layer of vitellophags. In some eggs the anterior surface of the mass is very irregular as if it were pushing forward between the yolk granules. In others, the yolk in the path over which the bacteroids have passed is more homogeneous than the surrounding yolk, suggesting that it had been dissolved. The small lateral masses unquestionably flow inward between the yolk gran- ules in company with one or two ameboid nuclei. The bacteroids in all cases appear to be carried along passively with the mycetocytes and vitellophags. So far as can be determined, all bacteroids in the egg are enclosed in the primary mycetome for about three days, after which some appar- ently leave the mycetome and migrate to the embryo. At eleven days incubation the embryo has straightened out along the ventral surface of the egg and is approximately one-half the length of the egg (Fig. 16). Segmentation is complete on the ventral side, but the dorsal wall is formed only on the most posterior three or four abdominal segments. The lateral walls are formed dorsally to a line running diagonally around the egg from the head to the most anterior BACTEROIDS OF ROACHES 447 point of the completed dorsal wall. The yolk is rapidly enclosed from this time on by dorsal and anterior growth of the lateral walls and elongation of the whole embryo so that by twelve days the dorsal wall is formed to the third abdominal segment (Fig. 17), and by fifteen days the embryo is as long as the egg and the yolk is entirely enclosed. At eleven days the fat bodies are rudimentary and strictly segmental, while the gonads are elongated masses of cells. By twelve days the gonads are definite cords, and by thirteen days the ovariole rudiments are recognizable as.segmental enlargements of the ovary. The gut epi- thelium is continuous along the ventral side of the yolk between the stomodeum and the proctodeum and extends dorsally around the yolk as a discontinuous cell layer. The body cavity ends blindly at its an- terior margin beyond the limits of the gut epithelium (Figs. 17, 18). In the period between eleven and twelve days of incubation, the bacteroids within the mycetome lose their definite arrangement. The mycetome elongates and breaks and bacteroids appear between the yolk granules posterior to the mycetome. They also appear against the gut epithelium, in the body cavity, and in cells of the lateral lobes of the fat body bordering the body cavity. Immediately after the mycetome breaks, its posterior wall appears to have moved into the yolk. After eleven days of incubation, the number of bacteroids against the gut epithelium increases, reaching a maximum at about the eighteenth day. After that time they rapidly disappear from the gut, apparently by degeneration. The nuclei of the mycetome become indistinguishable from the vitellophags, and apparently most if not all of them fuse with the gut epithelium late in embryonic development to become the defini- tive entoderm, much as described by Stuart (1935) for a grasshopper. The foregoing description of the primary mycetome, behavior of the bacteroids, and embryonic development in general, are matters of direct observations on fixed and stained material. Since it is impossible to see the bacteroids within the dense, living egg, their behavior must neces- sarily be interpreted from static material. Heymons (1895) stated that the bacteroids move centrifugally from the primary mycetome through the wall of the gut and into certain cells of the fat bodies, after the gut completely encloses the yolk. Consider- ing the stage of embryonic development and the places in which the bacteroids are found at the time, the following method of infection seems most probable. The vitellophags liquify the yolk around the pri- mary mycetome which then loses its compactness. This liquid digestion product suddenly flows toward the embryo, separating the posterior cells of the mycetome. The nuclei are detained by the dense yolk granules, but some of the bacteroids are carried along passively in the 448 EVER S © Ele biG ilar flow until they reach the embryo. If these bacteroids come in contact with the formed gut epithelium, they are retained while the liquid that carried them diffuses through. If they come in contact with the incom- plete margin of the gut, or pass immediately anterior to the gut, they enter the body cavity directly. Within the cavity they are carried more posteriorly and ventrally and brought into contact with the fat bodies. It seems possible, however, that in the early migration of the bac- teroids a few may have remained with the embryonic rudiment while the rest proceeded to the primary mycetome. Most, if not all, of the cells of the lateral lobes of the abdominal fat bodies bordering the body cavity take up a few bacteroids in the cytoplasm between the cell nucleus and the body cavity, thereby becom- ing mycetocytes. No differences between the cells of different parts of the fat bodies are discernible before infection, nor has any mechanical principle been found that would explain selective infection, yet the fact remains that neither the dorsal nor the ventral parts of the fat bodies become infected. As development proceeds, uninfected fat cells in- close the mycetocytes, or possibly as Buchner (1912) thinks, the infected cells sink into the fat body. The segmental fat bodies enlarge and fuse, having completely enclosed the gonads in the meantime. The bacteroids within each mycetocyte multiply rather slowly, causing an increase in mycetocyte size from approximately 12 » at the time of infection to 25 p at the time of hatching, and 30 at the first moult. Few mycetocytes get larger than 35 in diameter. The mycetocyte nuclei remain dis- tinctly eccentric until two or three weeks after the roach hatches. No indication of mycetocyte division has been found, and no increase in numbers of mycetocytes in a roach is discernible after the first infection. The final form of the fat body is attained by the time of the second moult by two processes; dispersal of the mycetocytes among the ever enlarging fat cells before the roach hatches; and later separation of the mass into lobes of varying sizes, most of which have cores of myce- tocytes. Infection of the ovary has been considered to be a more complicated process than has that of the fat body. Buchner (1912) considers it probable that the bacteroids pass directly from the mycetocytes (Fig. 5 and Buchner, 1912, his Fig. 1), through the young follicle cells to the oocytes. Fraenkel (1921) reported that she has seen steps in this process (her Figs. 1, 2). They do not specify, but intimate that this process is repeated for the infection of each oocyte. It seems improb- able, however, that a non-motile element could pass through the wall of the mycetocyte, across a potential cavity, through the very tough tunica propria and finally through the follicle cell. BACTEROIDS OF ROACHES 449 The relationships between bacteroids and oocytes described above (p. 439) exist in immature ovaries as young as four months (second moult). The ovaries of a young first moult roach are all like the tips of older ovarioles. Each ovariole is closely surrounded by mycetocytes outside the tunica propria. A cursory examination of the ovary at this stage reveals no bacteroids within the tunica. Before the second moult, a few oocytes in the posterior part of each ovariole have enlarged to two or three times their original diameter. Against some of these, there are clumps of bacteroids apparently dispersing around the oocytes (Fig. 21). These groups of bacteroids are often very close to mycetocytes from which they sometimes appear to be flowing, but no break in the tunica propria has been found. Careful study of ovaries of roaches younger than three months almost invariably reveals the presence of one or more bacteroids in each ovariole, usually among the more posterior oocytes. The bacteroids in this stage are easily overlooked because they stain much like the chromatin of the odcyte nuclei. Either they occur in greater numbers or they are more easily recognized between the enlarging oocytes. of the three-month ovaries than between the small oocytes of the one-month ovary. In consideration of the observations just given, the following is the most probable method of infection. At the time of the fat body in- fection, a few bacteroids find their way into the mass of ovarian cells and remain there, increasing very slowly for a time. Within twenty- four hours the ovary becomes a compact cord of germ cells, between which are scattered the bacteroids. The ovarioles are formed as out- growths from the original cord, enlarging at first entirely through multiplication of the odgonia. As this multiplication is greatest at the distal end, the bacteroids there become more scattered. The oogonia at the basal end of the ovariole soon cease division. Their chromatin passes through pachyphase before the roach hatches, and becomes diffuse soon after the first moult. With the advent of the diffuse chromatin the odcyte begins more rapid growth and the bacteroids begin dividing and gradually spread over the surface of the odcytes (Fig. 21). This dispersal is undoubtedly accelerated by the shifting of the oocytes in the ovarioles and by pressure on the ovariole from general body move- ments. With the completion of the follicle around the ‘most posterior odcytes, the spread of bacteroids is limited to a forward migration which keeps pace with the development of new odcytes to the infectable stage. In review, the following observed facts favor this hypothesis of ovarian infection: (1) the ovary is a rough mass of cells without defi- nite borders at the time of fat body infection; (2) there are a few 450 HERSCHEL T. GIER bacteroids in the very young ovary; (3) the first obvious infections are small clumps of bacteroids which appear to be spreading around the odcytes. These clumps may be in the center of the ovariole, far removed from a mycetocyte; (4) passage of bacteroids from myceto- cytes to ovarioles has not been convincingly demonstrated; (5) there are always several odcytes with bacteroids at their surfaces anterior to a completed follicle; and (6) in ovarioles like those of Periplaneta, there are usually bacteroids around all odcytes posterior to the tip, and so arranged as to give the impression that they had moved anteriorly in the angles between the odcytes, while in ovarioles like those of Parco- blatta where each oocyte fills the entire cross-section of the egg tube, the bacteroids are arranged evenly around the most anterior infected oocyte. The principal points against this hypothesis are: (1) the migrations of the bacteroids have not been seen in living roaches; (2) no bacteroids have been seen in the ovary for two weeks after the supposed time of infection (possibly due to lack of differential staining) ; and (3) the hypothesized retardation of bacteroid growth in the early ovary has not been established. This last objection, however, is apparently the same factor that limits the bacteroids to odcytes that have passed a certain stage of development. No differences have been found in the bacteroid behavior throughout the embryonic development of Periplaneta americana, P. australasie and Eurycotis floridana. However, Blatta orientalis, Parcoblatta penn- sylvanica, P. uhlerania, and P. latta differ from Periplaneta in the effect of a median ventral rather than a posterior embryonic rudiment. In Parcoblatta the polar masses migrate directly into the yolk. In this migration, the bacteroids may be in small separated clumps, each with two to four nuclei, and pass close to the embryo (Fig. 19), or they may be in a large loose mass and pass deeper through the yolk. The lateral clumps join the general mass while en route but remain distinct until they reach the center of the egg. Here all masses unite to form a pri- mary mycetome (Fig. 20) very similar to that of Periplaneta americana. In Blatta orientalis the polar masses are much larger, but the behavior is similar to that in Parcoblatia. Bacteroid behavior after formation of the primary mycetome is the same in all species studied, so far as could be determined. I have no embryological material of Cryptocercus punctulatus and so little of Blattella germanica that it warrants no com- parisons nor conclusions as to bacteroid behavior. Buchner (1912) described a primary mycetome in the last-named species. BACTEROIDS OF ROACHES 451 I wish to express my gratitude to Dr. Fernandus Payne and to the other members of the Faculty of the Zoology Department of Indiana University for their criticism and suggestions during the progress of this work. SUMMARY 1. The staining reactions, morphology and occurrence both intra- and intercellularly definitely preclude the intracellular bacteroids of the roaches from the category of mitochondria and link them closely to both the diphtheroids and the Rickettsia. 2. The number of bacteroids between the oocyte membrane and the follicle cells increases until there is a uniform layer two or three bac- teroids thick. By differential increase of the oocyte surface this layer becomes thinner along the sides. 3. Before the egg is oviposited, the original oocyte membrane breaks down and permits the bacteroids to enter the cytoplasm. 4. As the embryo develops, the bacteroids, accompanied by nuclei similar to the vitellophag nuclei, move in masses to the center of the yolk. 5. Later a few bacteroids from this central mass move posteriorly between yolk granules through the incomplete margins of the gut epi- thelium into the body cavity. From there most of them are taken up by the cells of the lateral lobes of the fat bodies while a few are caught between the cells of the forming ovaries. 6. The bacteroids within the ovariole seem to lie dormant for several weeks, then they multiply rapidly and spread over the surfaces of the enlarging oocytes. 7. Newly enlarged odcytes from the germarium are infected by bac- teroids moving anteriorly from older oocytes. 8. Bacteroid behavior varies between roach species with differences in number of bacteroids, in structure of the ovary, and in position of development of the embryonic rudiment. 9. All bacteroid migrations appear to be passive. LITERATURE CITED BrocuMAnNN, F., 1888. Ueber das regalmassige Vorkommen von bakterienahn- lichen Gebilden in den Geweben und FEiern verschiedener Insecten. Zeitschr. f. Biol., 24: 1. Bucuner, P., 1912. Studien an intracellularen Symbionten. I. Die intracellularen Symbionten der Hemipteren. Arch. f. Protist., 26: 1. Bucuner, P., 1921. Tier und Pflanze in intracellularer Symbiose. G. Borntrager, Berlin. Bucuner, P., 1923. Studien an intracellularen Symbionten. IV. Die Bakterien- symbiose der Bettwanze. Arch. f. Protist., 46: 225. 452 HERSCHEL We Girik Bucuner, P., 1925. Intracellulare Symbiose der Tiere mit pflanzlichen Mikro- organismen. Handb. d. biol. Arbeitsmeth., Abt. XII, Tiel 1: 325. Bucuner, P., 1930. Tier und Pflanze in Symbiose. 2 Aufl. G. Borntrager, Berlin. Cownry, E. V., anp P. K. Oxirsxy, 1922. Differences between mitochondria and bacteria. Jour. Exper. Med., 36: 521. Cuénot, L., 1896. Etudes physiologiques sur les Orthopteres. Arch. Biol., 14: 293. FLorENcE, L., 1924. An intracellular symbiont of the hog louse. Am. Jour. Trop. Med., 4: 397. FRAENKEL, H., 1921. Die Symbionten der Blattiden im Fettgewebe und Ei ins- besondere von Periplaneta orientalis. Zeitschr. f. wiss. Zool., 119: 53. Guaser, R. W., 1930a. The intracellular “symbionts” and the “ Rickettsia.” Arch. Path., 9: 71-96 and 557-576. Giaser, R. W., 1930b. On the isolation, cultivation and classification of the so- called intracellular “ symbiont” or ‘‘ Rickettsia” of Periplaneta americana. Jour. Exper. Med., 51: 59. Hennecuy, L., 1904. Les insectes. Mason et Cie, Paris. Hertic, M., 1921. Attempts to cultivate the bacteroids of the Blattide. Biol. Bull., 41: 181. Heymons, R., 1892. Die Entwicklung der weiblichen Geschlechtsorgane von Phyllodromia (Blatta) germanica L. Zeitschr. f. wiss. Zo6l., 53: 434. Heymons, R., 1895. Die Embryonalentwicklung von Dermapteren und Orthopte- ren. Gustav Fischer, Jena. Kocu, A., 1931. Die Symbidse von Oryzaephilus surinamensis L. (Cucujide, Coleoptera). Zeitschr. f. Morph. u. Okol. Tiere, 23: 389. LILIENSTERN, M., 1932. Beitrage zur Bakteriensymbiose der Ameisen. Zeittschr. f. Morph. u. Okol. Tiere, 26: 110. MaAwnsour, K., 1934. On the intracellular micro-organisms of some bostrychid beetles. Quart. Jour. Micr. Sci., 77: 243. Mansour, K., 1934. On the so-called symbiotic relationship between coleopterous insects and intracellular micro-organisms. Quart. Jour. Micr. Sci., 77: 25a: Mercier, L., 1907. Recherches sur les bactéroides des Blattides. Arch. f. Protist., 9: 346. Meyer, K. F., 1925. The “bacterial symbiosis” in the concretion deposits of cer- tain operculate land mollusks of the families Cyclostomatide and An- nulariide. Jowr. Infect. Dis., 36: 1. Neuxomm, A., 1927a. Sur la structure des bactéroides des Blattes (Blattela germanica). Compt. rend. Soc. Biol., 96: 306. Neuxomm, A., 1927b. Action des rayons ultra-violets sur les bactéroides des Blattes (Blattela germanica). Compt. rend. Soc. Biol., 96: 1155. ScHwartz, W., 1924. Untersuchungen tuber die Pilzsymbiose der Schildlause. Biol. Centralbl., 44: 487. Suirgr, E. H., 1934. Insect development. VI. The behavior of grasshopper oe. in anisotonic, balanced salt solutions. Jour. Exper. Zodol., 67: SW Suirer, E. H., anp R. L. Kine, 1933. Grosshopper eggs and the paraffin method. Science, 78: 366. Stuart, R. R., 1935. The development of the mid-intestine of Melanoplus dif- ferentialis (Acridide: Orthoptera). Jour. Morph., 58: 419. WN ceeaaee Se a is Studies on Rocky Mountain spotted fever. Jour. Med. es., 41: 1. AUTOGENOUS TRANSPLANTATIONS OF PIGMENTED AND UNPIGMENTED EAR SKIN IN GUINEAS Pigs, JOHN A. SAXTON, JR., MARY M. SCHMECKEBIER AND ROBERT W. KELLEY (From the Department of Pathology, Washington University School of Medicine, St. Louis, Missourt) The experiments to be described consist in a continuation of a series of investigations begun by Loeb and by Carnot and Deflandre, and studied further by Sale and Seelig and others, concerning the fate of autogenous transplants of pigmented and unpigmented skin in the ear of the guinea pig. Carnot and Deflandre (1896) and Loeb (1897) independently carried out experiments in which they transplanted pig- mented and unpigmented epidermis from the ears of guinea pigs. In certain respects the results obtained by these investigators were similar, but in other respects differences developed. When unpigmented ear skin is autotransplanted to a defect in pigmented skin, the graft is either lost by scaling, or is invaded and replaced by pigmented epithelium from the surrounding epidermis. An autogenous graft of pigmented skin to a defect in a white area, however, persists; and in the course of time the pigmented cells grow out and invade the surrounding white epidermis, thus increasing the size of the colored graft. Although it is not especially stated, Loeb’s transplants were autogenous, whereas Carnot and Deflandre state that they successfully carried out homoio- genous and even heterogenous grafts. The latter investigators further- more report that while grafts from one species to another as a rule failed to grow, grafts from guinea pig to rabbit, and from rabbit to guinea pig, appeared to persist for a while, although with certain dif- ficulties. These authors could not detect any differences in behavior between autogenous and homoiogenous transplants. The investigations of Loeb were carried further by Sale (1913) and Seelig (1913), who compared autogenous and homoiogenous grafts of pigmented and unpigmented skin in guinea pigs. Sale showed that auto- and homoiogenous transplants of pigmented epithelium to white skin behave differently. The large majority of autogenous grafts sur- vive and outgrowth of pigmented epidermis into the adjoining white 1 These investigations were carried out with the aid of a grant for research in science made to Washington University by the Rockefeller Foundation. 453 454 Jj. A. SAXTON, M. M. SCHMECKEBIER AND R. W. KELLEY epithelium is the rule. Only a few of the homoiogenous pigmented grafts remained alive, but in these cases invasion of the adjacent white epidermis did not occur. Instead, the pigment-producing function of the transplanted cells was injured and the graft gradually lost its pig- ment. As a rule an accumulation of lymphocytes underneath or in the transplanted epidermis preceded a partial separation of the trans- plant and its destruction in the majority of cases. Seelig has shown that unpigmented grafts when auto- or homoiotransplanted to defects in pigmented skin, are invaded and replaced by the adjacent pigmented epidermis. Rand and Pierce (1932), studying autogenous and homoiogenous transplants of white ventral skin to pigmented areas in the tadpole, concluded from gross and microscopic observations that pigmentation of these grafts occurred through invasion and replacement of the graft by the surrounding epidermis. Barriers of hen’s eggshell membrane could temporarily protect the graft from invasion and destruction. In autogenous grafts this invasion did not occur so regularly as in homoiotransplants. Herrick (1932) confirmed this observation so far as homoiogenous grafts in tadpoles were concerned, but concluded that autogenous grafts of unpigmented ventral skin retain their specificity and are not invaded for more than one hundred days. Seevers and Spencer (1932) were unable to confirm the results of Loeb in regard to outgrowth of autogenous pigmented skin grafts into the surrounding white skin and the invasion of unpigmented grafts by surrounding pigmented epithelium in guinea pigs. Rectangular grafts having areas of about two square centimeters were made and the animals were observed for as long as nine months. During this time the trans- planted black epidermis retained its pigment but it did not grow out into the neighboring white skin in any case. A few white grafts trans- planted to defects in pigmented areas did become slightly darker along their borders after four to six months, but many such grafts remained unpigmented throughout the period of observation. In no case was the color of the hair changed, either in the graft or surrounding tissue. Except in the case of transplants to back and belly in agouti guinea pigs, there is no definite statement as to the part of the animal’s body where these grafts were made; in the case of the experiments of Loeb and the other workers the ear was used without exception. The technique used in our experiments was similar to that used by Sale and by Seelig. Guinea pigs were selected which had either one pigmented and one white ear, or sufficiently large pigmented and white areas on either one or both ears. With aseptic precautions a circular area of each type of skin approximately 4 mm. in diameter was re- AUTOTRANSPLANTATION IN GUINEA PIGS 455 moved with a sharp razor. As little as possible of the connective tissue underlying the epidermis was included in the excised tissue. Bleeding, usually slight, was controlled by pressure. The skin pieces removed were placed in 0.9 per cent sodium chloride solution and cut to the desired size, usually about 2 mm. square. After oozing had ceased the grafts were placed on the denuded areas according to the experi- ments described, blotted, and alcohol-soaked cotton dressings were firmly applied, held in place with collodion. The dressings were usually loosened at the end of ten days, and subsequently observations and measurements were made at intervals of a few days or weeks for sev- eral months. Thirty-two animals in all were used and eighty-three individual grafts were made. Approximately one-half of the grafts did not survive, probably owing to errors in technique in our experi- mental procedure, or more often perhaps owing to scratching of the wound on the part of the animals. The following data give a brief summary of the experiments per- formed and the survival of the grafts. It is interesting to note that practically the same proportion of pigmented and unpigmented grafts survived. Woke Total Survived \yidiitetonatts toppieimentedsskamirjscscG ssc cte ssc eels 6 34 Single pigmented grafts to white skin .................... 24 10 Double pigmented grafts to white skin .................. 11 5) Eriplerpicmented sraits tovwhite skin’.....0:.2.c08... 068 1 1 In making the double and triple pigmented grafts a small area of intact white skin was usually left between the defects. Multiple transplants of white to pigmented skin were also made, but as they did not behave differently from single transplants, they are included in the discussion — of the latter. The behavior of some of the representative grafts is de- scribed below. Autogenous Transplants of Unpigmented Skin to Defects in Pigmented Skin Our observations in fifteen cases where we were certain that the grafts were present were in accord with those of Loeb, Carnot and Deflandre, and Seelig. Complete invasion of the graft occurred in all cases. During invasion by surrounding pigmented epithelium there was, we also noticed, some scaling of the superficial layers of the graft. This appeared to persist as long as any white epidermis was yet visible, but its exact degree could not be determined grossly. In Seelig’s ex- periments the time required for complete invasion varied from 13 to 90 days. We observed complete invasion in one case at 24 days, and in 456 j. A. SAXTON, M. M. SCHMECKEBIER AND R. W. KELLEY, the others the time required varied up to about ninety days or more. In one instance 200 days were required, and here the grafting was as- sociated with considerable thickening of the underlying tissues, and this reaction, or perhaps the condition which gave rise to it, may have injured the host epidermis and thus have been responsible for the slow growth. After complete invasion the site of the graft was usually somewhat thickened and appeared more deeply pigmented than the rest of the ear. Since similar observations have been made in connection with normal regeneration, this perhaps indicates that the presence of a graft of un- pigmented epidermis does not greatly alter the normal regenerative processes of pigmented epidermis except in so far as rate of repair is concerned. No correlation between the rate of invasion and the color or degree of pigmentation of the skin surrounding the transplant could be noted, grafts being made in defects of black, gray and brown skin. Single Autogenous Transplants of Pigmented Skin to Defects in Unpigmented Skin The results of these experiments likewise confirm the earlier ob- servations of Loeb, Carnot and Deflandre, as well as those of Sale. Outgrowth of pigmented epidermis from the graft occurred in each of the ten cases studied. Some of these grafts were observed for 280 days. An average of about twenty days is required before a recog- nizable increase in the size of the pigmented area occurs. Occasionally the graft shriveled somewhat in this period, and in two cases seemed to have been lost. However, in these two cases a small point of pig- ment could be recognized at the center of the defect after 10 days and subsequently grew at a rapid and constant rate remarkably parallel to that of those which had remained intact during the first 20 days. There appears to be a tendency for the outgrowth to be most rapid during the period between the twentieth and eightieth day following transplantation. During this time the diameter of the pigmented area increased from the original 2 mm. to a size somewhere between 4 and 6 mm., and hence it must have extended over the margins of the original defect in the skin. In most cases it was not possible grossly to deter- mine whether the removal of the skin and subsequent regeneration in the area surrounding the transplant affected the outgrowth of the pig- mented epidermis in this area. Since recognizable outgrowth did not occur until twenty days after grafting, regenerative processes of the epidermis and dermis over the denuded area had occurred by this time, AUTOTRANSPLANTATION IN GUINEA PIGS 457 so that a definite demarcation between the original defect and the sur- rounding skin did no longer exist at this time. After about eighty days the growth rate usually began to decrease, but the outgrowth continued at a decreasing rate for as long as two hundred and eighty days after grafting. In two cases where the pig- mented areas were oval in shape, growth in the largest diameter ceased at about one hundred and sixty days, but the width of the area increased slightly during the following one hundred days of observation. In Sale’s experiments no data were given concerning the duration of the period of outgrowth. As far as the direction of the outgrowth is concerned, it was of a fairly uniform character, roughly symmetrically surrounding the orig- inal graft. However, the margins of the zone of outgrowth often pre- sented a frayed appearance, due to fine pigmented processes which ap- peared to push ahead of the main outgrowth. These processes did not seem to be oriented in any particular way with reference to the adjacent blood vessels, scar tissue, or neighboring pigmented areas except in one case, described below, in which outgrowth was less rapid for a time in the direction toward a neighboring area of normal brown skin. The depth of pigmentation in the zone of outgrowth was usually of the same degree as in the original skin used for transplantation; occa- sionally, however, it appeared somewhat lighter, possibly due to the fine processes mentioned, between which there were strips of non- pigmented skin. The color of the outgrowth always corresponded to that of the original graft, whether the latter was black, gray or brown. We were unable to find any definite correlation between rate of out- growth and color of the grafted piece. With one exception described below, all appeared to grow equally well. One of the best growing cases was a graft of gray epidermis, pitch had increased in size between the final two observations. In this series we were not specifically concerned with the problem as to the possible effect of the amount of pigmentation or of the lack of it in the whole animal on the rate of outgrowth from the graft. In the majority of the animals used the area of pigmented skin was larger than the area of white skin, taking the body as a whole, but in one in- stance we used an animal that was almost entirely white except for one ear, which had a rather light brown color. From the skin of this ear two grafts were made to the white ear of the same guinea pig, and it was found that the diameters of the grafted areas increased 2 and 3 mm. respectively during a period of two hundred and forty days; whereas the average increase for the rest of the animals observed during this length of time was about 6 mm. In this single instance our results 458 J. A. SAXTON, M. M. SCHMECKEBIER AND R. W. KELLEY appear to confirm, perhaps, those of Carnot and Deflandre, who found that pigmented grafts extend more rapidly on animals showing a higher degree of total pigmentation. In our animals, however, the pigmenta- tion of the one brown ear was not intense, a fact which in itself may be sufficient to explain the slow outgrowth in this case, rather than the deficiency in total body pigmentation. . Where hair was present in the area into which pigmented epidermis extended, it remained unpigmented during the period of observation. This would indicate that the pigmented skin invades surface epithelium only and does not readily extend into the hair follicles. While we con- - cede the possibility that this may ultimately occur, it evidently requires a longer period of time than that during which our transplants were under observation. Multiple Autogenous Transplants of Pigmented Skin to Defects in Unpigmented Skin Double and triple grafts were made in order to obtain additional proof that the original white skin surrounding the defect becomes in- vaded by pigmented epithelium of the graft. When two pigmented grafts separated by an intervening area of normal white skin become connected, it is necessary to conclude that this white skin must have been invaded by the pigmented epidermis. Five double transplants and one triple transplant were made. They were separated in most instances by intact white skin varying in width from 1 to 4 mm.; but in one case both grafts were placed 2 mm. apart in the same defect. The double grafts behaved exactly like the single ones, in that out- growth of pigmented epidermis occurred. In two cases, where the grafts were fairly close together, fusion occurred during the period of observation. In one of these, in which the grafts were separated by 1 mm. of intact white skin, they had grown together after thirty-seven days, forming an elongated dark patch from which processes extended radially to the margin of the ear on one side, and medially to a neigh- boring area of normal black skin. In the case of the triple transplant two of the pigmented areas, both in the same defect, had joined by thirty days; and by sixty days all three grafts had fused across 2 mm. of intervening white skin to form a single lobated area of black epi- dermis. In cases where fusion did not occur the grafts had been either placed too far apart for fusion during the period of observation, or, in one case already described, the rate of outgrowth was unusually slow. In addition to the fusion of double transplants, invasion of normal pigmented skin by outgrowth from grafted skin was also observed. In one case already mentioned where the black graft was placed at a AUTOTRANSPLANTATION IN GUINEA PIGS 459 distance of about 2.5 mm. from a normal area of brown skin, outgrowth occurred toward this brown area, although less rapidly than away from it. The outgrowth met the autochthonous pigmented margin after 95 days, and by 200 days had definitely invaded it for a distance of 2 mm. along a border of 8 mm. The black margin of the outgrowth could be clearly seen extending into the lighter brown skin. Such an observation was possible because of the difference in the character of the pigment in the autochthonous and in the transplanted area. It would have escaped detection if grafts indistinguishable in pigmentation became connected, or when a black graft joined normal black epithelium, as occurred in two cases. In another experiment two defects, 6 mm. apart, were made in a wedge-shaped strip of white skin separating two large gray-brown areas. The smaller medial defect reached the pigmented skin on either side, while a very narrow margin of white skin was left on either side of the lateral defect. In the lateral defect a small piece of pigmented skin was placed, while the medial defect was left to heal without a eraft. After a period of 56 days had passed the medial ungrafted de- fect was filled in by pigmented epidermis, and during a period of 74 days the outgrowth from the small graft in the lateral defect had ex- tended across the two narrow white margins on either side. Accurate measurements of this original white zone were not made, but later ob- servations at least indicated that the whole zone was being invaded from the adjacent pigmented surfaces, a process which was most marked along the lateral margin of the ear, at a distance of several millimeters from the graft. This suggests that in some cases the normal balance between pigmented and unpigmented epidermis may be lost. The char- acter of the pigmented epithelium of this ear was somewhat different from that of the other animals used, being a rather uneven gray-brown, not unlike the margins of some of the outgrowths of grafted brown skin. DISCUSSION We notice following transplantations a change in the equilibrium between adjoining types of epidermis which differ in their pigmenta- tion. When, under the influence of such tissue disturbances, a shift in the balance between two neighboring types of epidermis takes place, the pigmented tissue seems to dominate over the unpigmented. How- ever, we found indications that such an imbalance may also occur be- tween neighboring types of epidermis which are both pigmented, but in which the kind and quantity of the pigment differs. It is conceivable that changes similar to those we have described may take place even between two adjacent areas of black skin, but such changes would not 460 j. A. SAXTON, M. M. SCHMECKEBIER AND R. W. KELLEY become manifest. In our case we observed movements in autoch- thonous epidermis under the influence of a wound in close proximity to this skin. As to the causes of these movements in epidermis, they must be distinguished from those of ordinary regeneration in which the growth takes place into a defect. Apparently the equilibrium be- tween different types of epidermis is a very labile one which can readily be disturbed by various interferences. SUMMARY AND CONCLUSIONS If white skin is autotransplanted into a defect in pigmented skin in the ear of the guinea pig, the pigmented epidermis gradually invades the white. At the same time some scaling takes place in the white skin, and in the end the white epithelium becomes replaced by pigmented epithelium. This process could be completed in a minimum of twenty- four days, but in other cases it took as long as two hundred days. In the end the area where the white epidermis was replaced by pigmented epidermis was thickened and showed deeper pigmentation. There was no difference noticeable in a limited number of cases in the rapidity of invasion of the white transplant by black, gray, or brown epidermis. Rate of invasion does not therefore seem to depend on the intensity or kind of pigmentation of the surrounding host epidermis. If pigmented skin is autotransplanted into a defect in white skin in the ear of the guinea pig, an outgrowth of the pigmented epidermis into white autochthonous skin takes place. .As an initial process pig- mented as well as white epithelium grows into the defect, but this is followed by the growth of pigmented into the original white skin. Defi- nite outgrowth was first observed at about twenty days following trans- plantation and the outgrowth became most rapid between twenty and eighty days. Following this period outgrowth usually continued at a decreasing rate but could still be observed as late as after two hundred and seventy days. The original hair of the invaded white epidermis remained unpigmented during this process, indicating that the pig- mented cells invaded the surface epithelium first and had not yet pene- trated into the hair follicle epithelium at the time when the experiments were concluded. There was no correlation observed between color and intensity of pigmentation of the transplanted epidermis and the rapidity of its outgrowth into the surrounding white skin. In case multiple transplantations were made, it could be observed that the two neighboring pigmented transplants gradually invaded the white skin separating them and thus a junction between the two pig- mented transplants was accomplished. This observation provides an AUTOTRANSPLANTATION IN GUINEA PIGS 461 additional proof for the conclusion that it is actually the original white host epithelium which is invaded and replaced by the transplanted pig- mented epithelium. In accordance with these findings, we could observe a junction taking place between a black transplant and an area of brown autoch- thonous epidermis from which the transplant had originally been sep- arated by an adjoining area of white skin. In this case it could be noted that the black skin invaded not only the white epidermis, but also the brown pigmented epidermis; and it may therefore be concluded that not only white skin may be invaded by black pigmented skin, but also pigmented skin may be so invaded. In one case it could be observed that originally white skin became partly invaded by adjoining pigmented skin of the host. Thus the pigment pattern of the skin became changed, apparently spontaneously. However, it is possible that transplantation of black skin into white skin at a short distance from the area where the first-named change took place may have been responsible for this disturbance in the equi- librium between adjoining pigmented and unpigmented autochthonous epithelium. BIBLIOGRAPHY Carnot, PAUL, AND CL. DEFLANDRE, 1896. Persistence de la pigmentation dans les greffes épidermiques. Compt. rend. Soc. Biol., 48: 178. Herrick, Earn H., 1932. Mechanism of movement of epidermis, especially its melanophores, in wound healing and behavior of skin grafts in frog tad- poles. Biol. Bull., 63: 271. Logs, Leo, 1897. Uber Transplantation von weisser Haut auf einen Defekt in schwarzer Haut und umgekehrt am Ohr des Meerschweinchens. Arch. f. Entwick. Mech., 6: 1. RAND, HERBERT W., AND MADELENE FE. Pierce, 1932. Skin grafting in frog tad- poles: local specificity of skin and behavior of epidermis. Jour. Exper. Zoos G2Z2125: SALE, LLEWELLYN, 1913. Autoplastic and homoeoplastic transplantation of pig- mented skin in guinea pigs. Arch. f. Entwick. Mech., 37: 248. SEELIG, M. G., 1913. Homodplastic and autoplastic transplantation of unpig- mented skin in guinea pigs. Arch. f. Entwick. Mech., 37: 259. SEEVERS, C. H., anp Donatp A. Spencer, 1932. Autoplastic transplantation of guinea-pig skin between regions with different characters. Am. Nat., 66: 183. RECOVERY CHANGES IN TRANSPLANTED ANTERIOR PITUITARY CELLS STRATIFIED IN THE ULTRA-CENTRIFUGE! M. F. GUYER AND PEARL E. CLAUS (From the Department of Zodlogy, University of Wisconsin) With the thought of possibly getting some light on the beginnings of malignancy, by upsetting the normal conditions within cells and observing the subsequent effects if any on the growth and functions of such cells, the authors subjected various rat tissues to intense centrif- ugalization and then implanted them in young rats. Inasmuch as they have been particularly interested in a type of vacuolation that ap- pears in the basophile cells of the anterior pituitary body following the implantation of carcinoma (Guyer and Claus, 1932, 1933, 1934, 1935), they devoted special attention to the transplants of centrifuged pituitaries, and the present paper is a report of their observations on this material. To avoid complications which might arise from the use of anzs- thetics the 39 young adult rats from which the pituitaries were to be taken were killed by a blow on the head, the glands were rapidly dis- sected out and placed in isotonic Locke’s solution. A small bit of each pituitary was used as a normal implant into a control rat, or in a few cases, for sectioning; the remainder was placed in the isotonic solution in the metal rotor of a Beams air-driven ultracentrifuge and rotated at a speed which produced a displacement pull of about 400,000 times that of gravity. The pituitaries from 20 rats were rotated for 20 minutes, then removed from the rotor, a bit was retained for cyto- logical inspection and the rest was immediately implanted subcu- taneously into twenty young rats of between 65 and 90 grams weight, one piece of pituitary toeach rat. Ina later experiment the pituitaries from 19 rats were removed and centrifuged, using exactly the same procedure as in the earlier experiment except that the material was centrifuged an hour before implanting it in the young rats. The effect of an hour of such centrifugalization is shown in Fig. 5. The materials of the cell are almost wholly stratified. After 20 minutes of rotation in the centrifuge the stratification is very evident but not complete. Asa result of such centrifuging the Golgi apparatus * These investigations were supported in part by a Brittingham research grant and in part by a grant from the Wisconsin Alumni Research Foundation. 462 RECOVERY IN CENTRIFUGED PITUITARY 463 becomes compacted into what seems to be a single liquid mass and comes to occupy the centripetal side of the cell; that is, it is among the lightest of the cell contents. The nucleus is displaced toward the opposite or centrifugal side of the cell. In basophile cells the nucleus seemed generally to be forced up against the cell wall but in the oxyphile cells a layer of deeply staining oxyphile granules often intervened between them. The oxyphile granules, indeed, seemed to be the heaviest particles in the cell and after prolonged rotation they became massed at the centrifugal side of the cell with the nucleus forced into their midst on the side toward the cell center. Examination of material rotated for 20 minutes shows, however, that the displacement of different cell contents is at different rates of speed and that stratification in one kind of cell may be induced much more rapidly than in another, thus indicating different degrees of viscosity. The oxyphile cells of the anterior pituitary, for example, respond more slowly than do the basophiles and with the less pro- longed centrifugalization one gets the impression at first that the nucleus is coming to lie at the centripetal side of the cell. This condi- tion is temporary, however, and is probably due to the fact that the heavier oxyphile granules are being concentrated into the center and opposite side of the cell, temporarily displacing the nucleus toward the centripetal side. Eventually it is forced to the centrifugal side and, as just described, becomes more or less imbedded in the mass of oxy- phile granules. From their respective behaviors under centrifuging it is evident that the oxyphile granules found in pituitary cells are heavier than the basophile granules. The centrifuge also reveals that even in the large so-called basophile cells there are considerable numbers of oxy- phile granules, since in such cells when differentially stained a thin zone of acid-stained particles is visible. The chromatin content of the nucleus is densely massed at the centrifugal side although strands of an achromatic substance, pre- sumably linin, remain stretched across the nuclear body, giving the appearance of being attached to the opposite nuclear wall. Evidence of this is seen in the nuclei shown in Fig. 5. Inasmuch as this condi- tion has been rather fully discussed and depicted in a recent paper by Dornfeld (1936) in connection with cells of the adrenal gland, based in part on material from the same rats as were used by us in the present study, further comment on the nature of the nuclear displacements is unnecessary; the conditions seem to be similar in all respects. The implanted tissue which had been centrifuged for 20 minutes 464 M. F. GUYER AND PEARL E. CLAUS was removed and fixed for cytological study after different periods of growth, as follows: Tissue planted 24 hours, from 1 rat. AQ ccles + ll A days, 1) se memati bc “6 7 oc ims 4 rats. bc bc 14 (73 oc 5 rats. bc be 21 bc (a9 4. rats. 66 bc DS 66 bc 1 rat. Implants did not grow in 3 rats. 20 rats. The most striking fact to be noted in the transplanted pituitary cells was the rapidity with which those which survived resumed relatively normal appearance. At the end of 24 and of 48 hours respectively the removed implants showed the presence of much ne- crotic tissue and a heavy invasion of polymorphs, but this was true of control transplants as well; hence it cannot be attributed to the centrifuging. As soon as the vascularization of the transplanted tissue became well established this condition rapidly cleared up and the cells, particularly those near blood vessels, began to look normal. The last thing to be restored to its characteristic normal appearance was the Golgi apparatus. This structure seems to be very sensitive to changes in general, for even in normal, non-centrifuged anterior lobe material it tended to become clumped and pyknotic looking. The return of the nucleus to the central region of the cell and the redistri- bution of the various cell granules seems to follow promptly the cessa- tion of centrifuging. In another set of experiments (ms. in press) carried on in this laboratory by Halcyon W. Hellbaum on centrifuged snake thyroid, the brief interval between the start of tissue fixation and its completion seemed to be sufficient for the nucleus to make an appreciable return toward the cell center; for with centrifuged thyroid tissues, in the halves of thyroid glands that were fixed in rapidly penetrating fluid such as Bouin’s, the nucleus was noticeably nearer the centrifugal margin of the cell than in the other halves of the same glands fixed in more slowly penetrating fluids such as Champy’s (bichromate-osmic-chromic acid) mixture. This would indicate that cytoplasmic elasticity enters as a significant factor. Immediately and for some time after centrifuging, the Golgi complex seems to be but a dense rounded drop of material at the centripetal side of the cell. It comes back to its network-like con- figuration very slowly, attaining it again only at the end of two or three weeks. It first loses its dense appearance by breaking up into a multitude of fine droplets, so small often as to give almost a powdery appearance. As time goes on the droplets become larger, apparently RECOVERY IN CENTRIFUGED PITUITARY 465 by more or less coalescence. In pituitary tissue which had been centrifuged for 20 minutes and then removed after an implantation of 14 days, the Golgi complex is usually back in the neighborhood of the nucleus though it is commonly still in a divided state of droplet-like particles. These are grouped more on one side of the nucleus although they may more or less surround it. A few examples of a return to the network type of Golgi body are to be seen in some of the basophile cells; still more occur in the oxyphiles. Even at the end of three weeks a considerable amount of granular Golgi material is still in evidence although the prevailing Golgi ap- paratus is of the network type. However, this slow return of the Golgi complex to the condition of a network is probably not deter- mined by the fact of earlier centrifugalization since control, non- centrifuged pituitary recovered after three weeks of transplantation may similarly show a considerable degree of granulation. By the end of the fourth week the transplants, both centrifuged and non-centrifuged, seem to be rather generally on the decline, and pos- _sibly final disintegration is approaching. This is particularly true of the basophile cells; their Golgi material shows much granulation or droplet formation. Figure 6 is a composite picture of two sections of recovered anterior pituitary which had been centrifuged for 20 minutes and then im- planted for seven days. That the tissue is anterior pituitary gland is evident. Certain of the basophiles (6) and the oxyphiles (0) have been lettered to identify them more readily. Of the 19 rat pituitaries which were centrifuged for an hour before implanting, only 2 survived. Operations at the end of two weeks revealed that 2 had been replaced by pus sacks and the remaining 15 had been resorbed. In the 2 bits of tissue which were alive at the end of two weeks the basophiles and oxyphiles seemed to be in a thriving condition. They did not differ in any noticeable particular from simi- lar implants of the 20-minute series. In one of the anterior pituitaries which had been centrifuged 20 minutes and then recovered after three weeks of implantation, a peculiar pituitary cyst was encountered (Figs. 3, 4). It strongly sug- gested the so-called colloid follicles so characteristic of the thyroid. Cysts of supposedly colloid-secreting cells are occasionally found in the pituitary glands of normal animals including man. The striking thing about the cyst in our transplant was the fact that it had ap- parently been stimulated to increased activity. The lumen was filled with a vacuolated, gluey looking mass and the secreting cells had be- come columnar instead of cuboidal, much enlarged, and in many cases they looked like goblet cells. This raised the question of whether 466 MM. ESGUYER AND PEAR Gia us they were typical colloid cells or were mucus-secreting cells, and also of just what the difference between a colloid and a mucous secretion is. With the Mallory triple stain the secretion gave the typical blue col- loid picture but when Mayer’s muci-carmine stain was used the picture was just as typically that of mucus-secreting cells with various of the cyst-wall cells staining as deeply as do characteristic goblet cells. This was true even in sections which had been stained in Mallory’s triple stain, and later destained, and stained again in Mayer’s muci- carmine. Thyroid follicles, however, subjected to the same treatment did not give the muci-carmine response. The pituitary cyst from a control animal (Figs. 1 and 2), on the other hand, did give the typical muci-carmine reaction. In Figs. 1 and 2 photographs of a cyst from the anterior pituitary of a control and supposedly normal rat are shown at different magnifica- tions. While the contents of the cyst lumens seem the same in the control and in the experimental animals, the secreting cells are very different in appearance in the two. Those from the control (Figs. 1 and 2) are cuboidal, those from the centrifuged and transplanted pituitary are conspicuously columnar in appearance. Both sets are ciliated. We have thoroughly canvassed the possibility that such appearances as those shown in our photographs (Figs. 1 to 4) might be due merely to striations in the mucin or in the shrinkage space between the mucoid mass and the epithelium lining, but careful inspection of these and other adjoining sections, both by ourselves and others in our laboratory accustomed to making cytological observation under high EXPLANATION OF PLATE [| 1. Photomicrograph of a colloid or mucoid cyst in the anterior pituitary of a normal rat, showing the ciliated epithelial lining of the cyst wall and the contents of the lumen. X 376. 2. Photomicrograph of a part of the foregoing cyst, under higher magnification. The cuboidal, ciliated epithelial cells are obvious. X 848. 3. Photomicrograph of a mucoid cyst in anterior pituitary material that had been centrifuged in an ultracentrifuge for 20 minutes, then grown in a young rat for three weeks, after which it was removed and sectioned. The epithelium which lines the cyst wall consists of very active, columnar cells, mostly ciliated, though some have the appearance of being goblet cells. X 376. 4. Photomicrograph of a part of the cyst shown in Fig. 3. Occasional smaller cells are observable between the bases of the columnar cells and the basement membrane. XX 848. 5. Photomicrograph of cells from a rat hypophysis centrifuged in an ultracentri- fuge for one hour. The deeply stained material of the nuclei is the chromatin sub- stance forced to the centrifugal side of the nucleus. In some of the nuclei strands of achromatic material, presumably linin, are visible stretching across the nuclear body, because still attached, seemingly, to the opposite nuclear wall. X 848. 6. Photomicrographs from two slightly different regions of a bit of anterior pituitary gland grown 7 days in a young rat after having been centrifuged for 20 minutes. Typical basophile (6) and oxyphile (0) cellsareto beseen. X 848. IRIBIClO NAMING JUN, (CHAIN TURIN OMG IID) Ie OMA AIR 467 PLATE I 468 MoE, GUYER AND PEAR E CeENUS powered lenses, convinces us that the objects in question are not artefacts but cilia. Whether the goblet-like cells of the transplant are transformed ciliated cells we are unable to say. And whether or not the wall cells of these cyst formations are to be regarded as unusual differentiations of pituitary tissue, as residual cells from an earlier epithelium, or migrant elements from the naso-pharyngeal region, we cannot say. The thing of chief interest in our present study is the enhanced secre- tory activity shown by the centrifuged cells. In the literature dealing with such hypophyseal cysts the nearest ap- proach to the types of cell shown in our preparations is that depicted by Rasmussen (1929) from sections of human hypophyses. Several of his pictures from human material look almost exactly like what we find in implanted centrifuged pituitary tissue from rats. SUMMARY Anterior pituitary tissue which has had its cell contents stratified through 20 to 60 minutes of rotation in an ultra-centrifuge with a dis- placement pull of about 400,000 times that of gravity, returns in many instances to normal appearance, displaying the characteristic basophile and oxyphile cells, when transplanted into young rats. An hour seems to be nearly the maximum time such tissues can be thus rotated and remain viable. Nuclei forced completely to one side of the cell rapidly resume their normal location near the cell center, with their displaced chromatin contents apparently restored to normal distribution. The Golgi apparatus, concentrated by the rotation into a liquid drop at the centripetal side of the cell, is the last part of the cell complex to resume its characteristic appearance, which is that of a network. A mucoid cyst with hypertrophied and extremely active epithelial wall cells, most of them ciliated, is described and compared with the type of cyst that is occasionally found in the normal pituitary. LITERATURE CITED DoRNFELD, Ernst J., 1936. Nuclear and cytoplasmic phenomena in the centrifuged adrenal gland of the albino rat. Anat. Rec., 65: 403. Guyer, M. F., anp PEARL E. Craus, 1932. Cell changes in the anterior lobe of the pituitary following cancer transplantation in rats. . Anat. Rec., 52: 225. Guyer, M. F., anp Peart E. CLaAus, 1933. Cellular constituents of the anterior hypophysis after uterine implants of carcinoma inrats. Anat. Rec., 56: 373. Guyer, M. F., anp Peart E. Craus, 1934. The Golgi apparatus in relation to vacuolation in the basophiles of the anterior pituitary of castrate and of cancerous rats. Anat. Rec., 61: 57. Guyver, M. F., AND PEARL E. Ciaus, 1935. Vacuolation in the anterior pituitary of castrated and of thyroidectomized rats. Anat. Rec., 61: (suppl.) 21. Rasmussen, A. T., 1929. Ciliated epithelium and mucus-secreting cells in the human hypophysis. Anat. Rec., 41: 273. COMPONENTS OF THE MITOTIC SPINDLE WITH ESPE-— CIA REE RENCE tO) iit CHROMOSOMAL AND ENTERZONAL FIBERS IN THE ACRIDIDA: E. ELEANOR CAROTHERS (From the Department of Zoology, State University of Iowa and Marine Biological Laboratory, Woods Hole, Mass.) INTRODUCTION While my own studies are concerned with specific components of the mitotic spindle, I believe the structure should be considered as a whole and I shall endeavor, accordingly, to summarize our knowledge of certain points.” Centrioles The beautiful demonstration of these structures in pole cells of Drosophila eggs which complete their development (Huettner and Rabinowitz, method published in 1933) must validate the reports of these structures in fixed material even for the skeptical. The following conclusions in regard to them seem to be warranted by careful work. Physical Characteristics —They are visible in living cells where they undergo vibratory motion and give evidence of being more rigid than the surrounding cytoplasm. They are usually spherical but their ability to vary in shape is well shown in the tree-crickets (Johnson, 1931) where the centrioles are minute spheres in the spermatogonia, compara- tively large V’s in the first spermatocytes and specialized rods in the second spermatocytes. In the protozoan, Barbulanympha, as shown by Cleveland et al. (1934), they are elongate, flexible rods of relatively large size. In all cases, they stain with iron-hematoxylin but in As- caris, at least, they do not give the Feulgen nucleal reaction (Carothers, unpublished). Position—Centrioles are located usually just outside the nuclear membrane but they may be near the cell membrane, as in the first spermatocytes of the tree-crickets, or intranuclear, as in certain of 1 This work was begun at the Zodlogical Laboratory of the University of Pennsylvania. 2A much fuller discussion of this subject is to be found in the third edition of E. B. Wilson’s excellent book, “The Cell in Development and Heredity.” Macmillan, 1925. 469 470 EK. ELEANOR CAROTHERS the Protista and in the male germ cells of Ascaris. During mitosis they come to lie on opposite sides of the nucleus, 180 degrees apart. Origin Normally, they are self-perpetuating structures which un- dergo division at each mitosis. In addition, a so-called de novo origin seems to have been demonstrated in certain instances. Wilson (1901) showed that in artificial parthenogenesis in sea-urchin eggs numerous cytasters each with a centriole-like body arise in one egg. These “central bodies’? and their cytasters divide and give rise to typical amphiasters which later are separated by cleavage furrows. Even more impressive is the fact that cytasters may unite with a nuclear aster to form a multipolar spindle with irregular distribution of the chromosomes to all poles followed by multipolar cleavage. A “de novo” origin may occur also in partheno-produced grasshoppers (Slifer and King, unpub- lished). Centrioles appear to be lacking during the maturation divi- sions of the eggs, yet the early cleavage cells of embryos arising from unfertilized eggs have typical centrioles. Here, however, there is the possibility of the survival of the egg centriole even if it does not function during meiosis. In the case of artificial parthenogenesis I would sug- gest that certain of the formed components of the cytoplasm may be able under especial circumstances to take on the function of centrioles. Distribution —Centrioles are present generally in the Metazoa. An exception is the odcytes in which they may either be lacking as in Ascaris or present as in Crepidula where Conklin (1902) showed that a division center is formed with each pronucleus. They are lacking in the higher plants, while in the lower plants (cycads, Ginkgo, bryophytes and pteri- dophytes) they are absent in the somatic cells and those of the early germ-line but appear as typical centrioles with asters at a definite stage in the life cycle, namely, the last divisions which give rise to the male sex-cells. Function—Aside from their prominent part as division centers in mitosis, centrioles are associated with motility as the blepharoplasts which form the locomotor apparatus of the sperm cells of the lower plants, the flagellate Protozoa and the flagellate cells of sponges as well as the sperm cells of animals. In the case of Barbulanympha, referred to previously, and spermatids, the centriole divides and the proximal one takes no part in forming the locomotor apparatus. This suggests, as has been pointed out by other workers, that centrioles may contain two components only one of which functions as a blepharoplast. Centrosomes This term denotes the regions of specialized cytoplasm surrounding the centrioles about which the constituents are arranged concentrically. COMPONENTS OF MITOTIC SPINDLE 471 Near the centrioles the groundwork is homogeneous or finely granular, and, in well-fixed material, the astral rays are seen to traverse it. Centrosomes vary greatly in size and complexity in different organisms and may be entirely abSent as in some of the hypermastigotes (Cleve- land et al., 1934). _ Astral Radiations and Central or Continuous Spindle Fibers Perhaps the most significant work to date on these structures is that just mentioned of Cleveland and his collaborators. They report, on what they consider absolutely conclusive evidence derived from both living and fixed material, that astral rays and central spindle fibers arise from the centrioles. Furthermore, and even more important, they state: “ The fibers of the achromatic figure . . . arise from the opposite ends of the same organelles (centrioles *) as the flagella; and they are like flagella in individuality, size, and appearance both in living and stained organisms; but, unlike flagella, they are not able to move inde- pendently.” Flagellates in roaches undergoing ecdysis form intracyto- plasmic bundles of flagella; these are used as a basis of comparison. Lucas (1932) described a ciliature anlagen for Cyathodinium in which the cilia are formed intracytoplasmically, before they are evaginated upon the surface. These two cases show clearly that definite, thread- like structures possessing elasticity and tenacity can be differentiated in the cytoplasm; hence, there is little reason to doubt that spindle fibers exist in living cells essentially as they appear in well-fixed material. Another well-established and critical fact, already mentioned, is that compound achromatic figures may be formed by the union of asters of different origin, due either to double fertilization where the two sperm may come from different males or to artificial fertilization, where cytasters may join with a normal amphiaster. Incidentally, tripolar spindles are the strongest evidence against the idea that the mitotic figure is due to a polarized field since lines of force form only between plus and minus poles. Another fact which does not fit with such a hypothesis is that in normal mitosis the rays from the two asters cross at the equator, whereas lines of force in a bipolar field are continuous from pole to pole. Both of these facts were recognized by early work- ers. J mention them now because several recent observers ignore them. Intergonal Fibers These are the fibers which connect the distal ends of daughter chromosomes for a brief time in anaphase. (See p. 482.) 8 The parenthetical expression is mine. 472 E. ELEANOR CAROTHERS Chromosomal Fibers (Half-spindle Fibers) These structures pass from a fixed locus on each chromatid to the poles. The shape of the chromosome is correlated with the position of these loci. Mitochondria Perhaps these elements should be mentioned since they surround the spindle densely, especially in insect meiosis (Bowen, 1920), and prob- ably prevent the fibers themselves from being seen in living cells. TECHNIQUE AND NOMENCLATURE The material was fixed in either strong Flemming or my own fixa- tive, the formula for which may be found in Dr. McClung’s Micro- scopical Technique. The stains used were Flemming’s tri-color and Heidenhain’s iron-hematoxylin. Most of the material was stained prior to 1920. In regard to terminology, I shall use the phrases ‘ chromosomal fibers’ or “chromosomal processes” interchangeably to designate the structures which pass from the chromosomes to the poles in preference to the expression “half spindle fibers” recently revived by Schrader, and I hope to show that the term “ point of fiber insertion ”’ should be changed to “ point of origin.” OBSERVATIONS My observations concern, chiefly, the chromosomal and interzonal fibers. The X-chromosomes of short-horned grasshoppers lend them- selves very readily to the study of chromosomal fibers for two reasons: 1. They are a step in advance of the other chromosomes in the first spermatocytes. Consequently, they reach the physical condition of euchromosomes in metaphase at late diakinesis and at the actual meta- phase are becoming diffuse for the following telophase. 2. The point of origin of spindle fibers varies in different species but is constant, typically, for the individuals of a given species; occa- sional exceptions to this rule occur, however. For example, most species have telomitic X-chromosomes but certain species of Trimero- tropis and allied genera have atelomitic X-chromosomes with rare indi- viduals which not only have both types but intermediate forms as well. Plate I, Fig. 1 shows partial complexes from three adjoining first spermatocyte metaphases of such an individual of T. coquilletti.. At COMPONENTS OF MITOTIC SPINDLE 473 C the X-chromosome (solid black) is telomitic, at B it is atelomitic with a process from each chromatid while at A, obviously, it is in a dilemma. The process for a terminal fiber is present, at the lower right end, the median processes are evident above, and in addition the opposite end has two subterminals going to opposite poles. The chromosome is flattened and distorted as though subjected to conflicting stresses. Even- tually such an X-chromosome will pass to one pole or the other without division, due apparently to the retraction of some of the processes. A count of 115 first spermatocyte metaphases gives the X-chromosome telomitic in 40, atelomitic in 50 and showing evidence of a conflict in 25. Perhaps the reader should be reminded that in the grasshoppers each egg which produces a male has only one sex-chromosome at the time of fertilization and that is contributed by the mother. All of these variations, therefore, have arisen in this individual. A change from telomitic to atelomitic or vice versa might be due to inversion of a segment of the chromosome. This, however, would not cause a change in the number of fibers. Conceivably, the extra fibers might be acquired by translocations involving spindle fiber loci from other mem- bers of the complex but in that case a cell in which the X-chromosome has fibers arising from three points should show two of the other chro- mosomes behaving eccentrically because of the loss. Such abnormali- ties are not present. Naturally a study of these multiple spindle fibers and of shifts in position of the fibers lead to a consideration of the origin and nature of the fibers which pass from the chromosomes to the poles. In 1917 I reported (p. 459) and figured (Plate 7, Fig. 45) a single X-chromosome in 7. fallax which showed several distinct fibers passing towards both poles of the spindle. At that time, I simply recorded it as an unmistakable fact with no surmises as to the cause. It is repro- duced as Fig. 2 of the present paper for the convenience of the reader and for comparison with other X-chromosomes which exhibit a similar phenomenon. Figure 3 is from the same individual as Fig. 1. Note that the apex of the V is marked by a group of delicate fibers. Schrader has empha- sized the existence of a like condition in other forms. Figure 4 is from Derotmema laticinctum, a species which usually has no atelomitics. Here, I think, the solution of the problem becomes self-evident when we compare this X-chromosome with the large tetrads in diakinesis just below it (Figs. 6 and 7). Delicate pseudopodial-like processes which contact either those from other chromosomes (Fig. 7, 4, B) or the nuclear membrane where they form enlarged or even plate-like 474 E. ELEANOR CAROTHERS contacts (Fig. 8) are put forth at this stage,* apparently, from every chromomere except those which resemble the X-chromosome in pre- cocity. These processes can be retracted almost instantly as is attested by the chromosomes of contiguous cells which show transitions such as are illustrated in Figs. 6, 7 and 11. This cyst was fixed, fortunately, at the moment when the cells were passing from late diakinesis to early metaphase; in a few cells the nuclear membrane had disappeared and chromosomes, only slightly smoother than the tetrad at the right in Fig. 11, were forming the equatorial plate. Short processes comparable to those on the X-chromosomes shown in Fig. 9 are evident, occasionally, on such tetrads at the spindle fiber loci. The whole testis appears to be normal so that I believe the variations shown to be the regular occur- rence. In fact, tetrads of like density, as denoted by staining reaction, vary within the same nucleus in regard to the degree of contraction or extension of their processes and even the parts of a single tetrad ex- hibit this diversity, as may be seen in Fig. 6, where the ring-shaped tetrad has some processes long and attenuated and others short and blunt as though they had been retracted suddenly. Figure 7A is an end view of one arm of a tetrad which illustrates the fact that the proc- esses are sent out not only at right angles to the plane of the next division but in various directions. Club-shaped ends on processes which have severed their contacts, such as are illustrated at 7B, suggest that external stimuli applied to the process itself may cause retraction as well as intrinsic forces which presumably usually cause their withdrawal. Figure 9 shows two late diakinesis X-chromosomes in their actual position relative to each other. Note that they are as fully condensed as the tetrads are at metaphase. The single process on each is in the usual position for the spindle fiber of the X-chromosome in this species and in these instances is certainly of chromosomal origin. Figure 10 from the same cyst represents an X-chromosome which has maintained an end-to-end association with a tetrad up to this late stage, and I sus- pect such an occurrence accounts for the telomitic X-chromosomes in this species. Figure 12 is from an earlier cyst. Approximately half of the X-chromosomes are U-shaped and half rod-shaped in this stage. At first thought one might assume that such forms are the direct pre- cursors of the V and rod-shaped X-chromosomes, respectively, of the metaphase, but this is improbable because in all species of short-horned grasshopper which I have observed, cytologically, the X-chromosomes 4 To obtain these processes at their best a number of precautions are necessary. First the animal should be killed in a suitable manner such as the application of xylol to the spiracles, next the testes should be removed from the body with delicacy and care in order to avoid unnecessary shock and quickly transferred to an abundance of good fixative. Finally the stain must not be extracted too much, the processes are chromophilic but very attenuated. COMPONENTS OF MITOTIC SPINDLE 475 form close U’s at about this stage and a little later thick rods, regard- less of whether the X-chromosomes at metaphase are telomitic or atelomitic. With the sex-chromosomes at their maximum density during late diakinesis and beginning to become diffuse just before the metaphase and with the roughened diakinesis tetrads becoming more compact as they approach the metaphase, a brief period occurs when the two types of chromosomes present similarly roughened contours. Such a stage is shown in Figs. 13 (X-chromosome below), 14 (X-chromosome above), 15 (X-chromosome only) and 11 (tetrads). Compare also the tetrad in Fig. 14, on which corresponding processes are still evident on sister chromomeres, with the sex-chromosome from Circotettix ver- ruculatus shown in Fig. 5 which has put forth a process toward both poles for each of five aggregates of chromomeres. The foregoing observations seem sufficient to raise a strong pre- sumption that the chromosomal fiber of the sex-chromosome is a process put forth by the chromosome itself.° Let us now consider some later stages; first, the metaphase spindle, next conditions which immediately precede the metaphase, and finally the anaphase. Naturally, many of the points which I shall mention have been observed by other cytologists. Four critical features, how- ever, seem to have escaped notice. Attention will be called to them in the following pages. In lateral views of well-formed metaphase spindles, two kinds of fibers are evident on a morphological basis. Thecontinuous fibers, which in the grasshoppers do not form a central spindle but are intermingled with the other spindle components, are slender, of uniform thickness and, in fixed material, somewhat wavy. The chromosomal fibers run a straighter course and are much heavier at their loci of émergence from the chromosomes, from which they taper gradually towards the centrioles—thus, they vary in thickness throughout their length. These differences are illustrated in Fig. 18 from a spermatogonium of Circotetiix rabula and Figs. 24 and 25 from first spermatocytes of Trimerotropis citrina. Convinced that the chromosomal processes pos- 5 At this place I should like to call attention to another point. The first three drawings in Fig. 11 are of the same tetrad in the size series from three cells. In addition to showing the transition of contour already mentioned, they are of interest on account of the characteristic spherules which mark this tetrad in late diakinesis and which probably are either secretory or excretory products. They are usually embedded in the surface chromatin but may be at a considerable distance from the parent chromosome with which they are still connected by what appear to be delicate fibrils. Later these connections may disappear, leaving some of the spherules free in the cytoplasm where for a time they continue to give the Feulgen reaction. I believe, for reasons which I hope to develop in a later paper, that they are metabolic products and not true chromatin in process of elimination, E. ELEANOR CAROTHERS COMPONENTS OF MITOTIC SPINDLE 477 sess considerable elasticity by evidence such as is represented by Figs. 23 and 29, which would be difficult to explain on any other basis, I expected them to become thicker as they became shorter in anaphase. Such, however, is not the case normally. The processes, beginning with the heavier basal parts, apparently, are reincorporated into the anaphase chromosomes as they move towards the poles, so that as they EXPLANATION OF PLATES All of the figures were drawn with the aid of a camera lucida at a magnifica- tion of 2,800 diameters. They were reduced one-third in reproduction. EXPLANATION OF PLATE I Fic. 1. Partial complexes from three contiguous first spermatocyte meta- phases, lateral views. X-chromosome, solid black. At b it is atelomitic, at c telomitic and at a the telomitic process is evident at lower right end which is of normal thickness, the apex of an atelomitic V with two processes is indicated above while the other end is flattened and has processes towards each pole. Trimerotropis coquilletii. Fic. 2. Partial complex of a first spermatocyte lateral view, T. fallax. Note the three prominent processes of the X-chromosome. Fic. 3. Similar to preceding. X-chromosome with numerous fibers or processes. TJ. coquilletti. Fic. 4. Similar to Fig. 3. Derotmema laticinctum. Compare this X-chro- mosome at metaphase with the diakinesis tetrads shown in Figs. 6, 7, and 8. Fic. 5. Similar to above, except the tetrads also are in solid black. Cuircotet- tix verruculatus. Fics. 6-15. Late first spermatocyte prophases, from T. coquilletti, showing transition of X-chromosome from the smooth contour characteristic for it in diakinesis and for the tetrads in metaphase (Figs. 6, 7, 8, and 12) to a roughened contour in very late diakinesis (Figs. 13, 14, 15). Note the long pseudopodial-like processes of the tetrads (Figs. 6, 7, 8) which are withdrawn at the approach of metaphase (Figs. 11, 12 and 13). Fic. 6. X-chromosome and tetrads. Note blunt processes of the ring at left where the fibers have been retracted. Fic. 7. Similar to Fig. 6. End views of arms of two tetrads at a and D. Processes go out in all planes and contact those of other chromosomes. Note the knobbed ends of retracting processes at b. Fic. 8. Similar to above. Processes of tetrad in characteristic contact with nuclear membrane. Fic. 9. X-chromosomes from two adjacent cells. Note the process on each and compare with spindle (chromosomal) fibers of tetrads shown in Fig. 5. Fic. 10. X-chromosome from the same cyst which will probably be telomitic at metaphase. Fic. 11. Tetrads from a cyst where metaphase plates are forming in some cells. The droplets associated with the three examples of one tetrad are char- acteristic. They may be either secretion or excretion. Fic. 12. Nuclei from two adjacent cells, earlier diakinesis. X-chromosomes at this stage are either rod or U-shaped in about equal numbers. Fic. 13. X-chromosome and tetrad approaching each other in roughness of outline. Very late diakinesis. Fic. 14. The same, a trifle later stage. Fic. 15. X-chromosome in very late diakinesis. Compare with X-chromo- some in metaphase shown Fig. 1, c. 478 E. ELEANOR CAROTHERS shorten they become more delicate. The linear order of their constitu- ents is thus preserved and there is no loss to the chromosome. It seems to follow that the chromatids move apart through some inner mechanism and that the chromosomal processes act more as guides than as “traction fibers.” If we turn now to premetaphase conditions, we find that as early cytologists observed, the nuclear membrane disappears first on the side traversed by the separating centrioles. Some early workers also noticed that the fibers from the two asters intersect in the plane of the equator at early metaphase, but later curve inward and become part of the central spindle. A characteristic which seems to have escaped men- tion, however, is that in this series of actions one side is still slightly in advance of the other (see Figs. 19 and 24). This condition applies also to the arrangement of the chromosomes on the spindle. The first of the critical features mentioned above is associated with this fact. (1) In lateral views of very early metaphases when the equatorial plate is form- ing, the chromosomes on one side of the spindle may have typical chromosomal fibers extending from the chromosomes to the poles, while on the other side identical fibers (processes) extend from the chromo- somes in the general direction of the poles, but are not yet oriented towards the centrioles. Such a case is illustrated in Fig. 16. The lower process on the small rod at the left is of especial interest because it bends outward towards the cell membrane crossing that of the larger, underlying tetrad. These structures are identical so far as they go with the unquestionable chromosomal fibers on the two tetrads at the right. Their homology with the processes on the sex-chromosomes EXPLANATION OF PLATE II Fic. 16. First spermatocyte metaphase, lateral view. Chromosome fibers reach the poles on one side, forming and not yet oriented on the other. Fic. 17. Similar to above. Note separation of the processes; also, that they are thicker at point of emergence from chromosomes. Fic. 18. Spermatogonial metaphase, lateral view, spindle very uniform. Chromosomal fibers thicker and straighter than continuous fibers. Circotettix rabula. Fic. 19. Second spermatocyte, lateral view, from same individual. Processes pass straight out from chromosomes until they approach the poles, then bend abruptly towards them. Fic. 20. Two tetrads where the point of spindle fiber origin is different for each homologue. T. fallax. Fic. 21. Six tetrads in anaphase illustrating interzonal fibers. T. fallax. Fic. 22. Three tetrads in metaphase showing relation of chromomeres to chromosomal fibers. TJ. fallax. Fic. 23. Tetrad displaced in sectioning. Chromosomal process distorted but unbroken. T. fallax. COMPONENTS OF MITOTIC SPINDLE 479 121 PEATE i 480 E. ELEANOR CAROTHERS shown on Plate I, especially Fig. 9, and on the tetrad at the right in Fig. 6 seems almost equally certain. (2) The second point which I consider critical is the sudden in- bending of such processes, which have tended to pass straight out from their point of origin in the chromosome, when they reach the vicinity of the centrioles (Figs. 19, 24, and 25). This abrupt inbending of the fibers lends some support to older views recently supported by Cleveland (1934, 1935a, 1935b) that fibers which lead from the chromo- somes to the centrioles have a dual origin. I am convinced, however, that in the grasshoppers they are wholly chromosomal in origin; partly because, even near the poles, they are always markedly heavier than the fibers of the central spindle. See especially the dyad at the right in Fig. 29, a first spermatocyte anaphase from Dittopternis. Their change of direction probably indicates that they have reached a region where the centrioles exert a more active influence. Such figures occur only in early metaphase where one side of the spindle is still slightly in ad- vance of the other. The spindle later becomes symmetrical, as shown by Fig. 18 from a spermatogonium of C. rabula. (3) The third suggestive characteristic is the behavior of the chromosomal fibers, one for each chromatid, often evident on each dyad in the first spermatocyte metaphase (Figs. 17, 21, and 23).° They taper gradually towards the centrioles and may diverge for some distance after leaving the chromosome, later converging as they approach the centrioles (Fig. 17). This behavior is perfectly comprehensible if they are processes put forth, one by each chromatid, which extend inde- pendently, in the general direction of the poles. On such a view they might diverge considerably until they come under the influence of the centrioles, when the sudden inbending would naturally cause them to converge. It need hardly be pointed out that this behavior, as well as that described under “ 2,” is wholly incompatible with any theory which EXPLANATION OF PLATE IIT Fics. 24 and 25. First spermatocyte metaphases, lateral views, T. citrina. Individual represented in 24 killed with potassium cyanide, that in 25 with xylol. Such differences in spindle configuration characteristic for these two methods of killing. Fic. 26. Metaphase, polar view, from same individual as Fig. 24. Fic. 27. Similar to above, from same individual as Fig. 25. Chromosomes form a more open plate. Fic. 28. Late anaphase, lateral view, chromosomes surrounding the poles. Same individual as Figs. 24 and 26. Spindle .69 of the length attained at meta- phase when animals are killed with xylol. Fic. 29. One pole of an anaphase. One chromosome and its fiber in plane of section. Dittopternis turbata. ® Henking (1890) noted this condition in first spermatocytes of Pyrrhocoris and recognized it to be associated with the valence of the chromosomes. COMPONENTS OF MITOTIC SPINDLE 481 PLATE III 482 E. ELEANOR CAROTHERS regards the spindle fibers as artifacts, produced in a homogeneous spindle substance by differential coagulation along hypothetical lines of force at the time of fixation. Neither does the observed behavior fit Bélar’s hypothesis (1929) that the chromosomal fibers are formed by a mucous secretion from the insertion points of the chromosomes which flows along underlying continuous fibers. (4) The fourth feature is the actual presence of chromatin along the chromosome fiber (Figs. 22 and 23, tetrad at the left). The re- semblance of these structures to satellites and constricted chromosomes is self-evident. Furthermore, the thicker basal parts of the chromo- somal fibers often show the chromatic stain (hematoxylin, safranin and even Feulgen). In Fig. 16, for instance, the fibers on the tetrad at the right retain the safranin for nearly half their length. The same is true of the heavier portions of the fibers in all of the figures. As was stated earlier, all of the slides used with the single exception of one of Dit- topternis (Fig. 29), were stained more than ten years ago and differ- entiated primarily for the chromosomes. Passing finally to the anaphase, let us briefly consider the interzonal fibers. Two or possibly three distinct structures are confused under this term in the literature. Mark (1881, p. 198) introduced the term in the following sentence: “ Between the two zones of thickenings are stretched delicate nearly parallel threads, which I shall designate as interzonal filaments.’ The “two zones” are the anaphase groups of chromosomes, though the word chromosome had not yet been coined. Later (p. 230) he definitely states that the “. .. free ends (of the interzonal filaments) terminate in the lateral zones of spindle-fiber thickenings.” Hermann (1891), on the contrary, concluded, from a study of mitosis in salamanders and other forms, that the connecting fibers apparent in anaphase are simply the fibers of the central spindle revealed by the separating chromatids. Hermann’s view has been widely accepted by cytologists who failed to recognize the fact that in addition to, and morphologically quite distinct from, the fibers of the central spindle, there are connections frequently between the “ free” ends of the chromatids as they separate in anaphase. These connections are thickest just as the ends begin to separate and at this time take the chromatin stain. As the separation increases the connection becomes more attenuated and, consequently, destains more rapidly so that in preparations differentiated for metaphase chromosomes all of the stain may be removed. A series of such stages is shown in Fig. 21 from first spermatocytes of T. fallax. The sixth tetrad from the left shows a con- nection which has persisted longer than usual. This tetrad is of further COMPONENTS OF MITOTIC SPINDLE 483 interest on account of the way in which the chromatin of the upper dyad is drawn out along the connecting fiber, a behavior directly comparable to the extension of chromatin along the chromosomal processes illustrated in Fig. 22. Ordinarily, in the Acridide, the interzonal connections sep- arate (I do not believe that they break) near the mid-region at about the stage shown in the third tetrad from the left and are retracted each end into its own chromatid. To me, the action seems to be ex- actly what one would expect under the existing conditions; namely, flexible, viscous, protoplasmic rods are separating in such a manner that the final contact is end to end. As these ends are separating, it is readily conceivable that viscous protoplasm belonging to each might be drawn out between them to a varying degree, depending upon such factors as the rapidity of the movement and the viscosity of the chro- matids. Whether or not interzonal fibers occur in normal, actively dividing cells may be questionable, but they certainly do occur in living cells which are sufficiently normal to complete division as well as in fixed preparations. If my conception is correct, they may well be lacking both in entire groups of organisms and in special cell divisions where the rigidity of the metaphase chromatids is adequate to prevent adhesion, or where the chromatids are separated widely. For instance, in the second maturation division of the short-horned grasshoppers, interzonal fibers are lacking entirely. This condition I believe to be due to the relatively wide separation of the chromatids when they come into the plate for this division. In my material I find no evidence of a chromosomal sheath or of tubular connections between separating chromatids in anaphase such as Schrader describes in certain Hemiptera. In Leptocoris first sper- matocyte anaphases fixed in Flemming’s strong solution, I have found no interzonal fibers but a pathway such as the dense daughter chromo- somes might leave when separating from each other in a fairly viscous cytoplasm if the clearer, more fluid part of the cytoplasm flows in be- tween them as they move apart. This condition is essentially the same as Ellenhorn (1933, p. 300) reported in living cells from Tradescantia anthers, where the cytoplasmic viscosity had been increased by pres- sure sufficient to cause a reversible gelation which did not stop the mitotic process. Schrader (1934, p. 520) states that Ellenhorn, like himself, interprets the interzonal connections as tubes. Fundamentally, their conceptions are entirely different since Ellenhorn considers the internal connections to be transient canals through viscous cytoplasm and Schrader (1932, p. 537) believes them to be a chromophobic outer layer of the chromosome which is pulled out in the form of a tube be- tween the separating daughter chromosomes and which may persist 484 BH, ELEANOR CAROTHERS with the lumen obliterated until the formation of the spindle for the succeeding division. A further phenomenon which must be concerned with the structure of the spindle, although I do not quite see its mechanism, should be considered here and that is the variation in size of the spindle which may be produced experimentally. Figures 24 to 28 are from first spermatocytes of T. citrina. Figures 24, 26, and 28 are from the same insect which was killed with cyanide fumes. Figures 25 and 27 are from another individual which was killed with xylol applied to the spiracles. The differences illustrated are characteristic. Cyanide gives short spindles with the centrioles well removed from the cell membrane and much of the cytoplasm apparently not involved, while xylol gives approximately twice as much distance between the centrioles which are consequently near the cell membrane with the spindle occupying nearly the entire cell. Correspondingly, the chromosomes are more crowded together in a smaller equatorial plate in the former case, as may be seen by comparing Figs. 26 and 27 from the same two insects. Even at the completion of the anaphase the distance between the poles in cyanide-killed animals is only slightly more than two-thirds of what it is in metaphase for an animal killed with xylol (Fig. 28)." Some question has existed as to which is the normal condition but recently an individual of Dittopternis killed with xylol gave both types of spindles. The short ones occur in a limited region which was injured, evidently, when the testis was removed. The cells have coarsely granu- lar, muddy staining cytoplasm and in extreme cases the chromosomes had fused into a pycnotic mass characteristic of dying cells. Clearly, then the large spindles are the normal ones. When one considers that less than five minutes previous to fixation both animals, from which the figures were taken, were presumably in the same condition so far as their mitotic spindles were concerned,* one marvels at a mechanism which permits such rapid rearrangement. If the chromosomal fibers are processes from the chromosomes, their rapid retraction is understandable. As to the astral rays and continuous spindle fibers, one can imagine elongate molecules to be involved. A. R. Moore (1935) has produced evidence for the existence of such elongate ' prestructural elements, molecules or micelles, in protoplasm by showing that the plasmodia of Physarum will pass through moist walls of hard filter paper with pores about 1 micron in diameter uninjured while they 7™McClung (1918a) reported the relation of this condition to the method of killing the animal but illustrations were not given. 8 This presumption is based on some fifty specimens killed with cyanide fumes and thousands with xylol. COMPONENTS OF MITOTIC SPINDLE 485 are killed if forced through silk gauze with pores less than 200 micra in diameter. DISCUSSION The observations of the earliest investigators in any given field frequently are strictly accurate and amazingly detailed. Probably be- cause they alone approach the subject with a completely open mind. It is interesting to note that Flemming (1882, Pl. VI, Fig. 3b) shows a late anaphase of a living “ Leydischen schleimzelle” of a sala- mander larva with achromatic threads, which he states are clearly visible. The general accuracy of Flemming’s observations is attested by his figures of mitosis from fixed material which far excel those of many later cytologists. Certainly, then, his statement that he saw achromatic threads, which his figure shows to be in the interzonal region, is to be taken at face value. Nevertheless, they might have been any one of three structures: (1) continuous spindle fibers, although this is not probable, (2) interzonal fibers which as drawn out parts of the chromo- somes should be visible in living cells, since the chromosomes themselves are visible; (interzonal fibers, however, seldom persist until late ana- phase) ; (3) mitochondria which are present in abundance at this stage in most forms and are readily visible in living cells. Belar (1929, Pl. I, Figs. 14 and 15) shows photomicrographs of what is probably a similar condition in living first spermatocyte ana- phases of Chorthippus lineatus. He interprets the apparent fibers as streaming mitochondria. On the other hand, the excellent ultraviolet photomicrographs of a similar stage in second spermatocytes of another grasshopper, Melanoplus femur-rubrum, by Lucas and Stark (1931), show nothing comparable to fibers in the interzonal region. These photomicrographs have been cited by some investigators as evidence that the mitotic spindle is homogeneous. In reality, they are evidences only that any structures which are present are completely permeable to ultraviolet rays in this species. Mitochondria almost certainly were present. In addition, the only anaphase figures are of second sperma- tocytes (Plate III, Figs. 27, 28) where, as I have mentioned previously, interzonal fibers are lacking. Incidentally, it should be noted that the refractive index of living chromosomes varies greatly in different species of Acridide. Consequently, some are more favorable than others for the study of mitosis in living cells. Knowledge of the various spindle components has been augmented greatly by the work of Cleveland and his collaborators on the flagellate parasites of the wood roaches. There is essential agreement in our conclusions. The chief point of difference is in regard to the chromo- 486 E. ELEANOR CAROTHERS somal fibers which Cleveland et al. believe to be of dual origin; the intranuclear part derived from the chromosomes and the extranuclear portion from the astral rays. In the grasshoppers, I am convinced that they are entirely chromosomal in origin. While such a divergence would not be surprising in organisms so widely separated, phylogeneti- cally, as the flagellate Protozoa and the grasshoppers, the possibility that the extranuclear chromosomal fibers in these flagellates are put forth from the “chromatin knobs” on the nuclear membrane would explain some of the difficulties encountered on the assumption of Cleveland and his coworkers. For example, describing the anaphase movement of the chromosomes, they state (p. 234) “... the extra- nuclear chromosomal fibers are greatly shortened, while the intranuclear chromosomal fibers are not greatly altered (Figs. 53, 56-58). The shortening of the extranuclear chromosomal fibers is difficult to under- stand. If, as all other observations indicate, they are astral rays that have been converted into chromosomal fibers by becoming fastened to the knobs on the nuclear membrane, why do they shorten while the unattached astral rays do not?” If the knobs send processes (fibers) through the nuclear membrane to the centrioles such fibers would not be expected to behave as do the astral rays, since their origin would be from the chromosomes. The paragraph just quoted continues: “ The unattached astral rays, as a rule, are considerably longer than the longest attached ones (extranuclear chromosomal fibers). This is due to the increased length of the astral rays after the intra- and extranuclear chromosomal fibers are united and will account for the fact that the extranuclear fibers sometimes appear slightly larger than the astral rays, 1f the chromosomal fibers increase in thickness as the astral rays increase in length.” This explanation of the fact that the chromosomal fibers are thicker than the astral rays seems rather far-fetched. If the latter originate from the centrioles and the former from the chromo- somes, correspondence in thickness would not be expected. As is well known, certain workers believe all spindle fibers to be artifacts. Schrader (1932, 1934) reviews the work of a number of such investigators in a very fair and sympathetic manner and gives an adequate bibliography. Schrader’s answer to such views is a series of experiments on the mitotic figures of higher forms which show that the components of the half spindle may be bent, independently of each other, by centrifuging and that such metaphases will complete their division; thus disproving, so far as the chromosomal fibers are con- cerned, the contention that the fibers seen in fixed preparations are ~ artifacts caused by the differential coagulation along lines of force of a homogeneous spindle substance. COMPONENTS OF MITOTIC SPINDLE 487 If one grants for the sake of argument that lines of force could cause such coagulation artifacts to result from fixation, one would at least expect them to be continuous from pole to pole and of uniform diameter, whereas the outstanding characteristics of chromosomal fibers are that they taper from their points of emergence from the chromosomes towards the poles, and that they are not continuous from pole to pole. In short, a great many of the arguments put forward by those who question the actual existence of spindle components of all categories would never have been advanced if their proponents had had adequate training in the study of mitosis in well-fixed material before undertaking the more difficult study of mitosis in living cells. Perhaps a brief statement of my conception of a chromosome is in otder. In the first place it is an individualized mass of living substance (protoplasm). We speak of the protoplasm of an amceba as being specialized into ecto- and endoplasm and the evidence indicates that the protoplasm of a chromosome is specialized in much the same way. The cortical part must meet its environment adequately. During meta- phase, anaphase and early telophase the immediate environment is the cytoplasm and the condensed chromosome exposes the least possible surface. These stages are concerned primarily with increase in cell number and not with the welfare of the chromosomes as individuals. In late telophase the chromosomes form the nuclear membrane (Wen- rich, 1916, pp. 86-88). That the chromosomes are the determining factors in this process is further indicated by the fact that in the sperma- togonial telophases the sex chromosome of the short-horned grass- hopper forms an almost separate and relatively much more commo- dious “vesicle”’ for itself while the euchromosomes are rebuilding the nuclear membrane.? From the late telophase through the “ resting period” and up to the end of the prophase the immediate environment of the chromosomes is the nucleoplasm. In this sheltered situation they expose a greatly increased amount of surface and probably reach the apex of their functional activity and individual well-being. Even this protected environment changes as tissue differentiation progresses and this may well be one of the factors which limit the number of cell divisions, and consequently, the size of a given organ. Support for such a view is found in the fact that in mature tissue cells which resume mitotic activity to repair an injury, the cytoplasm first dedifferentiates. In brief, a chromosome is a living entity clearly capable of growth, of reproducing itself by longitudinal fission, and of amceboid movement. Finally we come to the crucial point. Can the position of the 9McClung (19185) noted the increased diffusion and presumably greater activity of the sex-chromosome during the spermatogonial interphases. 438 E. ELEANOR CAROTHERS chromosomal fibers change without inversions or translocations of parts of the chromosomes ; in other words, can new loci arise and suppress the older ones? If so, what is the mechanism involved? The association of the chromosomal fibers with a fixed region of the chromosome is one of the most constant features of chromosomal organization. Cyto- geneticists, however, have long known that experimental breakage and fusions of chromosomes due to irradiation, abnormal temperatures and unknown intrinsic causes may result in shifts in the location of the chromosomal fiber region so that it does not invariably occupy the same position in a given chromosome; but in such cases, the region maintains its characteristics whatever its location. The first well-founded case of an association of the chromosomal fibers with a different region (chromomere in this instance) is that noted by Wenrich (1916) in Phrynotettix. His figure 65h—m illus- trates six tetrads of his type C,. Those shown at h-j are dividing equationally, those at k—-m reductionally. In the former case the ends marked by the polar granules (Pinney, 1908), as shown by a comparison with the earlier stages, are as usual oriented towards the poles. In the latter case the large distal granule, characteristic of the free end of the larger homologue in the spermatogonial telophases, is oriented towards one pole and the erstwhile free end of its homologue is, of necessity, directed towards the other pole. In other words, the chromo- somal fibers have shifted from the ends of these homologues marked by the polar granules to what in the spermatogonia were the distal or free ends, one of which is marked by a pronounced “ distal granule.” Wenrich found this change to have taken place in approximately fifty per cent of the 928 first spermatocytes recorded in the individual under consideration. His close study of the earlier stages of this tetrad renders it highly improbable that any inversion of parts of a chromo- some where the regions are so well-marked could have taken place in half of the cells without detection by so competent an observer.’° Wenrich also notes (p. 84) the critical point, that the polar granules, located at the spindle fiber ends of the euchromosomes, and the large dis- tal granules of tetrads “B” and “C,” as well as the whole of the sex- chromosome become roughened by the metaphase of the first sperma- tocyte division (Plate IX, Figs. 99 and 100) and suggests that some 10 Schrader, F. (1936), in a paper which appeared after this manuscript was sent to press, concludes on the basis of structure and staining reaction that the “kinetochore ” is quite distinct from an ordinary chromomere. I have no doubt that the degree of specialization varies in different organisms. However, I cannot agree with his suggestion that the polar granules of Phrynotettix (Pinney, 1908, Wenrich, 1916) are not “kinetochores” simply because tetrad “B” possesses a distal as well as a polar granule. In my opinion this is exactly the mechanism by which tetrad “C” shifts its spindle fiber locus. COMPONENTS OF MITOTIC SPINDLE 489 common physical or chemical properties underlie this correspondence in behavior. One physical characteristic marks all of the above structures as well as sex-chromosomes, chromoplasts and precocious chromomeres in general; their rhythm of condensation and diffusion is different from that of the rest of the chromatin, with the result that at the metaphase they are already beginning to become diffuse. The X-chromosome of Stauroderus scaleris has reached a vesicular telophase condition at the first spermatocyte metaphase (Corey, 1933). Likewise, the X-chromo- some of Notonecta indica goes into the first spermatocyte metaphase in a diffused condition and with multiple spindle fibers as was shown by Browne (1916, Pl. 6, Fig. 95). Carlson (1936, p. 132) gives a good description of the process of diffusion of the “ megameric” chromo- somes in the cryptosome stage in seven related genera of short-horned grasshoppers. If precocity of definite chromomeres ‘is associated with the ability to begin putting forth processes at the metaphase, a change (mutation) in the rhythm of activity of certain chromomeres might result in some which are usually relatively quiescent at the metaphase, speeding up and sending out processes capable of becoming chromosomal fibers with a resultant struggle for supremacy such as is exhibited by the X-chromosomes with multiple chromosomal fibers illustrated in Figs. 1 to 5. In my opinion, a specialized chromomere of each chro- matid normally sends out a fine pseudopodial-like process at the meta- phase which becomes the chromosomal fiber, but a change of rate of condensation and diffusion may result in a shifting of this function to a more precocious chromomere. SUMMARY The evidence presented is believed to justify the following conclu- sions: (1) Chromosomal fibers (half-spindle fibers) are in reality pseudopodial-like processes. They are put forth by definite chromo- meres, therefore normally are fixed in position. (2) Due to mutation, using this term in a broad sense, some other chromomere may assume this function with a resultant change in position of the spindle processes which does not involve either inversion or translocation of chromosomal segments. (3) At anaphase the processes, beginning with the heavier basal parts, are reincorporated into the sister chromatids as they move towards the poles. (4) The interzonal fibers behave as one would expect under the existing conditions; namely, flexible, viscous proto- plasmic rods are separating in such a manner that the final contact is end to end. As these ends separate the viscous protoplasm belonging 490 E. ELEANOR CAROTHERS to each is drawn out between them in varying degree depending on such factors as the rapidity of the movement and the density of the chro- matid. (5) Both chromosomal fibers and interzonal fibers are integral parts of the chromosomes and return to them during anaphase without loss of chromatin or linear derangement of their constituents. The work is based on six species belonging to the following four genera of Acridide; Trimerotropis, Circotettix, Derotmema and Brachystola. In addition, the paper contains a summary of what are believed to be well-established facts concerning the various components of the mitotic spindle. BIBLIOGRAPHY Berar, Kart, 1929. Beitrage zur Kausalanalyse der Mitose. II. Untersuchun- gen an den Spermatocyten von Chorthippus (Stenobothrus) lineatus Panz. Arch. f. Entw.-mech., 118: 359. Bowen, R. H., 1920. Studies on insect spermatogenesis. I. Biol. Bull., 39: 316. Browne, Etuet N., 1916. A comparative study of the chromosomes of six species of Notonecta. Jour. Morph., 27: 119. Carson, J. G., 1936. The intergeneric homology of an atypical euchromosome in sareiil closely related Acrididz (order Orthoptera). Jour. Morph., 59: 123. CaroTHErS, E. ELeanor, 1917. The segregation and recombination of homologous chromosomes as found in two genera of Acridide (Orthoptera). Jour. Morph., 28: 445. CLEVELAND, L. R., S. R. Hatt, EvizaABeTH P. SANDERS AND JANE Co Lier, 1934. The wood-feeding roach Cryptocercus, its protozoa, and the symbiosis between protozoa and roach. Memoirs Am. Acad. Arts and Sci., 17: 185. CLEVELAND, L. R., 1935a. The centriole and its rdle in mitosis as seen in living cells. Satenee, 81: 598. CLEVELAND, L. R., 1935b. The centrioles of Picidoteichenyaupna and their rdle in mitosis. Biol Bull., 69: 46. Conkiin, E. G., 1902. Karyokinesis and cytokinesis in the maturation, fertiliza- tion and clevage of Crepidula and other Gastropoda. Jour. Acad. Nat. Sci. Phila., 12: Corey, H. Irene, 1933. Chromosome studies in Stauroderus’ (an orthopteran). VOLS MOAN ty S53 OUS! ELLENHORN, JAKosB, 1933. Experimental—photographische Studien der lebenden MiGs VGOUGCINE fs AA fn, AUS Fasion FLEMMING, F., 1882. Zellsubstanz, Kern und Zellteilung. Vogel, Leipzig. HENKING, H., 1890. Untersuchungen tiber die ersten Ewtwicklungsvorgange in den Eiern der Insekten. II. Uber Spermatogenese und deren Beziehung zur ae bei Pyrrhocoris apterus L. Zeitschr. f. wiss. Zodl., 51: 685. HERMANN, F., 1891. Beitrag zur Lehre von der Entstehung der karyokinetischen Spindel. Arch. Mikro. Anat., 37: 569. Huettner, A. F. anp M. RasrnowitTz, 1933. Demonstration of the central body in the living cell. Science, 78: 367. Jounson, H. H., 1931. Centrioles and other cytoplasmic components of the male germ cells of the Gryllide. Zeitschr. wiss. Zool., 140: 115. Lucas, F. F. anp Mary B. Stark, 1931. A study of living sperm cells of certain grasshoppers by means of the ultraviolet microscope. Jour. Morph., 52: on COMPONENTS OF MITOTIC SPINDLE : 491 Lucas, Mirtam §S., 1932. The cytoplasmic phases of rejuvenescence and fission in Cyathodinium piriforme. II. A type of fission heretofore unde- scribed for ciliates. Arch. f. Protist., 77: 407. McCuiune, C. E., 1918a. Some considerations regarding microscopical technique. Anat. Rec., 14: 265. = McCiune, C. E., 19185. Possible action of the sex-determining mechanism. Proc. Nat. Acad. Sci., 4: 160. Mark, E. L., 1881. Maturation, fecundation and segmentation of Limax campestris. Mus. Comp. Zool. Harvard, 6: 173. Montcomery, T. H., 1905. The spermatogenesis of Syrbula and Lycosa, ete. Proc. Acad. Nat. Sci. Phila., 57: 162. Moore, A. R., 1935. On the significance of cytoplasmic structure in plasmodium. Jour. Cell. and Comp. Physiol., 7: 113. Pinney, M. E., 1908. Organization of the chromosomes in Phrynotettix magnus. Kan. Univ. Sci. Bull., 4: 307. ScHRADER, F., 1932. Recent hypotheses on the structure of spindles in the light of certain observations in Hemiptera. Zeitschr. f. wiss. Zool., 142: 520. Scuraber, F., 1934. On the reality of spindle fibers. Biol. Bull., 67: 519. ScHrabErR, F., 1936. The kinetochore or spindle fiber locus in Amphiuma tri- dactylus. Biol. Bull., 70: 484. WenricH, D. H., 1916. The spermatogenesis of Phrynotettix magnus, with special reference to synapsis and the individuality of the chromosomes. Bull. Museum Comp. Zodl., 60: 57. Witson, E. B., 1901. A cytological study of artificial parthenogenesis in sea- urchin eggs. Arch. f. Entw.-mech., 12: 529. SOME FRESH WATER PROTOZOA WITH BLUE CHROMATOPHORES JAMES B. LACKEY (From the U. S. Public Health Service, Stream Pollution Investigations Station, Cincinnati, Ohio) Among the Protista blue is one of the rarest of the colors. The blue-green algze may be grass-green or shades of brown or red. A culture of a marine Spirulina I maintained for two years was constantly violet. Such variations in the color of the Myxophycee are due to combinations of chlorophyll, phycocyanin, and phycoerythrin, and to the make-up of the gelatinous envelope at times. Some Protozoa have a blue or violet color diffused through the cyto- plasm as the ciliate, Stentor cereuleus Ehrenberg. Other Protozoa may occasionally have a blue color due to specific pigments contained in chromatophores. A single fresh-water rhizopod, Paulinella chromato- phora Lauterborn, fulfills this condition, but most of the blue Protozoa are Mastigophora. Various marine Dinoflagellida have blue or blue- green plastids, but according to Eddy (1930) none of the fresh-water armored species in the United States are so colored and I have been unable to find a record of such a species of Gymnodinium, the unar- mored genus. The genera Chroomonas and Cyanomonas of the Crypto- monadida have blue chromatophores but Smith (1933) does not record the occurrence of either in this country; Pascher (1913), however, holds Cryptoglena americana Davis to be identical with Cyanomonas americana Oltmanns. West and Fritsch (1927) record Chroomonas nordstetu Hansgirg as reported once from England. From such scanty records of their occurrence it would be imagined that the blue Protista are very rare. In the past several years I have collected some hundreds of water samples from various field locations. Many of these have been centri- fuged to concentrate their microorganisms and examined before great temperature or other changes could occur. In such a sample taken from Newtown Creek at Collingswood, New Jersey, in 1932, I found a blue flagellate in abundance. Since then I have found Paulinella chro- matophora, a new species of Chroomonas, and three other new flagel- lates with blue chromatophores. I have also noted a rather widespread field occurrence of some of these organisms. ‘Table I shows those which have occurred at nine locations. 492 SOME FRESH WATER PROTOZOA 7 493 From this table a comparative idea of the rareness of occurrence of these blue Protozoa may be obtained. It should be stated, however, that Cyanomonas americana has never been seen by me from but two stations, and the same is true of Cyanomastix morgam. Species of Chroomonas, however, are relatively common; thus bi-weekly samples from a small pool in South Mountain Reservation, Essex County, New Jersey, rarely failed to show some of these flagellates, while in 76 plank- ton samples from the Muscle Shoals area in Tennessee, Mississippi and Alabama, 21 contained one or more species and in a few they con- stituted the most abundant protozoan organism. This is a much more frequent occurrence than available records indicate. Their small size Protozoa with Blue Chromatophores Occurring at Nine Sampling Stations Locality where found || < By MS ver laser deers : eS ee ee Ble es ae |e Name of organism Sr GEES 82} us a | A 5 | 2 eile |e2/3 (82 |e [ez /221e Baa Seales ealeccy ca [eis iil eeu Overt gu salt ee ae | ee oO = On. oe {asia = — $8s| 6. | 86 | $5 | ges] es |S95| Ss | 32 Moe) eZ | eo | we |esa] Qe (She Of | He Paulinella chromatophora..... > x Chroomonas pulex........... x x x x x x Chroomonas nordstetw........ x x x Chroomonas cyaneus......... x x x x x Cyanomonas americana...... x Cyanomonas cereuleus....... Cyanomastix morgant........ x Gymnodinium limneticum..... x x and activity prevents ready identification, unless they are concentrated in some manner. In all these Protozoa the color is a bright blue, localized in chromato- phores from which it diffuses out as a blue liquid after death of the organism. No experimental work on its nature has been done, but it is brighter than extracted phycocyanin. Paulinella chromatophora Lauterborn. Plate I, Fig. 1. This rhizopod was found at three stations in the spring of 1936, two of them shown in Table I. Its recurrence was several times noted at one of these. All the stations were clear streams with considerable debris and an abundance of filamentous blue-green alge. Hydrogen ion concentrations were 6.2, 6.8, and 7.4. The animals conformed rather closely to the published descriptions, but are more pyriform in shape 494 JAMES, By RACKEN and with nine rows of plates instead of eleven or twelve. The cell does not fill the test and the cytoplasm is clear and homogenous. The nu- cleus is visible in the upper portion. All animals had two contractile vacuoles, definite in location as shown. The few pseudopodia are long and branch a few times. They are moved sluggishly. The two chro- matophores are curved bands, decidedly blue with no pyrenoids. It is unlikely that they are symbiotic alge as Kudo (1931) suggests, for every individual contained two; eighteen in one sample alone were examined. Cyanomastix morgani, gen. nov., spec. nov. Plate I, Figs. 2, 3. This organism occurred rather abundantly in a fresh water lime sink pond in Morgan County, Alabama, during September, 1935. It is difficult to classify, for its color places it in the Phytomastigoda, and its thick membrane, lack of a gullet and simple vacuole system should place it in the order Phytomonadida. While the peculiar shape and color of the chromatophores, and lack of a stigma are exceptional to the Chlamydomonadide, I believe it should be placed at least provision- ally in that family. No resting stages or reproductive phases were observed. Organisms practically uniform in size, 15 to 20 microns in length, 8 to 15 in width, slightly pyriform in shape. Membrane thick with five shallow grooves at round end, a small opening at top. Cytoplasm homogenous with an anterior nipple from which emerge two slightly subequal flagella, the longer about cell length. A contractile vacuole is just below this nipple. No nucleus was observed. All individuals had two chromatophores, band-shaped, twisted or turned, a vivid blue green. On them are a few large bodies resembling pyrenoids with irregular edges. The only noticeable cell inclusions are small spheres, probably oil. Locomotion is rapid, rather like that of Chlamydomonas. ‘The or- ganisms also have the trick of attaching themselves by the anterior nip- ple, with the flagella opposed. Nutrition seemed to be holophytic. Chroomonas setoniensis, spec. nov. Plate I, Figs. 4, 5. This species conforms to the generic description as given by Pascher (1913). In size it is the largest member of its genus, attaining a length of 20 microns. Its flattened oval form, deeply insunk gullet and size seem to merit calling it a new species. It has been common at two stations in the Muscle Shoals area, and in the pool in South Mountain Reservation referred to above, it appeared constantly during several months in considerable numbers. SOME FRESH WATER PROTOZOA 495 PLATE I 496 JAMES B. LACKEY Cell constant in shape, flattened, oval, slightly obliquely truncate at the anterior end, with a deeply insunk gullet or mouth depression, open along one side. About 20 microns long, 15 wide and 8 thick. Two subequal flagella emerge from the mouth depression, the longest about 25 microns long, the shorter about 20. The single chromatophore is bright blue, having the form of a peripheral cylinder, split from anterior to posterior end, with irregular edges. One simple contractile vacuole near the point of emergence of the flagella. The nucleus is median with a thin membrane and no peripheral chromatin granules. Stained preparations show the flagella to end in small basal bodies, with no visible rhizoplast. There are several types of inclusions, spheres which might be oil, and bodies resembling paramylum. It swims rapidly re- volving on its main axis. No evidence of reproduction has been seen. Cyanomonas cereulus, spec. nov. Plate I, Figs. 6, 7. This organism has been seen in only two plankton samples; one from a lake, small and partly shaded, in Morgan County, Alabama, the other from a lake in Hardin County, Tennessee, at Pickwick Dam. This water was densely shaded by cypress and tupelo trees. Cell somewhat egg-shaped, but flattened on one side. The anterior end is pointed. There is a small depression anteriorly, on the right side; two flagella emerge here. The longer is carried backward, about 20 to 25 microns long, and does not trail on the substratum at all times. The anterior flagellum is about cell length. The animal varies some- what in size, 15 to 20 microns long and about half as wide. It is some- what metabolic in that it assumes a discoid shape at times, as when changing direction of movement. The pellicle is not evident and its surface is smooth. The nucleus is small and central; there are no visible cell inclusions but the cytoplasm is finely homogeneously granular. No contractile vacuole is to be found. There are from 15 to 25 bright blue discoid chromatophores, without pyrenoids, small and peripherally located. Its nutrition and reproduction were not noted. Davis (1894) described a blue-green motile cell from the salt marshes of the Charles River in Massachusetts as Cyanomonas amer- icana. He thought it to be a motile cell of the blue-green alga Polycystis pallida. In general this organism is like his, except that the depression from which the flagella emerge is lateral instead of anterior, and there are no eyespots. A few flagellates more like his were found in a pool in Tishomingo County, Mississippi, in the summer of 1935, but could be studied only superficially. They came from a small pool which did not seem to have any blue-green alge in it, and I have found nothing to connect these blue-green or blue flagellates with such alge. On the contrary, they exhibit a high degree of organization. SOME FRESH WATER PROTOZOA 497 Gymnodimum limneticum, spec. nov. Plate I, Figs. 8, 9. This was the most common dinoflagellate at one station on Reelfoot Lake, Tennessee, in September, 1935. It was eaten in considerable numbers by the larve of malarial mosquitoes there. Cell 25 to 35 microns long, somewhat dorsiventrally flattened, but a regular oval in outline. Hypocone somewhat larger than the epicone. Transverse furrow wide, deep, with no displacement. The longitudinal furrow does not enter the epicone. ‘Trailing flagellum about one and a half times the body length. Commonly a large pusule in the hypocone, the nucleus being subcentral. The blue color is due to 8 to 12 oval dis- coid chromatophores scattered through the cell. No oil or other in- clusions were noted, and as no food bodies were found it is inferred that nutrition is holophytic. The animal is naked and the surface is smooth. No reproduction was seen. These blue flagellates present a perplexing question as to the sig- nificance of their color and its development. It appears improbable that they have any close relationship to the blue-green alge; they are too highly organized. Morphologically their relationship to amceboid and flagellate Protozoa is unmistakable. Since all appear to nourish themselves holophytically, it is perhaps simplest to regard them-as species in which chromatophores have developed containing a blue pig- ment in place of, or in addition to, chlorophyll. REFERENCES Davis, Braptey M., 1894. Notes on the life history of a blue-green motile cell. Bot. Gazg., 19: 3. Eppy, SAMUEL, 1930. The fresh-water armored or thecate dinoflagellates. Tyrans. Am. Micros. Soc., 49: 277. Kupo, R., 1931. Handbook of Protozodlogy. C. C. Thomas. Springfield, Il- | linois. Lackey, J. B., 1932. Oxygen deficiency and sewage Protozoa, with descriptions of some new species. Biol. Bull., 63: 287. PascHer, A., 1913. Die Susswasserflora Deutschlands, Osterreich, und der Schweiz. 2. Flagellate. 2. Gustav Fischer, Jena. SMITH, Gitpert M., 1933. The Fresh-water Algz of the United States. McGraw Hill Book Co. New York. West, G. S., anp F. E. Frirscu, 1927. A Treatise of the British Fresh-water Alge. Cambridge University Press, London. THE MOLECULAR WEIGHTS AND PH-STABILITY REGIONS OF THE HEMOCYANINS INGA-BRITTA ERIKSSON-QUENSEL AND THE SVEDBERG (From the Institute of Physical Chemistry, The University, Upsala, Sweden) Previous ultracentrifugal investigations carried out in this labora- tory have brought to light a number of different molecular weights and sedimentation constants among the hemocyanins (Svedberg and Chirnoaga, 1928; Svedberg and Heyroth, 1929, a, 6; Svedberg and Eriksson, 1932; Svedberg, 1933; Svedberg and Hedenius, 1933; Svedberg and Hedenius, 1934). It has also been observed that a change in the pH of the solution often causes the appearance of one or several new molecular species and the disappearance—partial or com- plete—of the species present before the change was brought about (Svedberg and Hedenius, 1934; Svedberg, 1934, a, 6, c, d). The pH-stability regions characterized in this way have often been found to be quite different even in the case of hemocyanins possessing the same molecular weight at the isoelectric point. This circumstance seems to be of considerable interest from a biological point of view and we have therefore made an attempt to study in some detail the molecu- lar weights and pH-stability regions of a number of hemocyanins in order to find out whether the diagram giving the molecular weight (or the sedimentation constant) as a function of the pH of the solution could be used to define an animal genus or species. METHOD The ultracentrifugal technique developed for the determination of molecular weights and sedimentation constants has been described elsewhere (Svedberg, 1934, a, 6, c, d). In the case of sedimentation equilibrium measurements the molecular weight is given by the formula: 2RT In Col C1 MO = Volare = a) s where M = molecular weight, R = gas constant, 7 = absolute tempera- ture, Cc. and c,; = concentrations of solute, V = partial specific volume, p = density of solvent, x2 and x; = distances to the centre of rotation, w = angular velocity. For the determination of c2/c; a procedure based upon measure- 498 M.W. AND PH-STABILITY OF HEMOCYANINS 499 ments of light absorption has previously been used almost exclusively. A new method for determining the concentration gradient utilizing the change in refractive index has been developed by O. Lamm (1933) and has in most cases been found to be superior to the absorption method in its present shape. The refraction method in its two forms (the scale-method and the slit-method) has therefore been used in this investigation. Sedimentation equilibrium measurements require the expenditure of very much time and work and such determinations can therefore hardly be used for mapping out completely the pH-stability regions. In cases where more than one molecular species is present in the solu- tion the equilibrium method has also the great drawback that it is unable to give definite information about the number and nature of these components. We have therefore used the sedimentation velocity method for collecting the data necessary for tracing the pH-stability pictures supplementing these measurements with equilibrium deter- minations. In the following diagrams the changes in the sedimenta- tion constant will accordingly indicate the dissociation reactions occur- ring at different pH values. The sedimentation constant results from a measurement of sedimentation velocity by reducing the directly observed values to standard conditions with regard to centrifugal force, buoyancy and viscosity: aL 9 (1 a Vpo) s = dx/dt 1/wx n/n G26) (2) where dx/dt = observed sedimentation velocity, w = angular velocity, x = distance from centre of rotation, » = viscosity of solvent, m) = vis- cosity of water at 20°C., V = partial specific volume of solute, p = density of solvent, pp = density of water at 20° C. In order to minimize the danger of causing artificial changes in the state of aggregation all the hemocyanins have been studied as they occur in the blood without any treatment other than dilution with NaCl or buffer solutions. In the case of the crustaceans the fibrin has been removed by shaking the blood with glass beads. As a rule the light absorption method has been used in the determination of the sedimentation constants (A = 366 my, nickel oxid glass filter or the region } = 250-290 mu, chlorine and bromine filter) but some check runs with the refraction method have been carried out. With regard to the details of procedure the previous communication (Svedberg and Hedenius, 1934) should be consulted. Recent work in this laboratory has shown that some of the earlier sedimentation equilibrium measurements have given too low molecu- lar weight values (probable causes of error: systematic deviations in 500 I-B. ERIKSSON-QUENSEL AND THE SVEDBERG the evaluation of the photographic blackenings, incomplete attainment of equilibrium). The discrepancy believed to exist between the fric- tional constants deduced from sedimentation velocity and sedimenta- tion equilibrium measurements on the one hand and diffusion measure- ments on the other (Tiselius and Gross, 1934; O. Lamm and A. Polson, 1936) therefore vanishes and the possibility of computing molecular weights by combining sedimentation velocity and diffusion measure- ments as pointed out long ago by one of us (Svedberg, 1925) and used for such computations at the earlier stages of our investigations (e.g. Svedberg, 1927) is accordingly actualized, The molecular weight is given by: RTs AE Seas) Me) where D = diffusion constant and the other symbols have their pre- vious meaning. A systematic study of the diffusion constants of the proteins under- taken by A. Polson has furnished a number of data also for the hemo- cyanin. His diffusion constants combined with our sedimentation constants according to equation (3) give molecular weight values which agree very well with those calculated from our sedimentation equilib- rium measurements (see Table XXVI). SEDIMENTATION CONSTANTS! The main part of the work consisted in the carrying out of sedi- mentation velocity measurements at different pH-values. From these data the pH-stability curves were traced. In some cases the reversibility of the dissociation reactions have been tested. Arthropoda Hemocyanin is found in the blood of the crustaceans and the Arachnomorpha, but not in that of the myriapods and the insects. We have only studied representatives of the subclasses Malacostraca and Xiphosura, the species belonging to the other subclasses yielding too little blood for an extended investigation. Crustacea: Malacostraca The respiratory proteins contained in the blood of the following animals belonging to the order Malacostraca of the class Crustacea 1The sedimentation constants are expressed in units of 10°. Centrifugal force is expressed in terms of the gravitational constant. The letters A-K are used to characterize the different hemocyanin components. Components having the same sedimentation constant are designated by the same letter. A + in the column of a table indicates that the component in question has been observed but that it was impossible to calculate its sedimentation constant. M.W. AND PH-STABILITY OF HEMOCYANINS 501 have been investigated: Pandalus borealis, Palinurus vulgaris, Nephrops norvegicus, Homarus vulgaris, Astacus fluviatilis, Cancer pagurus, Carcinus menas. The two sedimentation constants 16 and 23 are characteristic of the group. In Pandalus and Palinurus the sedimenta- tion constants for the main components are 16.4 and 17.4. In all the others the “‘normal’’ sedimentation constant is about 23 and dissocia- TABLE I Pandalus borealis Dilution of blood, 15 times; centrifugal force, 200,000 (speed 52,000 r.p.m.); thickness of column of solution, 0.6 cm.; source of light, mercury arc; light filter, chlorine and bromine; plates, Imperial Process; exposure time, 30 seconds; aperture of lens, F : 36; developer, metol-hydroquinone, 1 minute. | PH of! Total | se 0 S20 Solvent sol- molar | G H vent FUNG RNa Ac wNa Chae sent) 2. 3.6 | 0.22 16.1) Containing inhomogeneous components with Soo = 12.0 and 4.7 ef ie Se Mee eae ee 4.0 | 0.22 18.4] Containing inhomogeneous components with soo = 4.7 i " Re i ee Se Stee 4.2 | 0.22 | 22.3] 16.9 oe a PER gre ated oe 4.6 | 0.22 | 22.4 Inhomogeneous Ms e Bible ir aati crys 5.0 | 0.22 | 21.9 Inhomogeneous a ‘ He RESIN NS eee NON S32) | O22 17.6| Slightly inhomogeneous ic ie cies career emai fy 5.) || O22 17.4 KH2PO,, NasH POs, NaCl 6.0 | 0.22 18.4 ‘ s A 6.8 | 0.22 17.4 ii s rf 6.8 | 0.22 0 “ a a0 6.8 | 0.22 ted iS Pe 7.4 | 0.22 16.4 ee = Tp cic-all 0) |) OLA 16.9 KH2PO:, NazB.O7, NaCl..... 8.5 | 0.27 17.0 sf SS Prise es 9.0 | 0.27 17.9 NazCOs, NaeB.Oz, NaCl. Bee thee ieee et) 0.25 77 = RS ea Aree eae 9.5 | 0.25 17.4 ¢ “ Paiva a ae 10.0 | 0.25 17.0 fy e Gay oes 10.5 | 0.25 16.6] 7 times diluted; NiO-filter ie rs Nae, eet 10.8 18.1) Inhomogeneous lower part 22.9] 17.4 tion products with sedimentation constants around 16 are formed. Pandalus shows an aggregation product with the sedimentation con- stant 22.9. All the final alkaline-splitting products have sedimenta- tion constants approximating 5. 502 L-B. ERIKSSON-QUENSEL AND THE SVEDBERG 0 0:2 0:4 0:6 08 1:0 1:2 Distance from meniscus, cm. Fig. 2a. Hires 2a Fic. 2. Sedimentation pictures (a) with photometric records (6) for hemocyanin from Palinurus vulgaris at pH 9.0 (Soo = 16.4); centrifugal force 120,000; time between exposures 5 minutes. 0 0:2 0-4 0:6 0:8 1:0 42 Distance from meniscus, cm. Fic. 3a. Fic. 30. Fic. 3. Sedimentation pictures (a) with photometric records (0) for hemocyanin from Palinurus vulgaris at pH 10 (Sx = 16.4 and 4.10); centrifugal force 280,000; time between exposures 5 minutes. Pandalus borealis Kristineberg, Sweden The main component of this hemocyanin, component H, has the sedimentation constant 17.4. It is present in the blood from pH 3.6-10.8 with the exception of the pH region 4.2-5.0. It is homo- geneous with regard to molecular weight from pH 5.5-10.5. At 10.8, 4.0, and 3.6 a large part of H dissociates into smaller molecules. In the region 4.2—-5.0 the blood contains a higher very inhomogeneous M.W. AND PH-STABILITY OF HEMOCYANINS 503 component, G, the sedimentation constant of which is 22.9. Figure 1 (Plate I) gives the pH-stability diagram. f Palinurus vulgaris Roscoff, France In the pH-region 3.6—9.4 one molecular species is found in the hemocyanin (component H). This has the sedimentation constant 16.4 and is perfectly homogeneous. At pH 9.4 it dissociates to some TABLE II Palinurus vulgaris Dilution of blood 5 times; centrifugal force, 150,000 (speed 45,000 r.p.m.); in the last two runs 250,000 (speed 59,000 r.p.m.); thickness of column of solution, 2.2-1.0 cm.; light filter, nickel oxid glass; time of exposure 10-15 seconds; other con- ditions as in Table I. pH of | Total | sso S20 Solvent solvent | molar | H K Nav Ne aideNG NaCI seo eis ay tint Mees ex SrOn a LOPZOM ses (a9 (a3 (77 Batis Genes eC AOE 4.0 | 0.20 | 17.4 . ie Gg SON areas cree eee 4.6 | 0.20 | 16.6 Naa nROgMI>... 3.4 23.3 IBUAGS INEVAG, IN opogusoeses 4.0 | 0.12 | 24.5 ; Purified hemocyanin (Am:SO.-prec.) UAC NANG eNaGle ne oe ae 4.0 | 0.20 | 22.8 HAc, NaAc, KCl........... 5.0 | 0.12 | 22.6 ISTA\G, INVA! INENCIE Se ecipo ons 5.0 | 0.20 | 23.4 KH2PO,, NasHPO,, NaCl. ...| 6.0 | 0.20 DIRS oe Me “,...| 6.8 | 0.20 | 23.6 KH2POs, NasHPO,, KCl Ga0m0 6.8 0.12 20.7 Oe aR OWe ce 8.0 | 0.12 | 22.6 KH2PO,, NasB.O7, NaCl peoone 9.0 0.20 21.9 NaOH, Na2HPO,, KCl...... 9.8 | 0.15 | 23.4 | 16.9 NasB.O7, Na2COs, NaCl Sirus 10.0 0.25 } 17.3 tf i aap te 10.1 | 0.23 14.8 . i Teh oe 10.0 | 0.20 | 22.2 NaOH, NazHPO:, NaCl..... 10.2 | 0.05 | 19.3 | 15.5 NazBsO7, NasCO;3, NaCl..... 10.3 | 0.20 15.5 Tea naan 10.3 | 0.20 21.8 st oe PANN oie 10.6 | 0.17 16.0 | + NaOH, NazHPO:, NaCl..... 10.7 | 0.06 15.6 | 6.64 NaeB.O7, Na2COs, NaCl emaihareetre 10.5 0.20 21.5* wh cept u es 10.3 | 0.20 LO re : fee ahenetars 10.7 | 0.20 | 19.8* | 16.8 NazCO3, Naz2BsO7, NaCl..... 10.8 | 0.20 16.6 NaOH, Na2HPQ:, NaCl..... 11.0 | 0.06 13.8*| + 2 os Se ee ee oie 11.0 | 0.32 12.2* | 9.72| Inhomogeneous * oF Ree aetied ne 11.4 | 0.07 5.81] Inhomogeneous 22.6 | 16.1 * Not used for calculation of mean value. Homarus vulgaris Havstensund, Sweden The sedimentation constant of the main component G is 22.8. G is present alone in the blood from pH 3.4 to 9.8. At pH 9.8 and 506 I-B. ERIKSSON-QUENSEL AND THE SVEDBERG above a dissociation product, H, is formed with the sedimentation constant 16.1. When both G and H are present together it is difficult to determine their separate sedimentation constants. Generally G is observed in solutions of pH up to around 10.5 and H upto 11. At pH 11.4 the sedimentation picture shows only low-molecular inhomo- geneous material. Figure 6 (Plate I) gives the pH-stability diagram. Astacus fluviatilis Upland, Sweden The hemocyanin shows more than one component in the whole pH-range investigated. Inthe pH region 3.6-10, 80-85 per cent of the TABLE V Astacus fluviatilis Dilution of blood 10 times; centrifugal force 150,000 (speed 45,000 r.p.m.); for the last three runs 250,000 (speed 59,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. Solvent pile poet By Se S20 lBLAVeS INE Vac NECA S 56 erie Secs « S200 O20 M2322) alieo Inhomogeneous KELPOR NEB OMENACl int “ r “ 9.0 | 0.25 | 23.7| 16.1 Na2CO;, NaeBiO7, NaCl........ 9.4 | 0.25 | 24.0) 17.5 se ae Ma atest LOS) O23" 2A eo lno hs i ty vole enh ud eee ee LOG) (O21 2229 aS Ome NaOH NasliPO,, NaCl 2) .5.- 10.9 | 0.27 WES? || Sail a oe Lec eee eae es 2 OT, 4.65] Inhomogeneous a i A iat aecats ria Se O32 4.19 be 23.3] 16.3 | 4.02 * Not used for calculation of mean value. M.W. AND PH-STABILITY OF HEMOCYANINS 507 40 3 5 7 9 1! 13 3 5 7 9 il 13 40 20 ! ! | 3 5 7 9 11 3 8 40 20 3 5 7 9 M 5 11 PLATE I pH-stability diagrams for hemocyanins of species listed below. Abscissz in all figures, pH; ordinates, soo. The dotted lines in Figs. 5, 6, 8, 10 and 13 indicate the positions of the isoelectric points. Fic. 1. Pandalus borealis. Fic. 4. Palinurus vulgaris. Fic. 5. Nephrops norvegicus. Fic. 6. Homarus vulgaris. Fic. 8. Astacus fluviatilis. Fig. 10. Cancer pagurus. Fic. 11. Carcinus menas. Fic. 13. Limulus polyphemus. 508 I-B. ERIKSSON-QUENSEL AND THE SVEDBERG protein has the sedimentation constant 23.3 (component G). The other component present, H, has the sedimentation constant 16.3. Around pH 10 the amount of H starts to increase, G to decrease. At Distance from meniscus, cm. Fic. 7a. Fic. 70. Fic. 7. Sedimentation pictures (a) with photometric records (6) for hemocyanin from Astacus fluviatilis at pH 7.7 (soo = 23.3 and 16.3); centrifugal force 150,000; time between exposures 5 minutes. 0 0:2 0:4 0:6 0-8 1-0 4-2 Distance from meniscus, cm. Fic. 9a. Fic. 98. Fic. 9. Sedimentation pictures (a) with photometric records (6) for the hemo- cyanin from Cancer pagurus at pH 4.2 (Soo = 32.7, 23.6, and 16.4); centrifugal force 120,000; time between exposures 5 minutes. M.W. AND PH-STABILITY OF HEMOCYANINS 509 | pH 10.3 a low molecular component, K, is formed. In a narrow pH- range, 10.3-10.6 G, H, and K appear together. At pH 10.9 G has totally disappeared and there is very little left of H. In more alkaline buffers all of the material is low-molecular and inhomogeneous. Figure 7 gives an example of a sedimentation run and Fig. 8 (Plate I) the pH-stability diagram. Cancer pagurus Havstensund, Sweden The main component, G, of this blood has the sedimentation con- stant 23.6 and is observed in the wide pH-range 3.6-11.0. G is always accompanied by another component H with the sedimentation constant 16.4. H is present only in very small quantity. In the pH range 3.6—4.8 appears a third component, F, aggregation product of G, TABLE VI Cancer pagurus Dilution of blood 6 times; centrifugal field 115,000 (speed 40,000 r.p.m.); in the last run 220,000 (55,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. Salcent "sol nota S20 S20 | Seo | Seo HCl Na-citrate, NaCl....| 3.0) 0.20 Aggregated and inhomogeneous FAG Nance, NaCiacn a. - 3.6 | 0.20 | 32.7 | 24.7) 15.3 ce ‘ feat ics nt acre 4.0} 0.20 | 32.5 | 24.9) + ot i pts WA ke cere 4.2 | 0.20 | 34.6 | 25.0 x Una Se 4.5 | 0.20 | 31.0 | 24.2] 16.0 o SNe sheers 4.8 | 0.20 | 29.9* | 24.1) + a fe SCR SAL ni 5.0 | 0.20 23.8} 16.5 ‘3 me ee oS SAD 5.4 | 0.20 23.8) 16.1 KH2PO., NasHPO,., NaCl.| 6.0 | 0.20 23.7| 16.6 * iy 25} OS |) O20 23.2| + ee a = Al Tos || O20 22.5} + KH2PO., Na2B.O7, NaCl..} 8.0 | 0.25 22.7) + + ‘a 5 Bl ies) | OL5) 23.1) + NazCOs, Na2B.07, NaCl. 5 9.0 0.25 2D) Al + “ ry mea Os om Os2o Poet Mies a iS % 10.0 | 0.25 24.3) 17.7 4 i“ aS 10.5 | 0.25 DTN Neo - CS 5 OSV OZS 23:0) =- NazHPO., NaOH, NaCl. .|11.0 | 0.30 23.9} 16.2] 4.60) Inhomogeneous - - ieee le) 0:30, 4,82 ui 32.7 | 23.6) 16.4) 4.71 * Not used for calculation of mean value. with the sedimentation constant 32.7. At pH 11.0 component K with sedimentation constant 4.77 is observed together with Gand H. At 510 I1-B. ERIKSSON-QUENSEL AND THE SVEDBERG pH 11.5 G and H have disappeared. K is not quite homogeneous. At pH 3.0 the hemocyanin is aggregated and non-uniform. Figure 9 gives an example of a sedimentation run and Fig. 10 (Plate I) the pH- stability diagram. TaBleE VII Carcinus menas Dilution of blood 6 times; centrifugal force 175,000 (speed 50,000 r.p.m.); in the last two runs 280,000 (speed 63,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. Sven eee IBUAVE SIN EVA INBKEN 4 gh G Ggat oo a ue 30) 71) 0206 2320|hti at 4 os Pao beser neha ts 4.0 | 0.20 | 24.0) 17.1 3 iG Ap Men PRN crc 4.6 | 0.20 | 23.9) 16.8 iy : BE Dh aarti AION ae 540) |) Dass) | ZAI War KH.2PO,u, NasHPO:, NaCl....... LON | 07S5ii 23:0 Ges z fe KCl. eee O80 OM R235 elo:2 os os Nia Clear cose 74 | 0.20 | 22.6) + a Clee ee 8.0 | 0.12 | 23.4] 16.6 te NaeB.O7, NaCl....:... S55) 0225) 2S eilello:3 - i OO ae ee SOM O25 e230 ies ts os ay PLS aE OEY of OS) || AAs vill Malpas) se . CAI are 10:0 | O25 | 22.7) 16:7 si os isthe Niche 10°59) 0:25) 22-5) 1A SF NaOH a, Sica Mee cee a i) Oza 5.00} Inhomogeneous ¢ x BENE ep a 11.5 -| 0.30 4.17| Inhomogeneous 2Se Si lO * Not used for calculation of mean value. Carcinus menas Kristineberg, Sweden The sedimentation constant of the main component G is 23.3. It is stable from pH 3.6-10.5. A smaller component with the sedimenta- tion constant 16.7 is present throughout the same range but to a very small amount. At pH 11.2 both are dissociated into component K, a low-molecular inhomogeneous split-product. Figure 11 (Plate I) gives the pH-stability diagram. Arachnomorpha: Xtphosura: Limulus polyphemus Woods Hole, Mass., U.S. A. In the blood four different components are present at the same time M.W. AND PH-STABILITY OF HEMOCYANINS Sl in the pH-range 5.2—-10.5. Their sedimentation constants are 56.6 (D), 34.6 (F), 16.1 (H), and 5.87 (K). The proportions between their concentrations are unchanged inside this region. Above 10.5 D, F, and H disappear; they split up into K. Below pH 5.2 D and F dis- TABLE VIII Limulus polyphemus Dilution of blood 4 times; centrifugal force 125,000 (speed 41,000 r.p.m.); in the last three runs 300,000 (speed 65,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. Solvent ‘ous EE) S| ee eae HAc, NaAc, NaCl...... 3.6 | 0.36 24.2) 16.2) 6.02 ss ae RE I 4.0 | 0.19 23.5| 16.1] 6.15 i Es Sey oo ty. 4.2 | 0.36 23.5] 16.4) 5.24 5 st Fee eSNG 4.4 | 0.19 23.3} 16.2) + i % ot pa hehehe 4.6 | 0.19 | 55.2* |. 23.5) 15.8} + % a ri ata Ba 4.6 | 0.19 24.5} 16.5} + S ny Ce te See a 4.8 | 0.19 35.2) 23.1) 15.8) + - ‘ Sih eee aces 4.8 | 0.19 24.7| 17.0) + i a Re ai Seca 5.0 | 0.19 25.8] 16.8] 6.3 « is i anoathafotees 5.2|0.19| 56.3 | 34.8 16.1) + KH2PO:, Na2HPO,u, NaCl| 5.4 | 0.24] 58.2 | 35.3 17.2) + % ‘a ot 5.6 | 0.19 | 59.7 | 38.1 16.2) + “ bi af 6.3 | 0.36 | 56.5 | 36.7 15.5) + ue ot ae 6.3|0.36) + + 15.5) 5.0F Y ce 6.8 | 0.19 | 57.3 | 36.4 16.3) + ee a i 7.2 | 0.36) 57.1 | 35.1 15.5] + oy 4 7.6 | 0.19 | 56.4 | 36.0 15.9) + a i af 8.0 | 0.36] 58.0 | 36.0 NOW) Se NasHPO,s, NaCl........ 8.9 | 0.20) 58.2 | 33.8 16.1] 6.28 “ NaOH, NaCl .| 9.8 | 0.22} 56.1 | 35.1 56} Se ai a MNO | O22) S20 1) SOM 17.6} 6.25 Na2CO3, Na2B.O7, NaCl .|10.3 | 0.22 | 53.4 | 29.6 14.4] 5.70 oS i “10.5 | 0.23 | 54.6 | 32.3 16.1] 5.86 NazH POs, NaOH, NaCl .|10.7 | 0.23 5.14 Na2COs, NasB.O7, NaCl é 10.8 0.23 i + 4.88* NazHPO., NaOH, NaCl ./11.5 | 0.27 4.45* | Dialysed blood NazHPO:, NaOH, NaCl .|12.1 | 0.27 4,26* | Dialysed blood 56.5 | 34.6) 24.0) 16.1) 5.87 * Not used for the calculation of mean value. + Made at 60,000 r.p.m. and with blood diluted 2 times to get component K. sociate in two ways. The concentration of K apparently increases; 1.e€., some part of them must form more of K. A new component, G, with the sedimentation constant 24.0 appears in a rather high concen- tration; and G must be a dissociation product of D and F as the con- centration of H does not change and the concentration of K increases. Sa L-B. ERIKSSON-QUENSEL AND THE SVEDBERG Figure 12 gives an example of a sedimentation run and Fig. 13 (Plate I) the pH-stability diagram. Mollusca The bloods of the Amphineura, the Gastropoda and the Cephalo- poda contain hemocyanin. Because of the difficulty of obtaining sufficient material from specimens of the first class our investigation has been limited to species belonging to the two latter classes. Conchifera: Gastropoda: Blood from the following animals belonging to this group has been investigated: Littorina littorea, Neptunea antiqua, Buccinum undatum, 0 0:2 04 06 0:8 1-0 1:2 Distance from meniscus, cm. Fic. 12a. Fic. 120. Fic. 12. Sedimentation pictures (a) with photometric records (0) for the hemo- cyanin from Limulus polyphemus at pH 6.8 (seo = 56.6, 34.6, 16.1, and 5.87); centrif- ugal force 120,000; time between exposures 5 minutes. Busycon canaliculatum, Helix pomatia, Helix arbustorum, Helix nemora- lis, Helix hortensis, Limax maximus. They show extremely varied stability diagrams and a great number of dissociation products. The main component has a sedimentation constant of about 100. In blood from two of the animals, Busycon and Limax, a higher component with the sedimentation constant near 130 is found. All, except Neptunea, show a dissociation product with a sedimentation constant of about 60. Very often a component with sedimentation constant of approximately 16 is found and the final alkaline dissociation products have sedimentation constants around 11. It is also striking that the four kinds of Helix have different stability diagrams, so that each of them could be characterized thereby. M.W. AND PH-STABILITY OF HEMOCYANINS 513 Littorina littorea Kristineberg, Sweden Between pH 5 and 8.5 this hemocyanin shows one homogeneous component B, with the sedimentation constant 99.7. At pH 8.5 B is partly split up into components C and H. In the pH-range 8.5-10.7 most of the hemocyanin is present in form of C, with the sedimentation constant 63.3. H appears only in very small amount up to pH 10.8 TABLE IX Littorina litiorea Dilution of blood 4 times; centrifugal force 49,000 (speed 25,000 r.p.m.); in the last three runs 180,000 (speed 50,000 r.p.m.); thickness of column of solution, 1.2 cm.; other conditions as in Table II. H of Solvent eee oe be NaAc, HAc, NaCl...... 4.0 | 0.22 Solubility too low for determination ie sf Pasko crake 5.0} 0.22 | 96.9 Se ee CER UIC 5.8) O22 | Coy NasHPO,, KH:PO,, NaCl| 5.9| 0.22 | 99.8 “ ts «| 6.2| 0.22 | 98.4 us ts S| O22 | OR5 “ “ A O20 | ORO “ ve © | 810,| 0.22 | 94.6 Na2B.O;, i i Seon OL2 a LOZ O3z5, == te i ° 9.0 | 0.27 | 102.1 | 64.8 = a Na:CO;, “‘ 9.5 | 0.25 | 101.4 | 63.0 =. o . SOO | O25) | 00.0 | C24 a 3 - SLO Sa O:25 01.979) 6225 = . “110.5 | 0.25 | 104.6 | 64.3 =e = ie Se LOLOnOLZ >a 10220) 6223 + i‘ Po eee LOMO. 2am elO229R tO See: = “t ‘ (O83! O28. | Ores) So.5 |) as NaeHPO., NaOH, ilileal Pp @28 15.6 es a we falas) |p eZ) 1g i s ala. S) IP O25 229% 99.7 | 63.3 | 15.0 * Not used for the calculation of mean value. where most of B and C are dissociated into H and above it is the only component present. When first observed H has a sedimentation constant of about 15, but it decreases towards the alkaline side and the protein becomes inhomogeneous. At pH 4.0 precipitation takes place. Figure 14 (Plate II) gives the pH-stability diagram. 514 I1-B. ERIKSSON-QUENSEL AND THE SVEDBERG 0 0:2 0-4 0-6 0:8 1-0 1:2 Distance from meniscus, cm. Fic. 15a. Fic. 150. Fic. 15. Sedimentation pictures (a) with photometric records (0) for the hemocyanin from Neptunea antiqua at pH 8.0 (sx = 104.1); centrifugal force 40,000; time between exposures 5 minutes. 0 0:2 04 0:6 0:3 1-0 1:2 Distance from meniscus, cm. Fic. 16a. Fic. 160. Fic. 16. Sedimentation pictures (a) and photometric records (6) for hemocyanin from Neptunea antiqua at pH 8.6 (se = 104.1 and 14.3); centrifugal force 60,000; time between exposures 5 minutes. Neptunea antiqua Kristineberg, Sweden In the pH-range 4.6—8.6 the hemocyanin is homogeneous and shows one component, B, the sedimentation constant of which is 104.1. At 8.6 there begins a partial splitting into component H with the sedi- mentation constant 14.3. In the pH-region 8.6-10.0 B and H are M.W. AND PH-STABILITY OF HEMOCYANINS Sul present together. At pH 10.8 the splitting is total, B disappears and a new component, I, with a sedimentation constant near 13 appears. The sedimentation constant of I decreases at a higher pH and the protein becomes inhomogeneous. In acid solutions aggregation takes place. The material becomes more and more inhomogeneous, the particles larger. At pH 3.0 most of the protein is precipitated. Figures 15 and 16 give examples of sedimentation runs and Fig. 17 (Plate II) the pH-stability diagram. TABLE X Neptunea antiqua Dilution of blood 15-20 times; centrifugal force 49,000 (speed 25,000 r.p.m.); in the last three runs 180,000 (speed 50,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. pH of | Total $20 $20 $20 B H I Solvent solvent] molar NaAc, HAc, NaCl........ 3.0 | 0.19 Most of the material precipitated 116.8* Inhomogeneous 138.2* e 110.6* es 102.4 104.7 102.7 101.1 106.7 102.8 NasHPO,, KH»PO,, NaCl .| 5.5 “ re ent ae “ OCI Tl on (=) Sessssssse PRR RPP PPE Noe Sooo ol ooo) i “| 8.0 | 0.19 | 106.9 NaeB.07, i | 2 8:6)1-0:20)| 100.7 | 14.0 y eon Sa O2Z0N LOL OF e126 re Fe ees | On Ol2 On elO 1 Ae int 326 NasCO;, ~~ =| -9:25)/0:20| 107.8 | =- a St GS OADM ORS) ieee i s oh MOO | OOP OO te 2 a ee nar Se O22 0 15.7) 13.8 | Inhomogeneous NAOH NGIPOM) mene) 101 10.97 9.48 “ 104.0 | 14.3 * Not used for calculation of mean value. Buccinum undatum Kristineberg, Sweden The main component, B, of this hemocyanin is observed in the pH-range 5—10.8. Its sedimentation constant is 102.1. B is present alone from pH 5 to 9.5. Here a very small quantity of C, a component 516 L.-B. ERIKSSON-QUENSEL AND THE SVEDBERG with sedimentation constant 63.8 is formed. The relative amount of C increases with pH. Above 10.8 B and C are dissociated into more inhomogeneous material with a sedimentation constant around 11. At pH 4.0 the hemocyanin is precipitated. Figure 18 (Plate II) gives the pH-stability curve. TABLE XI Buccinum undatum Dilution of blood 7 times; centrifugal force 49,000 (speed 25,000 r.p.m.); in the last four runs 220,000 (speed 55,000 r.p.m.); other conditions as in Table II. H of Solvent eon eee ope ot NaCl.. 0.17 | 104.5 4.0 Protein precipitated HAc, NaAc, ee Se ONOeZ OM O20 IXIBIEOn, INaDRNEOn, ocl| 9 PUD | Sheet oe O24) O20 7] uO “ iB IOs ¢ | CE | CMA Oza a ‘t NaCl..| 7.4 | 0.20 | 100.7 rk Neely, ICI. .|) 801) O12 | Ids . INaGlEA | Sron | POl25) | 103-2 : ce a CLOW OS) Ie CES Na2COs, i ssi) D5) 225 Oss oa x 5 (NO0:| O25 | L050 | GAec es s ~ [MO |] O.28 ) W005 | OSL) $F a ts “(10.8 | 0.25 12.8 NaOH, NasHPO,s, ‘ ../11.3} 0.30 10.5 | Very inhomogeneous Me oe Cpe LS | O80 11.0 ey He ih oe es a (LAI | ©.935 9.2 ay i 102.1 | 63.8 Busycon canaliculatum Woods Hole, Mass., U.S.A. The hemocyanin from this animal shows a very complicated stability diagram. Four well-defined components, A, B, C, and H are observed within the investigated pH-range 3-12. The sedimenta- tion constants are: for A 130.4, for B 101.7, for C 61.1 and for H 13.5. A and B are present together in the region 4.5-7.7._ The relative amount of A is larger towards the acid side. It is not perfectly homogeneous; the sedimentation curves have a shape indicating the presence of slightly faster sedimenting particles. At 4.0 the protein aggregates, and becomes inhomogeneous. In the region 7.7—9.0 there M.W. AND PH-STABILITY OF HEMOCYANINS 51 ~I are three components present, B, C and H. The amount of B is larger in the less alkaline part of the region, at 9.0 it is only several per cent of the whole protein. The percentages of C and H increase with pH. In the pH-region 9.0—10.8 the components C and H are Distance from meniscus, cm. Fie. 19a. Fic. 19d. Fic. 19. Sedimentation pictures (a2) and photometric records (6) for hemo- cyanin from Busycon canaliculatum at pH 6.5 (seo = 130.4 and 101.7); centrifugal force 40,000; time between exposures 5 minutes. 0 0-2 0:4 0-6 0:3 1-0 12 Distance from meniscus, cm. _ Fig. 20a. Fic. 200. Fic. 20. Sedimentation pictures (a) and photometric records (6) for hemocyanin from Busycon canaliculatum at pH 8.5 (seo = 101.7, 61.1, and 13.5); centrifugal force 70,000; time between exposures 5 minutes. more inhomogeneous. Figures 19 and 20 give examples of sedimenta- tion runs and Fig. 21 (Plate II) the pH-stability curve. 120 100 80 20 140 120 100 80 60 40 20 18 PLATE II pH-stability diagrams for hemocyanins of species listed below. Abscissz in all figures, pH; ordinates, sx. The dotted lines indicate the positions of the isoelectric points. Fic. 14. Littorina luttorea. Fic. 17. Neptunea antiqua. Fic. 18. Buccinum undatum. Fic. 21. Busycon canaliculatum. M.W. AND PH-STABILITY OF HEMOCYANINS 519 Helix pomatia Upsala, Sweden The hemocyanin appears in four different well-defined molecular species, B with the sedimentation constant 98.9, C with 62.0, H with TABLE XII Busycon canaliculatum Dilution of blood 20 times, centrifugal force in the pH-range 3-8, 35,000 (speed 23,000 r.p.m.), in the pH-range 8-11, 60,000 (speed 30,000 r.p.m.), above pH 11, 200,000 (speed 55,000 r.p.m.); thickness of column of solution, 0.6 cm., other condi- tions as in Table II. soo) [gee ee penis HCl, Na-citrate, NaCl...) 3.0 | 0.30 45.5 Inhomogeneous HAc, NaAc, NaCl....... 3.4-3.6| 0.19 Solubility too 2 . SORA ee 4.0 | 0.19 106.0 low for deter- a i Se ANOS Piosaiaeg 4.5 | 0.19 | 133.0} 103.8 mination “ e pea coreg 3 te 5.0 | 0.19] 137.0) 107.2 ie ce Bok ash aera 5.0 | 0.19 | 134.9) 104.6 ‘s i - 6.0 | 0.19} 130.1} 102.8 i i < 6.5 | 0.19} 128.1) 102.1 fi i. KCl. 6.8 | 0.12 | 128.5} 102.0 oh c exCl Wot Os) Wa rieed enter ne a 7.4 | 0.19 | 127.3] 98.3 “ a oe 7.7 | 0.19 98.6] 64.1) + is i KCl . 8.0 | 0.12 102.9) 65.5} + i NaeB.O7, NaCl 8.3 | 0.23 98.9] 62.7) + a : ss 8.4 | 0.23 96.1; 59.6] 12.2 u - oi 8.4 | 0.23 102.2) 60.3} 14.1 % = ae Se OLS 97.1| 60.0) 14.2 i - es 8.6 | 0.23 107.4| 64.0} 14.9 NazCOs, a ef 9.0 | 0.23 103.3| 61.7) 14.7 “4 wa * 915) 0:23 60.9) 13.0 és os 10.0 | 0.23 60.5] 13.0 ia “ i 10.3 | 0.23 58.0} 14.0 i - 10.5 | 0.23 58.2] 12.6 i. os e 10.8 | 0.23 58.5] 13.6 NazPHO,, NaOH, i lied || O27 12.0 rh mi 11.5 | 0.30 11.5* | Inhomogeneous zs tf a 12.1 | 0.30 8.52* " 130.4) 101.7] 61.1} 13.5 * Not used for calculation of mean value. 16.0, and I with 12.1. The main component is B, present in solutions with a pH from 3.6 to 8.2. It is alone in the pH-region 4.6—7.4 and accompanied by C in the regions 3.6-4.6 and 7.4-8.2. H is present in very small quantity together with B and C in the pH-range 7.9-8.2, 520 I-B. ERIKSSON-QUENSEL AND THE SVEDBERG TABLE XIII Helix pomatia Dilution of blood 15 times; centrifugal force in the range pH 3-8, 50,000 (speed 27,000 r.p.m.); above pH 8 150,000 (speed 45,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table IT. Solvent eae ee aes aa HCl, Na-citrate, NaCl. .| 3.3 | 0.27 . 22.5 Inhomogeneous HAG Nac; NaCl: o..-. 3.6 | 0.20] 98.2 | 61.9 4 rs pon Ate BeONl073/1 60.5 26.5 ne 5 cance 3.8 | 0.37 | 102.5 | 63.1 ui ~ coe a eae 3.8|0.20| 95.6 | 59.2 “ ut Se OnO2 0 98:3 “ 66 “| 74 | 0.20) 94.3 | =- “ “ “| 7.5|0.20| 89.6*} + Bad separation of the 2 components Ho NasB.O7, “ | 8.1 | 0.23) 94.1 | 65.8 es “ os 8.2 |0.23| 99.3 | 62.8} 19.5* ve os 4 8.3 | 0.23 16.5 ry ue se 8.4 | 0.24 15.3 us i ” 8.5 | 0.24 16.5 2 Me ud 8.6 | 0.24 16.6 ay “ ‘ 8.9 | 0.24 ard a ae we 9.3 | 0.24 15.7 + NazCOs;, as WS 9.4 | 0.22 15.8 | 13.4 we i oe 9.5 | 0.23 15.7 | 11.9 Be ee Hs 9.7 | 0.23 15.4 + re 4g He 9.8 | 0.22 14.8 os te es 9.8 | 0.23 14.7 | 11.9 ue oe He 9.8 | 0.23 15.6 | 11.1 a 1 "110.0 | 0.23 18.1 + Ma sh ““ 110.3 | 0.23 15.0 | 11.5 i“ a “110.5 | 0.23 Soi |) Lilsa se i: 110.6 | 0.42 18.1 | 13.9 ss He “110.8 | 0.25 + 11.3 NaOH, NasHPOs,, “111.5 | 0.30 9.9* |Inhomogeneous “6 be cc 12.1 0.30 8.4% 66 98.9 | 62.0) 16.0 | 12.1 * Not used for the calculation of mean value. M.W. AND PH-STABILITY OF HEMOCYANINS Sz alone between pH 8.2 and 9.3, and together with I in the pH-range 9.3-10.8. In more alkaline solutions the hemocyanin is inhomogeneous and the sedimentation constant decreases. Below pH 3.6 the material is non-uniform and the sedimentation constants are irregular but apparently decreasing. Figure 22 (Plate III) gives the pH-stability diagram. TABLE XIV Helix arbustorum Dilution of blood 5 to 30 times; centrifugal force in the pH-region 3-8, 50,000 (speed 26,000 r.p.m.); above pH 8 175,000 (speed 50,000 r.p.m.); thickness of column of solution 0.4—2.2 cm.; other conditions as in Table II. pH of | Total | soo S20 $20 S20 Solvent solvent] molar | B Cc H I HCl, Na-citrate, NaCl....| 2.5 | 0.20 Very inhomogeneous a ss Sch ral enon msn () 2019 S225 | si HAc, NaAc, NaCl....:.. 3.6 | 0.20 | 83.8} 56.0 “ “ « | 30 | 0.20 | 87.2) + ie NaeB.O7, NaCl..| 8.2 | 0.25 16.1 i " Teal) een |) O28) 16.1 i Ne a Sad |) O28) N50) ae Na2COs, 4 a 9.0 16.4! 13.0 if a a 9.4 14.9] 11.0 o x cs oF 16.8} 12.5 2 ra a 10.14 | 0.25 1S)5)) 2) Sf i ef 10.3 14.0} 10.9 o Ba OS) 10.9} 9.7 NaesHPO., NaOH, NaCl. .| 10.8 | 0.27 Tad lteS st af a MLO! es 13.7| 10.4 % ay peers elle sin | 0277 14.6} 9.3 91.2) 64.1) 15.0} 11.1 Helix arbustorum Upsala, Sweden The hemocyanin has two components, B and C, in the pH-range 3.6-8.0. The two components seem to influence each other in a way 322 I.-B. ERIKSSON-QUENSEL AND THE SVEDBERG that makes accurate determination of their sedimentation constants rather difficult. The mean value for B is 91.2 and for C 64.1. The proportions between B and C change in an irregular way. 0 0:2 0-4 0-6 0-8 1-0 1:2 Distance from meniscus, cm. Fic. 23a. Fic. 230. Fic. 23. Sedimentation pictures (a) and photometric records (6) for hemo- cyanin from Helix arbustorum at pH 7.7 (soo = 91.2 and 64.1); centrifugal force 50,000; time between exposures 5 minutes. 0 0-2 0:4 0-6 0:8 1-0 4:2 Distance from meniscus, cm. Fic. 25a. Fic. 250. Fic. 25. Sedimentation pictures (a) and photometric records (6) for hemocyanin from Helix nemorals at pH 8.5 (soo = 16.6); centrifugal force 125,000; time between exposures 5 minutes. Some runs indicate that in more dilute solutions the amount of C incteases. At pH 8.2 B and C are dissociated into a uniform compo- nent H with sedimentation constant 15.0. H is the single molecular species present in the pH-region 8.0-8.7. At pH 8.7 a small part of it is split up into a fourth component I with the sedimentation constant M.W. AND PH-STABILITY OF HEMOCYANINS 23 11.1. H and I are present together in the entire pH-region investi- gated above pH 8.7. The amount of I increases with pH. At pH 2.9 and below the hemocyanin is inhomogeneous. Figure 23 gives an example of a sedimentation run and Fig. 24 (Plate III) the pH- stability curve. Helix nemoralis Upsala, Sweden The main component, B, of this hemocyanin, has the sedimentation constant 101.0. It is present in the pH-range 3.6-8.0. Between pH 3.6 and 4.2 it is accompanied by component C with the sedimentation constant 65.0. At pH 7.7 and 8.0 there are also traces of C. Above pH 8 B and C are split up into the very homogeneous dissociation product H with the sedimentation constant 16.6. H is present alone up to 10.5 where I with the sedimentation constant 11.9 appears. H and I coexist to 10.8. In the region 10.8-11.5 I is alone. At 11.5 another dissociation product, still lower, is present together with I. Around pH 3 the material is rather inhomogeneous. At pH 2.5 H and I are formed. Figure 25 gives an example of a sedimentation run and Fig. 26 (Plate III) the pH-stability curve. Helix hortensis Upsala, Sweden In the pH-region 4.0—-8.0 this hemocyanin has two components, B with the sedimentation constant 100.0 and C with the sedimentation constant 61.9. B is the main component, C only forms 15 per cent of the whole protein up to pH 7.2. Between 7.2 and 8.0 B is present in a higher relative amount. In the pH-range 8—9.7 the material is split up into component H with the sedimentation constant 15.9. At 9.7 the splitting goes further, beside H component I with the sedi- mentation constant 12.2 is present. At pH 11.3 and above H dis- appears. In the region 3.6-4.0 C is present alone and in more acid solutions the hemocyanin is split up into inhomogeneous, more low- molecular dissociation products. Figure 27 (Plate III) gives the stability diagram. Limax maximus Varmdon, Sweden The main component, B, has the sedimentation constant 97.3. It is present in the pH-range 4.0-8.0 together with a higher component, A. A is rather inhomogeneous and low in concentration, therefore the determination of its sedimentation constant is uncertain. As mean value 136 has been found. At pH 4.0—4.2 a lower component C with 524 L-B. ERIKSSON-QUENSEL AND THE SVEDBERG PLATE III pH-stability diagrams for hemocyanins of species listed below. Abscisse in all figures, pH; ordinates, so. The dotted lines indicate the positions of the isoelectric points. Fic. 22. Helix pomatia. Fic. 24. Helix arbustorum. Fic. 26. Helix nemoralis. Fic. 27. Helix hortensis. M.W. AND PH-STABILITY OF HEMOCYANINS 525 the sedimentation constant 61.4 appears. At pH 3.6-3.8 the hemo- cyanin is further dissociated into three components, G, H, and I, with the sedimentation constants 25.5, 18.0, and 13.7. Above pH 8 A and B are dissociated into component H. Figure 28 (Plate IV) gives the pH-stability diagram. TABLE XV Helix nemoralis Dilution of blood 10 times; centrifugal force in the pH-range 2-8 50,000 (speed 26,000 r.p.m.); above pH 8 210,000 (speed 54,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. PH of! Total molar $20 $20 $20 $20 Solvent Cc HH I vent 0.30 US 40 lal) HCl, Na-citrate, NaCl... ths on Sent 0.30 34.7, 23.2 inhomogene- ous 52.3, 23.7 inhomogene- ous 98.8 99.3 96.9 98.9 99.5 99.1 100.4 104.4 107.7 99.5 100.7 103.4 101.0 103.7 100.1} + OS) By | ae 104.9 |69.4 Bees 4.6 5.0 KH2PO;, NasHPO,, NaCl bc 3 LCibe - Na2B,O7, NaCl . bé ims bc NasCO:NesBiOn) Nac ae 11.6 11.8 10.5 NaOH, NasHPO,, NaCl.. sc“ bc bc 101.0 |65.0 1257, 12.8 16.6 |11.9 9.0 8.5 526 I-B. ERIKSSON-QUENSEL AND THE SVEDBERG Conchifera: Cephalopoda: Decapoda. WHemocyanins from three animals belonging to this group: Sepia officinalis, Rossia owenit, and Loligo vulgaris, have been investigated. The main component of their blood has the sedimenta- tion constants 55.9, 56.2, and 56.7 respectively. They are identical within limits of experimental error. Moreover, their final alkaline dissociation products are the same. Sepia and Loligo have components with the sedimentation constants 18.7 and 16.9. These probably represent a splitting of the original molecule into the same number of TABLE XVI Helix hortensis All conditions as in Table XIV Solvent feo) eee Na-citrate, HCl, NaCl...]| 2.5 | 0.20 Very inhomo- geneous a om See S229) OLS0. 28.1, 19.3, 13.0 Very inhomo- geneous ue a Senta e ese le O2S0 Son 31.1 inhomo- geneous HAc, NaAc, Na@l.. 2... 3.6 | 0.20 60.7 24.0 inhomo- geneous ee < Cah ae 4.0 | 0.20 | 96.5} 60.3 4 “ OGG i a8 a 4.2 | 0.20 | 100.5 | 62.8 tc “a COPE AOE a 4.6 | 0.20 | 101.1 | 58.9 5) ; , , is Es a 6.5 | 0.20 | 100.1 | 61.4 . in i 6.6 | 0.20 | 98.6] 62.0 i Fe v 7.0 | 0.20 | 99.5 | 66.7 a a % 7.2 | 0.20 | 98.0} 63.4 : ah i 7.7 | 0.20 | 99.1 | 65.4 - ‘ IKECI et WOR + i a Nasb.O7 NaGlo jl) 18-04 (90:23 16.6 Na2CO3, NaeBsO7, NaCl..| 9.5 | 0.23 14.8 i cn cae 9.7 | 0.20 IG5) Se . ne i 10.0 | 0.23 VO28 | iste i i lb MOS | R28) 1.3} |) se tt ee eal OKO OAD 14.7 | 12.3 zi Sa UO OES) 15.8 | 13.8 NaOH, NasHPO,, NaCl. .| 11.3 | 0.30 ae ys) G i tel tes) J) OES 8.99* HOOLOR ROICO Ss eno eat Arat * Not used for the calculation of mean value. M.W. AND PH-STABILITY OF HEMOCYANINS 52M, fragments. In all three of them there is observed an inhomogeneous component of rather high sedimentation constant which exists in a narrow range inside the pH-region where the main component is stable. Octopoda. This group is represented by two animals, Octopus vulgaris and Eledone moschata. The normal sedimentation constants for the hemocyanin in their blood are 49.3 for Octopus and 49.1 for Eledone. Only one dissociation product is found and this is identical TABLE XVII Limax maximus Dilution of blood 10 times; centrifugal force in the pH-range 3.6-8.0 50,000 (speed 26,000 r.p.m.); above 8.0 175,000 (speed 50,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. Solvent ia ot | Boe ea ea alpen (ogee a HAGCINaAcWwNaGl aw aces ees. 3.6 | 0.20 24.3 |18.6 |13.7 oH ° SN Casares ins ean tare 3.8 | 0.20 26.6 119.9 |13.6 ie a Menus IU Fair Va 4.0 | 0.20 =e 95.2 |60.8 - Mcafee pes ear eres 4.2 | 0.20 | 143.0] 96.3 |63.0 ps he OP ipelatenast Reltat Alito ene 4.6 | 0.20 + | 100.7 * oS cra ters ate SOP WW O:207 26s eo 720 oO a Near, etn ARS. 5.0 | 0.20 + | 100.5 KH»2PO., Na2HPO,, NaCl...... 5.0 || O20 a 97.8 “ i ig eres ee 5.9 | 0.20 | 128.9] 93.0 - e Sa cae tert Uo 6.4 | 0.20 | 129.1} 93.2 i " ICs ae 6.8 | 0.20; + 99.3 ra a Nal@l a. ee 2.2 | 0.20 | 153.1) 92.4 “ a paeeisdtose 7.4 | 0.20 = 96.1 nf * pedehe er han3 8.0 | 0.20 + | 105.6 a Na2BiOz, NaiG@lrn.) ar 8.5 0.25 17.8 F mn PREM oe oie 9.0 | 0.25 17.1 NazCOs, NasBsOz, NaCl Cue en ad 10.0 0.25 16.6 ISOHO I Gs) (Ol 25) sO pil sa with the final one in the blood from the decapods. The molecular weight of this small component probably represents 1/7 of the normal molecular weight of the octopod hemocyanin and 1/8 of the molecular weight of the decapod hemocyanin. Loligo vulgaris Naples, Italy The hemocyanin appears in four different molecular species, namely, the components D, F, H, and I. Of those D, H, and I are well-defined and homogeneous; they give the sedimentation constants 56.7, 16.9, and 12.1. F is very inhomogeneous and it appears irregu- 528 1-B. ERIKSSON-QUENSEL AND THE SVEDBERG larly. The mean value of its sedimentation constant is 35.9. Component D exists in the pH-range 5.0-10.0 with the exception of a narrow region around 8. It is present alone between pH 5.0 and 7.2, accompanied by H from pH 7.2 to 10.0 and by F from 7.7 to 8.7. Component I appears together with D and H in the pH-range 9.6-10.0 and is the single component present above 10.0. Above pH 11 it is TaBLeE XVIII Loligo vulgaris Dilution of blood 10 times; centrifugal force in the pH-range 5—10 60,000 (speed 30,000 r.p.m.); above pH 10 260,000 (speed 60,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. Salven ate, | Soe eae [BUA INGVANG. INEGI GW auog ode ae see 5.0 0.20 | 55.6 KEEPO, NasiPOs NaCl......°.. 5.5 0.20 | 55.9 rt is Se ot Siemens 5.9 0.20 | 56.3 i ah PMP es Roemer chy 6.5 0.20 | 56.9 i a pi daceem sect ye 8 6.8 0.20 | 54.1 a a a ar URE RR 7.2 0.20 58.9 Diez, A o Spat clei enna dad 0.20 | 57.9 | 39.3 “ os SP Se nee 8.0 0.20 =e 18.9 i 4 gratis SNe eis iets 8.0 0.20 31.5 | 18.3 ae Na2B.07, CSN Riad Syren veh 8.4 0.25 57.8 + + i i Pomnne oeb bares acte 8.7 0.25 57.1 | 37.0 | 16.6 a mee Ricsyaisratt 9.2 0.25 Silos 14.0 Na2COs, Tey Rene ae oa 9.4 0.25 57.6 fa) a Ne Se ane eee 9.6 0.25 55.3 16.9 | 12.6 ot Zi Tye Eacenac atte 9.8 0.25 + 16.2 | 12.5 a ri Bee te ceteris 10.0 0.25 56.6 + = i fe e 10.3 0.25 12.0 i f a il rrentnatesye es 10.6 0.25 11.4 4 Be DEN sieioronche eatiods 11.1 0.27 10.5 rf e a ES 0.30 9.4* ‘ Ba Meer ar te ited 0.35 9.0* 50.7 73529 OO neato * Not used for the calculation of mean value. still perfectly homogeneous with regard to molecular weight but its sedimentation constant drops gradually. This must be due to a change in shape of the particles. Below pH 5 aggregation sets in. Figures 29 and 30 give examples of sedimentation runs, Fig. 31 (Plate IV) gives the pH-stability curve. Rossia owent Kristineberg, Sweden The main component, D, has the sedimentation constant 56.2. It M.W. AND PH-STABILITY OF HEMOCYANINS 529 exists between pH 5.0 and 10.1; it is present alone in the pH-ranges 5.0-7.7 and 8.9-9.4, and together with component F in the pH-range 7.7-8.9, and together with component I in the pH-range 9.4-10.1. Component F is very inhomogeneous and present in small quantity, therefore its sedimentation constant is not well-defined. In one run the value 40.9 was found. Component I appears first in small quantity 0 0:2 0:4 0-6 0:3 1:0 1-2 Distance from meniscus, cm. Fig. 29a. Fic. 290. Fic. 29. Sedimentation pictures (a) and photometric records (6) for hemo- cyanin from Loligo vulgaris at pH 6.8 (seo = 56.7); centrifugal force 60,000; time between exposures 5 minutes. 0 0:2 04 06 06 10 12 Distance from meniscus, cm. Fie. 30a. Fic. 300. Fic. 30. Sedimentation pictures (a) and photometric records (6) for hemo- cyanin from Loligo vulgaris at pH 7.7 (soo = 56.7 and 35.9); centrifugal force 50,000; time between exposures 5 minutes. 530 L-B. ERIKSSON-QUENSEL AND THE SVEDBERG together with D, but its relative amount increases with pH. Above pH 10.1 it is present alone. It shows a behavior similar to component I in the Loligo-blood. In solutions more acid than pH 5 the hemo- cyanin is precipitated. Figures 32 and 33 give samples of sedimenta- tion runs and Fig. 34 (Plate IV) the pH-stability curve. TABLE XIX Rossia owentt Dilution of blood 15 times; centrifugal force below pH 10 60,000 (speed 30,000 r.p.m.); above pH 10 215,000 (speed 55,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. sven ne | Be | ee EbNe Neen NaGls. 2504... 5.0 0.24 59.3 KELEOM NasHEO,) NaCl). |. 5.5 0.24 56.0 a5 a oe ae a 5.9 0.24 55.0 : r SO be 6 0.24 53.7 : a iH A aaa 64 0.24 52.9 u r Epes 3 6.8 0.24 57.0 : a mere 2 Og 7A 0.24 53.8 : “ a eee ie 0.24 57.7 ac SOREN ERB Op vile e 8.4 0.30 56.4 li : “ (rie 8.9 0.30 57.5 40.9 NazCOs, “ ee Fae 9.2 0.30 56.5 u ue aay a 9.4 0.30 56.3 ae « a COR ee 9.8 0.30 58.8 ae c « re I 10.1 0.30 59.0 13.0 e “ ‘ae a 10.3 0.30 11.4 ‘ e Cinta 10.6 0.30 10.4 NaOH, NasHPO,, NaCl....... 11.1 0.39 10.3 a « CR is Sot 11.5 0.47 10.4 i “ ek aN Ahh 0.40 10.0 56.2 10.9 Sepia officinalis Roscoff, France The sedimentation picture of this blood shows two components, D and H, in the pH-range 4.0-8.5. D has the sedimentation constant 55.9 whereas the sedimentation constant of H is 18.7. From pH 4 to 7 the proportions between D and G seem to be constant. From the centrifuge curves it can be estimated roughly that the percentage of D is 75, of H 25, assuming that they have the same light absorption. The amount of H rises towards the alkaline side. In the region 8.5-10.4 another component is observed, which is very inhomogeneous M.W. AND PH-STABILITY OF HEMOCYANINS 531 and with a sedimentation constant higher than H. In this region the inhomogeneity of the material makes the measurements rather uncer- tain. At pH 10.6 a homogeneous dissociation product, I, is formed. Its sedimentation constant is originally 11.8 but drops gradually as in the case of Loligo and Rossia. At pH 4 an increase in the sedimenta- tion constants of D and H shows that aggregation setsin. At 3.6 the 0-2 0:4 0:6 0-8 1:2 Distance from meniscus, cm. Fic. 32a. Fic. 326. Fic. 32. Sedimentation pictures (a) and photometric records (6) for hemocyanin from Rossia owenti at pH 6.0 (Seo = 56.2); (centrifugal force 70,000; time between exposures 5 minutes. 0 0-2 0-4 0-6 0:8 1-0 12 Distance from meniscus, cm. Fic. 33a. Fic. 330. Fic. 33. Sedimentation pictures (a) and photometric records (b) for hemocyanin from Rossia owentt at pH 10.1 (soo = 56.2 and 13.0); centrifugal force 120,000; time between exposures 5 minutes. 532 I.-B. ERIKSSON-QUENSEL AND THE SVEDBERG protein is precipitated. Figures 35 and 36 give examples of sedimenta- tion runs, Fig. 37 (Plate IV) the pH-stability diagram. Octopus vulgaris Roscoff, France The main component, E, of this hemocyanin has the sedimentation constant 49.3. It is present in the blood in the pH-region 4.0—-10.3. TABLE XX Sepia officinalis All conditions the same as in Table XVIII. pH of | Total S20 S20 S20 Solvent solvent] molar D H I INevave, ISA, INC cocescodass 4.0 | 0.20 | 62.6* |27.3* Inhomogeneous “cc sé tas Systolic ae rae 4.6 | 0.20 | 58.9 |23.9 A BTN Poe tots 5:07 | 0:20) 53741829 ui ze Oe Cee SR 5254/0220) 5608 ae KH2PO,, NasHPO,, NaCl......| 6.0 | 0.20 | 54.7 |20.9 oe i He oie) Os. | O40) Ssissi lS.) ei i Oak oebel) Osa OHO Sieh \ih76 f Na2B.0;7, i See 8.5 | 0.25 | 54.3 |18.5f NasCOs, i * 9.0 | 0.25 | 55.2 +f eo iS OO gocuocl OF | O25 15380 200° ne s Hears 10.0 | 0.25 + |20.5* Inhomogeneous +e ie 8 sccccol| LOG |) O25 11.8 INBOE PNA de OpuNaGlen es. - 11.1 | 0.30 10.5 e LON Ae aartae 11.5 | 0.30 10.9 a “ Seer Se pice 11.7 | 0.30 9.1 55:9 LSE Ose * Not used for the calculation of mean value. + Three components present, the third with a sedimentation constant between DandH. At 9.4 the value 42 was obtairjed. At pH 9.5 it is dissociated to a small degree into component I. The relative amount of I increases with pH. It appears together with E in the pH-region 9.5—10.3, but above pH 10.3 it is alone. The sedi- mentation constant at pH 10.3 is 12.1, but above it drops. I seems to be identical with the alkaline dissociation products of the decapods. In the pH region 4.0-9.5 E is generally the only component present; in three runs it was accompanied by a lower component each time with different sedimentation constant. Figure 38 (Plate IV) gives the pH-stability diagram. M.W. AND PH-STABILITY OF HEMOCYANINS 533 The pH-stability diagram has been studied here before (Svedberg and Eriksson, 1932). If the value given there for the sedimentation constant of the main component is corrected for the influence of density and viscosity of the buffer solutions, it becomes 52.6. This agrees fairly well with the present determination. After correction 0 0-2 0-4 0:6 0:8 1-0 1-2 Distance from meniscus, cm. Fic. 35a. Fic. 350. Fic. 35. Sedimentation pictures (a) and photometric records (6) for hemocyanin from Sepia officinalis at pH 5.5 (soo = 55.4 and 18.7); centrifugal force 71,000; time between exposures 5 minutes. 0 -0:2 0:4 0-6 08 1-0 1-2 Distance from meniscus, cm. Fic. 36a. Fic. 360. Fic. 36. Sedimentation pictures (a) and photometric records (6) for hemocyanin from Sepia officinalis at pH 11.5 (sx = 10.6); centrifugal force 220,000; time between exposures’ 5 minutes. 534 L.-B. ERIKSSON-QUENSEL AND THE SVEDBERG this value is constant inside the pH-region in which the main compo- nent is present. During the earlier investigation the blood was not stored in frozen condition as now and was dialyzed for some days against the buffer instead of being diluted just before the run. This might explain the appearance of a low-molecular component with a sedimentation constant changing with pH. A similar component has only appeared in a few runs in this work. TABLE XXI Octopus vulgaris Dilution of blood 20 times; centrifugal force 100,000 (speed 37,000 r.p.m.); thick- ness of column of solution 1.2 cm.; other conditions as in Table II. sven ee | 7a ae INBVANG, JRUAG. INTE CI eae a eas ane 4.0 0.20 51.6 2 a Baers tare tues, de 3 4.6 0.20 52.4 = Es dete eben 2 sel 5.0 0.20 49.9 NazH PO,, KH:PO,, NaCl areliciosona 5.6 0.20 47.6 es es Seigtaor ut 6.0 0.20 48.8 a pe Wraaraten ates 6.3 0.20 48.5 S57) e s sere Mea 6.8 0.20 47.6 cs - Sn aes saree es 0.20 48.0 N Se faceplates Toll 0.20 48.9 20.1 in a Fae ea ts 8.0 0.20 47.8 15.2 Na2B.O7, A ue a6 Jone 8.4 0.25 45.7 os ‘ Se teeenee ts 8.8 0.25 45.2 zs INasC Ose Se aeeee 9.0 0.25 49.5 ae a Baee eee 9.3 0.25 51.1 2 i “ii eae 9°5 0.25 51.9 ae fy “e Se ora 9.8 0.25 52.8 aR 4 a obey tiie 2) 10.3 0.25 51.4 12.1 cs Hh AN eee 10.8 0.25 11.3 49.3 Eledone moschata Naples, Italy The main component E of the hemocyanin is present from pH 5 to 11.0. The sedimentation constant is 49.1. At 10.6 about ten per cent of the protein splits up into smaller molecules having the sedimentation constant of 11.8 (component I). At 11.0 the percentage ° of the lighter protein goes up to 50. At pH 11.1 the splitting is complete. The sedimentation constant of the very uniform dissocia- tion product drops with increasing pH. At pH 5.4 the protein is very inhomogeneous. In more acid solutions precipitation sets in. M.W. AND PH-STABILITY OF HEMOCYANINS 535) Figures 39 and 40 give examples of sedimentation runs, Fig. 41 (Plate IV) the pH-stability diagram. REVERSIBILITY Some experiments have been carried out on the reversibility of the dissociation reactions and the equilibrium between the dissociation TABLE XXII Eledone moschata Dilution of blood 15 times; centrifugal field 75,000 (speed 32,000); in the four last runs 230,000 (speed 57,000 r.p.m.); thickness of column of solution 1.2 cm.; other conditions as in Table II. Sven ae SP UNC wUNaANGy NaCl ohn de 5.4 | 0.20 | 49.4 Very inhomogeneous KEG Om NaskiieO xe NaGlie 25-244) .5.9 | O20) F497 + ss Ser eA 6.2 | 0.20 | 47.8 oe es Ban aie ee ee 6.5 | 0.20 | 48.7 % of be 6.8 | 0.20 | 48.1 ¥ ese aioe 7.1 | 0.20 | 46.7 y By Sr ACAD cathy 8 7.4 | 0.20 | 48.5 i “ Pats Sa teanse es 7.7 | 0.20 | 49.7 pi sh sae Wee eee 8.0 | 0.20 | 50.0 i Na2B.07, RON nets 8.5 0.25 48.2 NasCOs, € Stl wes a ee 9.1 | 0.25 | 50.8 * i rurvehene te oes 9.5 | 0.25 | 50.3 * a SOI ie yee 9.8 | 0.25 | 49.9 ve 4 Ui ae eae 10.3 | 0.25 | 50.1 ES i Ne pe ee fe 10.6 | 0.25 | 49.8 | 11.8 ss * Ean eeacets core HOR SOR25 917482211025 hy ra eae ace Lele te O25) 11.4 INGOURE, INGRENEO KE INEKOQNE Sos 660 46 11.5 | 0.30 9.46 io a Bik Sareea eee 11.5 | 0.30 9.59 ss + AUREL ots 11.7 | 0.30 8.57 49.1 | 10.60 products. Two hemocyanins have been chosen for these tests, viz. the blood pigments of Helix pomatia and Limulus polyphemus. They can be regarded as representatives for two different groups. In the natural blood and in a large part of the investigated pH-region the Helix hemocyanin consists of just one molecular species. The Limulus hemocyanin, on the contrary, contains four different components in its natural state and inside the main part of the stability region (from pk Stor!) 536 I-B. ERIKSSON-QUENSEL AND THE SVEDBERG Helix pomatia The reversibility of the dissociation caused by a change in pH was studied. The results are summarized in Table XXIII. The blood was taken to pH, chosen so that some sort of dissociation took place, and then to pHyr; where it normally would have been less dissociated. In Table XXIII are given the values for the sedimentation constants pened soon «ROSNER ENN ANNO RUM (0) 0:2 0:4 0-6 0-8 1:0 1:2 Distance from meniscus, cm. Fic. 39a. Fic. 390. Fic. 39. Sedimentation pictures (a2) and photometric records (6) for hemocyanin from Eledone moschata at pH 9.8 (Soo = 49.1); centrifugal force 72,000; time between exposures 5 minutes. 0 0-2 0:4 0:6 0-8 1-0 1:2 Distance from meniscus, cm. Fic. 40a. Fic. 400. Fic. 40. Sedimentation pictures (a) and photometric records (6) for hemocyanin from Eledone moschata at pH 11.0 (sx = 49.1 and 10.6); centrifugal force 72,000; time between exposures 5 minutes. M.W. AND PH-STABILITY OF HEMOCYANINS S)0// at pHry, the normal sedimentation constants at pHu, and the sedi- mentation constants found at pHn when the protein had first been taken to pH;. The results show that the dissociation is reversible inside a certain pH-region, but if the solution is made too alkaline only a part of the hemocyanin is associated into the normal compo- nents. Another part of it aggregates, but into smaller particles than normally found. If the solution is brought to pH 12 only inhomo- geneous material is formed by neutralization. The equilibrium between the components was studied in the TABLE XXIII Helix pomatia s20 for comp. pH, | pHy;z| szo At pHy normal at seo for comp. found at pHy, pHyy 8.0 | 6.8 | 98.9] 62.0 16.0 | 98.9 95.7 8.9 | 6.8 16.0 | 98.9 99.0 Sto ded 16.0 | 98.9 62.0 60.7 17.9 Inhomogeneous 9.5 | 5.9 16.0 12.1 | 98-9 106.8 66.1 22.3 10.3 | 5.0 16.0 12.1 | 98.9 94.1 63.4 23.0 10.5 | 6.8 16.0 12.1 | 98.9 97.9 64.8 26.2 10.8 | 5.9 16.0 12.1 | 98.9 ODS CO. BRT 10.8 | 7.7 16.0 12.1 |98.9 62.0) + 59.8 + 10.8 | 5.9 16.0 IZA iL | Oso 109.8 70.5 23.6 12.1 | 5.9 Inhomogeneous] 98.9 | 32.3 Inhomogeneous 4.0 | 6.8 | 98.9} 62.0 98.9 85.7 Two compo- nents, bad separation 4.0 | 5.4 | 98.9} 62.0 98.9 100.9 following way: a solution at pH 8.0, containing the components B and C, was run in the ultracentrifuge until component B had sedimented to the bottom of the cell. The centrifuge was stopped and the liquid pipetted off. The sediment was stirred up in a small quantity of the same buffer. Runs were made on these two solutions. The initial solution contained 67 per cent B and 33 per cent C, the solution pipetted off 73 per cent C and 27 per cent B and the sediment 67 per cent C and 33 per cent B. These figures agree within limits of error. The relative amounts are obtained from the microphotometric curves of the sedimentation pictures assuming that the different components have the same light absorption. The dissociation products seem to be in equilibrium. If the relative concentration of one is diminished the same proportions are reéstablished. 538 I-B. ERIKSSON-QUENSEL AND THE SVEDBERG Limulus polyphemus The reversibility on changing the pH was studied at pH 10.8, where the protein is dissociated into a product with sedimentation constant 5.87, and at 4.0, where it contains three components. The dissociation reactions were completely reversible. In order to study the equilibrium between the components the same procedure as in the case of the Helix hemocyanin was carried out. A solution containing the four components D, F, H, and K was centrifuged until D had reached the bottom. In the initial solution the composition was: 22 per cent D, 14 per cent F, 41 per cent H, and 23 per cent K. In the pipetted-off solution 14 per cent D, 11 per cent F, 55 per cent H, and 20 per cent K. In the sediment 58 per cent D, 17 per cent F, 25 per cent of H and K together. Evidently the sediment contains a larger percentage of the heavy component D and less of the light components H and K, while the supernatant solution contains less of D and more of H and K. The solutions were stored for a week and run again. The values then obtained agreed with those given above within the limits of error. The components therefore do not seem to be in equilibrium in this case. If the components are in equilibrium the equilibrium must be disturbed during the ordinary runs. The concentration of the higher components in the solution decreases faster than the concentration of the lower components. Whether or not this shows in the sedimenta- tion diagrams must depend on the rate of reéstablishment of the equilibrium and the time taken by the run. In some cases such an effect has been observed. EXPLANATION OF PLATE IV pH-stability diagrams for hemocyanins of species listed below. Abscissz in all figures, pH; ordinates, sy. Fic. 28. Limax maximus. Fic. 31. Loligo vulgaris. Fic. 34. Rossia owenit. Fic. 37. Sepia officinalis. Fic. 38. Octopus vulgaris. Fic. 41. Eledone moschata. The dotted line in Fig. 41 indicates the position of the isoelectric point. PLATE IV 540 I.-B. ERIKSSON-QUENSEL AND THE SVEDBERG MOLECULAR WEIGHTS A number of sedimentation equilibrium measurements have been carried out so as to make it possible to codrdinate sedimentation constants and molecular weights. In cases where more than one molecular species is present the computation of molecular weights from equilibrium data is uncertain and we have therefore refrained 100 Percentage of initial concentration s 0 0:2 04 0-6 0:8 4:0 Distance from meniscus, cm. Fic. 42a. Fic. 420. Fic. 42. Sedimentation pictures (a) and concentration curves (0) for hemocyanin from Helix pomatia at pH 5.5 (Soo = 98.9); centrifugal force 45,000; time between exposures 5 minutes. {00 Vie 0 0-2 0-4 06 08 10 Percentage of initial concentration Distance from meniscus, cm. Fic. 43a. Fic. 430. Fic. 43. Sedimentation pictures (a) and concentration curves (6) for hemo- cyanin from Helix pomatia at pH 8.2 (soo = 98.9, 62.0, and 16.0); centrifugal force 60,000; time between exposures 5 minutes. M.W. AND PH-STABILITY OF HEMOCYANINS 541 from studying solutions of this kind. The results are summarized in Table XXIV. Mg DISCUSSION OF RESULTS The most striking result of this investigation is perhaps the perfect homogeneity of the various hemocyanins with regard to molecular weight. Not only the main component or components which exist in the region comprising the isoelectric point (Fig. 42) but also the TABLE XXIV Sedimentation equilibrium determinations on hemocyanins S n “a 4 yo] =] ag ae ea|¢ fe) = Species 38 Solvent oo | &|# O15 esl z 5 M M Mean 2a Se zo, 4 se} = = a |e A a Pandalus 7 |\NazHPOs, KH2POa, NaCl] 0.20} 6.8 4 | 3800] Scale] 160 408000 397000 borealis 7 ne a “| 0.20} 6.8 4 | 3800] Scale] 132 386000 Palinurus 10 |NasHPOs, KH2PO., NaCl} 0.20] 6.8 | 12 | 3330} Slit | 120 485000 vulgaris 10 oe oy ““ 10.20] 6.8 | 12 | 3060] Scale} 112 462000 10 i sr oe 0.20 | 6.8 12 | 3540] Scale} 154 450000) 447000 10 * YS ** 10.20] 6.8 | 12 | 3300} Scale] 131 409000 10 ih oS us 0.20) 6.8 12 | 3300] Scale} 117 429000 Homarus 7 |\NaeHPO.s, KH2POs, NaCl] 0.20} 6.8 | 12 | 3700} Slit | 137 793000 vulgaris 7 - NS ““ 10.20] 6.8 | 12 | 3700} Slit 112 798000} 803000 7 a cate a 0.20} 6.8 A | 3700] Scale} 214 817000 Calocaris 4 |NasHPOs, KH2PO:, NaCl) 0.20) 6.8 4 | 2600} Scale} 249 | 1.3060C0 1.329000 macandre 4 Bo 3 “ 10.20] 6.8 4 | 2700} Scale} 226 | 1.352000) ~~ Helix 20 |Na2HPO:, KH2ePO:s, NaCl] 0.20] 6.8} 12 | 1500} Slit | 168 | 6.520000 pomatia 60 oe nN ** 10.20] 6.8 | 12 | 1500) Slit | 208 | 6.658000) 6.680000 120 ve oy Ss 0.20 | 6.8 12 | 1620} Scale] 156 | 6.864000 20 |KH2POs, Na2Bs07, NaCl | 0.25} 8.5 4 | 3180} Slit 161 819000 797000 20 re i i 0.25} 8.5 4 | 3162] Scale] 238 774000 products of reversible dissociation (Fig. 43) are definite molecular individuals. At very low and very high pH-value where irreversible dissociation occurs the products are inhomogeneous (Fig. 44). Com- pare with Fig. 45. The components of highest molecular weight always exist in a region comprising the isoelectric point. As a rule the number of components increase from the isoelectric point towards the alkaline and acid side. Among the 22 species studied 13 have only one component at the isoelectric point, 7 have two, 1 three and 1 four components. The sedimentation constants observed may be arranged in ten classes (A-K), as shown by Table XXV._ For each class the variations are almost within the limit of experimental error. To a certain molecular weight may correspond more than one sedimentation 542 1-B. ERIKSSON-QUENSEL AND THE SVEDBERG constant according to the more or less pronounced deviation from the spherical shape which the molecule shows (Table XXVI). The molecular weights of the hemocyanin found in the blood of a certain species are always simple multiples of the lowest well-defined compo- —_ Cc oO Percentage of initial concentration 3 0-4 0:6 08 1-0 1-2 o 2 1 Distance from meniscus, cm. Fig. 44a. Fic. 440. Fic. 44. Sedimentation pictures (a) and concentration curves (0) for hemo- cyanin from Nephrops norvegicus at pH 11.7 (sx = 8.95); centrifugal force 220,000; time between exposures 10 minutes. 100 Percentage of initial concentration Ss 0 0:2 0-4 0-6 0:8 1-0 Distance from meniscus, cm. Fia. 45a. Fic. 450. Fic. 45. Sedimentation pictures (a) and concentration curves (0) for hemo- cyanin from Nephrops norvegicus at pH 4.5 (soo = 24.5); centrifugal force 150,000; time between exposures 5 minutes. M.W. AND PH-STABILITY OF HEMOCYANINS 543 “nent. Thus for the Malacostraca the relationship is 1 : 2 and for the Gastropoda 2:8:16:24. Moreover, the weights of all the well- defined hemocyanin molecules seem to be simple multiples of the lowest one among them (Table X XVI). In most cases the hemocyanin components in the blood of a certain species are interconnected by reversible, pH-influenced dissociation- TABLE XXV Sedimentation constants arranged in classes Species A B Cc D E F G H IL K CRUSTACEA: Malacostraca Pandalus borealis......... 22.9) 17.4 + Palinurus vulgarts........ 16.4 4.10 Nephrops norvegicus....... 24.5} 17.1 5.97 Homarus vulgaris......... 22.6) 16.1 + Astacus fluviatilis......... 23.3) 16.3 4.02 Carcinus M@nds.......... 23.3} 16.7 + Cancer pagurus........... 32.7| 23.6) 16.4 4.71 ARACHNOMORPHA: Xiphosura Limulus polyphemus....... 56.6 34.6} 24.0) 16.1 5.87 CONCHIFERA: Gastropoda Littorina littorea.......... 99.7| 63.3 15.0 Neptunea antiqua......... 104.0 14.3) + Buccinum undatum........ 102.1) 63.8 + Busycon canaliculatum..... 130.4} 101.7) 61.1 13.5 Helix pomatia............ 98.9} 62.0 16.0} 12.1 SNL OUSTOTIITL Ee asa 91.2) 64.1 15.0} 11.1 UN nemoralas 2. esi. seen: 101.0} 65.0 16.6) 11.9 ST OVLEMSUSin Miwa Spare e 100.0} 61.9 15.9} 12.1 Limax maximus.......... 136.1) 97.3) 61.9 25.5] 18.0} 13.7 Cephalopoda Decapoda Loligo vulgarts........2... 56.7 35.9 16.9} 12.1 Sepia officinalis........... 55.9 + 18.7] 10.6 Rossta owentt............ 56.2 + 10.9 Octopoda OOPS THOUS 6 os coe bs 49.3 + Eledone moschata......... 49.1 10.6 association reactions. At certain critical pH values a profound change in the number and percentage of the components takes place. The shift in pH necessary to bring about the reaction is not more than a few tenths of a unit. Consequently the forces holding the dissociable parts of the molecule together must be very feeble. This fact makes it all the more surprising that an equal backward shift in pH does 544 1.-B. ERIKSSON-QUENSEL AND THE SVEDBERG really cause an association of the parts to exactly the same molecule. For this reason it is justified, we think, to use the term molecule for the hemocyanin particles despite their enormous size. Their mass cannot be changed continually as in the case of an ordinary colloid particle and they have a finite stability range with regard to the surroundings. The hemocyanin particles, therefore, most probably possess a definite structure. TABLE XXVI Species pH | Sa | Doo Ms Me Meale. Pandalus borealis........ 6.8 | 17.4 397,000 Palinurus vulgaris........ 6.8 | 16.4 |3.4 446,000 | 447,000 Helix pomatia........... 9.7 | 12.1 |2.23 | 502,000 1x 406,000 Busycon canaliculatum....| 9.6 | 13.5 |3.29 379,600 =1233,800 Eledone moschata......... 11.6 | 10.6 |2.16 | 457,000 Nephrops norvegicus...... 6.8 | 24.5 |2.79 | 812,000 Homarus vulgaris........ 6.8 | 22.6 |2.78 752,000 803,000 | 2 406,000 Helix pomatia........... 8.6 | 16.0 |2.06 | 719,000 | 797,000 = 812,000 Helix nemoralis.......... 8.6 | 16.6 |1.92 799,000 Calocaris macandre...... 6.8 | 34.0 1,329,000 | 4 406,000 = 1,624,000 Octopus vulgaris.......... 6.8 | 49.3 |1.65 | 2,785,000 7X 406,000 Eledone moschaia......... 6.8 | 49.1 |1.64 | 2,791,000 = 2,842,000 Rossid OWeN11............ 6.8 | 56.2 |1.58 | 3,316,000 8 x 406,000 = 3,248,000 Helix pomatia........... 6.8 | 98.9 |1.38 | 6,630,000 | 6,680,000 | 16x 406,000 Busycon canaliculatum....| 6.8 |101.7 |1.38 | 6,814,000 = 6,496,000 Busycon canaliculatum....| 6.8 |130.4 |1.17 | 9,980,000 24 < 406,000 =9,744,000 The pH-stability diagrams show a number of regularities indicative of biological kinship. All the Malacostraca give similar diagrams and so do the Gastropoda. The Cephalopoda show two types of diagram, one characteristic of the Decapoda and one of the Octopoda. Only one species of the Xiphosura, viz. the Limulus, has been studied. It shows a diagram entirely different from that of the Malacostraca species studied. The placing of Limulus among the crustaceans as was formerly done is therefore not supported by our sedimentation analysis of fits blood pigment. M.W. AND PH-STABILITY OF HEMOCYANINS 545 Within the large groups mentioned above an attempt at classifica- tion based upon the similarity of the stability diagrams would give, roughly, the following result: Malacostraca: Pandalus and Palinurus—main comp. s = 16. Nephrops and Homarus—main comp. s = 23. Astacus, Cancer and Carcinus—both comp. s= 16 and s = 23 through most of the stability range. Xiphosura: Limulus—four comp. s = 57, 34, 16, and 6 through most of the stability range; at the acid side also comp. s = 23. Gastropoda: Pulmonata: Helix and Limax—comp.'s = 100 and 61 on both the alkaline and acid side of the isoelectric point. Aspidobranchia: Iittorina, Neptunea, Buccinum, and Busycon—comp. s = 61 only on the alkaline side. Cephalopoda, Decapoda: Loligo, Rossia, and Sepia—main comp. s = 57. Cephalopoda, Octopoda: Octopus and Eledone—main comp. s = 49. Really striking differences in the stability diagrams only occur for species belonging to different orders. All species of one and the same order have similar diagrams. Within the order Cephalopoda the two sections Decapoda and Octopoda show rather different diagrams, though. From the point of view of the blood sedimentation analysis, therefore, the Decapoda and the Octopoda ought perhaps to be regarded as systematic units of the same magnitude as the Gastropoda, i.e. as different orders within the class Mollusca, having developed as parallel branches from common ancestors far down in the Mollusca cladus. It is interesting to see that there is a notable difference between the diagrams for the species belonging respectively to the suborders Pulmonata and Aspidobranchia. Similarities between species belonging to the same family or the same genus are sometimes revealed by their stability diagram (e.g. Nephrops and Homarus; Helix pomatia and Helix nemoralis) but species from two different although closely related genera may have very similar diagrams (e.g. Helix hortensis and Limax maximus). Among the species studied by us no two have quite identical stability diagrams. It should be possible, therefore, to make a species determination based solely on the stability diagram. In such cases where the stability diagrams are rather similar (e.g. Helix pomatia and Helix nemoralis) a determination of the isoelectric point would settle the question (5.05 for Helix pomatia and 4.63 for Helix nemoralis) (Pedersen, 1933). 546 I-B. ERIKSSON-QUENSEL AND THE SVEDBERG The expenses of this investigation have been defrayed by grants from the Andersson Foundation, the Astra Co., the Nobel Foundation, the Rockefeller Foundation and the Wallenberg Foundation. Living material of the following species were put at our disposal by the Zodlogical Station at Kristineberg, Sweden, viz. Pandalus boreals, Nephrops norvegicus, Carcinus menas, Buccinum undatum, Littorina littorea, Neptunea antiqua, and Rossia owenw. For this valuable help we are indebted to Professor E. Lénnberg and Dr. G. Gustafson. Samples of blood from Loligo vulgaris and Eledone moschata were kindly sent us by Professor R. Dohrn and Dr. L. Califano of the Zoodlogical Station at Naples. The Limulus and Busycon blood was given by Dr. A. C. Redfield of the Oceanographic Institution at Woods Hole, Massachusetts. Blood from Palinurus vulgaris, Sepia officinalis and Octopus vulgaris was collected for us at the Marine Station, Roscoff, France, by Dr. Lindahl. For the permission to make this collection we are indebted to Professor Charles Pérez. SUMMARY 1. Blood from twenty-two species containing hemocyanin as respiratory pigment has been subjected to a detailed ultracentrifugal study. 2. Sedimentation constants have been determined throughout the pH-stability ranges. 3. In some cases sedimentation equilibrium measurements have been carried out. From these data molecular weight values were obtained and correlated with the sedimentation constants. 4. The various hemocyanins are perfectly homogeneous with regard to molecular weight both at the isoelectric point and in the regions where reversible dissociation or association occurs. 5. The stability range for the components of highest molecular weight always comprises the isoelectric point. 6. Except at extremely low and high pH values the dissociation reactions are reversible. 7. The molecular weights of the different hemocyanins and their reversible dissociation products show a relationship of simple multiples. 8. Species belonging to the same order have similar pH-stability diagrams. Each species has a characteristic diagram by means of which it can be identified. In doubtful cases the isoelectric point may be used as complementary evidence. M.W. AND PH-STABILITY OF HEMOCYANINS 54a REFERENCES Lamm, O., 1933. Nature, 132: 820. Lamm, O. AND A. Poison, 1936. Biochem. Jour., 30: 528. PEDERSEN, K. O., 1933. Koll.-Zeitschr., 63: 277. SVEDBERG, T., 1925. Koll.-Zettschr. Erg.-Band 2., 26: 53. SVEDBERG, T., 1927. Zeztschr. physikal. Chemie, 127: 51. SVEDBERG, T., 1933. Jour. Biol. Chem., 103: 311. SVEDBERG, T., 1934a. Chem. Rev., 14: 1. SVEDBERG, T., 19346. Science, 79: 327. SVEDBERG, T., 1934c. Koll.-Zettschr., 67: 1. SVEDBERG, T., 1934d. Naturwiss., 22: 225. SVEDBERG, T. AND E. CHIRNOAGA, 1928. Jour. Am. Chem. Soc., 50: 1399. SVEDBERG, T. AND I.-B. ErtKsson, 1932. Jour. Am. Chem. Soc., 54: 4730. SVEDBERG, T. AND A. HEDENIUus, 1933. Nature, 131: 325. SVEDBERG, T. AND A. HEDENIUus, 1934. Bzol. Bull., 66: 191. SVEDBERG, T. AND F. F. HEyrotn, 1929a. Jour. Am. Chem. Soc., 51: 539. SVEDBERG, T. AND F. F. HEyrotu, 19296. Jour. Am. Chem. Soc., 51: 550. TisELius, A. AND D. Gross, 1934. Koll.-Zettschr., 66: 11. INDEX A BRAMOWITZ, A. A. Physiology of the melanophore system in the catfish, Ameiurus, 259. Abstracts of scientific papers presented at Marine Biological Laboratory, summer of 1936, 386. Aenoplex smithii, spermatogenesis of, cytological phenomena, 375. Alcohol, effect on gastrulation in Arbacia, 413. Ameiurus, physiology of melanophore system, 259. Arbacia punctulata, cleavage without nuclei, 101. ARMSTRONG, P.B. Mechanism of hatch- ing in Fundulus heteroclitus (ab- stract), 407. Ascidians, excretion in, rdle of blood cells, 249. BACTERIA, multiplication in different volumes stored sea water, influence of oxygen tension and solid surfaces, 324. Bacteroids, intracellular, of roaches, 433. BALLARD, W. W. Observations on lens regeneration in Amblystoma (ab- stract), 388. Beams, H. W., AND R. L. Krnc. The effect of ultracentrifuging upon chick embryonic cells, with special reference to the ‘‘resting” nucleus and the mitotic spindle, 188. Bear, R. S. See Schmitt, Bear and Young (abstract), 402. BISSONNETTE, T. H. Modified sexual photoperiodicity in ferrets, raccoons and quail (abstract), 395. Blood cells, in excretion, ascidians, 249. Bowen, R. E. Sensory hairs of the dog- fish ear, 429. BozLER, E. The double refraction of smooth muscle (abstract), 402. BRAMBEL, C. E. See Cowles Brambel, 286. Brooks, MatirpA M. The effect of methylene blue on the spectropho- and tometric picture of hemoglobin, CO-hemoglobin and CN-hemoglobin (abstract), 386. : Burton, Atan C. The basis of the principle of the master reaction in biology (abstract), 393. —, —. See Sichel and Burton (abstract), 397. ALCIUM, free, in action of stimulat- ing agents on Elodea cells, 306. CAMERON, J. A., AND K.O. Mitts. Be- havior of frog tadpole epidermal cells during seven successive re- generation periods (abstract), 405. CaMERON, J. A. See Fisher and Cam- eron (abstract), 404. Carotenoid metabolism, in surf perches, structure and function of gut in, 217. CarotuHEers, E. ELEANOR. Cellular be- havior in abnormal growths pro- duced by irradiation of grasshopper embryos (abstract), 394. ; : Components of the mitotic spindle, 469. Catfish, melanophore system, physiology of, 259. CHAMBERS, R. Experimental studies on the oil wetting property of the plasma membrane (abstract), 398. Chick embryonic cells, effect of ultra- centrifuging, with reference to “‘rest- ing’’nucleusand mitotic spindle, 188. Chromatophores, blue, in some fresh- water Protozoa, 492. Chromosomes, salivary gland, structure as affected by mechanical distortion, 238. Ciark, E. R., AND ELEANOR L. CLARK. Observations on conditions affecting growth of cells and tissues, from microscopic studies on the living animal (abstract), 405. CLARK, JEAN M. See Mazia and Clark, 306. CLaus, PEARL E. See Guyer and Claus, 462. 548 INDEX Cleavage without nuclei in Arbacia punctulata, 101. Cowes, G. H. A. See Keltch, Clowes and Krahl (abstract), 399. ,—: See Krahl, Clowes and Taylor (abstract), 400. Coates, C. W. See Smith and Coates, 282. Cor, W. R. Environment and sex in the oviparous oyster, Ostrea vir- ginica, 353. 5 Sequence of functional sexual phases in Teredo, 122. CoonFIELD, B. R. Regeneration in Mnemiopsis leidyi Agassiz, 421. CowLes, R. P., anp C. E. BRAMBEL. A study of the environmental condi- tions in a bog pond with special reference to the diurnal vertical distribution of Gonyostomum se- men, 286. DAN: Kartsuma, T. YANAGITA AND M. SucryvaAMA. The behavior of the cell surface during cleavage (ab- stract), 395. DELAMATER, ARLENE JOHNSON. The effects of heavy water upon the fission rate and the life cycle of the ciliate, Uroleptus mobilis, 199. Differentiation, energetics of, 59, 82. Distribution, diurnal vertical, of Gonyo- stomum semen in a bog pond, 286. Dogfish ear, sensory hairs, 429. Dogfish, reactivation by cutting of severed melanophore nerves, 255. puBuy, H. G. Separation of the con- ducting and contractile elements in the retractor muscle of Thyone briareus (abstract), 408. AR, dogfish, sensory hairs of, 429. Egg, membranes and germinal vesi- cle, Sabellaria vulgaris, 46. Elodea cells, free calcium in action of stimulating agents on, 306. Environment and sex in oviparous oys- ter, Ostrea virginica, 353. Environment, conditions of, in a bog pond, and diurnal vertical distribu- tion of Gonyostomum semen, 286. ERIKSSON-QUENSEL, INGA-BRITTA, AND THE SvEDBERG. The molecular weights and pH-stability regions of the hemocyanins, 498. 549 Excretion of inorganic phosphate in sculpin, 360. , role of blood cells, in ascidians, 249. FIGGE, FRANK H. J. The effect of some oxidation-reduction indicator dyes (phenol indophenol) on the eyes and pigmentation of normal and hypophysectomized amphibians (abstract), 393. FisH, CHARLES J. The biology of Oi- thona similis in the Gulf of Maine and Bay of Fundy, 168. FisHerR, K. C., AND J. A. CAMERON. The effect of light on the CO- poisoned embryonic Fundulus heart (abstract), 404. Fission rate, effects of heavy water on, of Uroleptus mobilis, 199. Fox, Denis L. See Young and Fox, 217. FRENcH, C. S. Efficiency of photo- synthesis in purple bacteria (ab- stract), 394. ASTRULATION, effect of alcohol on, in Arbacia, 413. GeorGE, W. C. The rile of blood cells in excretion in ascidians, 249. Germinal vesicle, of egg of Sabellaria vulgaris, 46. GrER, HERSCHEL T. The morphology and behavior of the intracellular bacteroids of roaches, 433. GLassMAN, H.N. See Jacobs, Glassman and Parpart (abstract), 401. Gonyostomum semen, diurnal vertical distribution of, and environmental conditions, 286. GRAFFLIN, ALLAN L. Renal function in marine teleosts. IV. The _ excre- tion of inorganic phosphate in the sculpin, 360. Grave, B. H., AND J. SmirH. Her- maphroditism in Mollusca (ab- stract), 407. Gut, structure and function, in surf perches with reference to carotenoid metabolism, 217. Guyver, M. F., anp PEARL E. CLAUS. Recovery changes in transplanted anterior pituitary cells stratified in the ultra-centrifuge, 462. AIRS, sensory, of dogfish ear, 429. Harting, H. K. See Luckeé, Ricca and Hartline (abstract), 397. 550 ,—.—. The discharge of impulses in the optic nerve fibers of the eye of Pecten irradians (abstract), 403. Harvey, ETHEL Browne. Partheno- genetic merogony or cleavage with- out nuclei in Arbacia punctulata, 101. Harvey, E.N. See Hunter and Harvey (abstract), 396. Heavy water, effects on fission rate of Uroleptus mobilis, 199. Hemsrunn, L. V. Protein lipid bind-— ing in protoplasm, 299. Hemocyanins, molecular weights and pH-stability regions of, 498. HensHaw, P. S. The question of re- covery from X-ray effects in Arbacia sperm (abstract), 399. Horstapius, S. Investigations on de- termination in the early development of Cerebratulus (abstract), 406. HunNNINEN, A. V., AND R. WICHTERMAN. Hyperparasitism: A species of Hex- amita (Protozoa, Mastigophora) found in the reproductive systems of Deropristis inflata (Trematoda) of marine eels (abstract), 409. Hunter, F. R., anp E. N. HARVEY. The effect of lack of oxygen on the permeability of the egg of Arbacia punctulata (abstract), 396. Hunter, Laura N. Some nuclear phe- nomena in the Trichodina (Pro- tozoa, Ciliata, Peritrichida) from Thyone briareus (Holothuroidea) (abstract), 406. RVING, Laurence. Physiological ad- justments to diving in the beaver (abstract), 391. ACOBS, M. H. See Parpart and Jacobs (abstract), 400. ——, —. —., H. N. GLAssMAN AND A. K. PARPART. Further studies on spe- cific physiological properties of erythrocytes (abstract), 401. EIL, Evsa M., ann F. J. M. SICHEL. The injection of aqueous solutions, including acetylcholine, into the isolated muscle fiber (abstract), 402. KELLEY, ROBERT W. See Saxton, Schmeckebier and Kelley, 453. KettcH, ANNA K., G. H. A. CLOWES AND M. E. Kraut. The respiratory INDEX effects exerted by certain organic compounds in relation to their molecular structure (abstract), 399. KINDRED, J. E. An interpretation of the secondary lymphoid nodules in the albino rat (abstract), 390. Kine, R. L. See Beams and King, 188. Koonz, Cart H. Some unusual cyto- logical phenomena in the spermato- genesis of a haploid parthenogenetic hymenopteran, Aenoplex smithii (Packard), 375. Kopac, M. J. Interfacial films between oil and cytoplasm (abstract), 398. Kranu, M. BE. See Keltch, Clowes and Krahl (abstract), 399. ——, —.—., G. H. A. CLowEs anp J. F. Taytor. Action of metabolic stim- ulants and depressants on cell division at varying carbon dioxide tensions (abstract), 400. LACKEY, J. B. Some freshwater Protozoa with blue chromatophores, 492. LipMAN, Harry J. The phosphatase content of the developing chick embryo (abstract), 386. Lucas, ALFRED M. Nerve cells without central processes in the fourth spinal ganglion of the frog (abstract), 387. Lucxé&, B., R. Ricca AND H. K. Hart- LINE. Comparative permeability to water and certain solutes of the egg cells of three marine invertebrates (Arbacia, Cumingia and Chaetop- terus) (abstract), 397. ACDOUGALL, Mary S. Cyto- logical studies of the genus Chilo- donella, III. The conjugation of Chilodonella labiata, variation? (ab- stract), 410. Marine Biological Laboratory, thirty- eighth report, 1. Martin, Ear: A. Asexual reproduction in Dodecaceria fimbriatus (abstract), 396. Maza, DANIEL, AND JEAN M. CLARK. Free calcium in the action of stimu- lating agents on Elodea cells, 306. Melanophore nerves, reactivation by cutting, in dogfish, 255. system, physiology of, in catfish, 259. INDEX Melanosis, cutaneous, in lungfishes, 282. Membranes and germinal vesicle of egg of Sabellaria vulgaris, 46. Merogony, parthenogenetic, or cleavage without nuclei in Arbacia punctu- lata, 101. Metz, C. W. Effects of mechanical distortion on the structure of salivary gland chromosomes, 238. Mitts, K. O. See Cameron and Mills (abstract), 405. Mitotic spindle, components of, 469. , effect of ultracentrifuging on, in chick embryonic cells, 188. Mnemiopsis leidyi, regeneration in, 421. Molecular weights and pH-stability re- gions of the hemocyanins, 498. Mosquito larvae, utilization of solutes by, 343. Mustelus, reactivation by cutting of severed melanophore nerves, 255. ITRITE in the sea, occurrence and significance of, 133. Q!THONA similis, biology of, in Gulf of Maine and Bay of Fundy, 168. Oyster, environment and sex in, 353. Oxygen tension, influence on multiplica- tion of bacteria, 324. ARKER, G.H. The reactivation by cutting of severed melanophore nerves in the dogfish, Mustelus, 255. PARPART, A. K. See Jacobs, Glassman and Parpart (abstract), 401. , —. —. The permeability of the erythrocytes of the ground-hog (abstract), 387. , — —, AND M. H. JaAcoss. Paradoxical osmotic volume changes in erythrocytes (abstract), 400. Parthenogenetic merogony, or cleavage without nuclei, in Arbacia punctu- lata, 101. Phosphate, inorganic, excretion of, in sculpin, 360. pH-stability regions of the hemocyanins, 498. Pituitary, anterior, cells, recovery changes in, after stratification in ultra-centrifuge, 462. Program and abstracts of scientific papers presented at the Marine Biological Laboratory in the summer of 1936, 386. Protein lipid binding in protoplasm, 299. Sol RAKESTRAW, Norris W. The oc- currence and significance of nitrite in the sea, 133. Reactivation by cutting of severed melanophore nerves in dogfish, 255. Recovery changes in transplanted an- terior pituitary cells stratified in the ultra-centrifuge, 462. Regeneration in Mnemiopsis leidyi Agas- siz, 421. Renal function in marine teleosts: ex- cretion of inorganic phosphate in sculpin, 360. Ricca, R. See Lucké, Ricca and Hart- line (abstract), 397. RIETHMILLER, GRACE. See Wolf and Riethmiller (abstract), 391. Roaches, intracellular bacteroids of, 433. RucGuH, Roperts. A quantitative analy- sis of the anterior pituitary-ovula- tion relation in the frog (abstract), 389. GABELLARIA vulgaris, membranes and germinal vesicle of egg of, 46. Salivary gland chromosomes, effects of mechanical distortion on structure of, 238. SANDow, ALEXANDER. Diffraction pat- . terns of striated muscle and sar- comere behavior during contraction (abstract), 392. Sastow, G. Prevention of edema in frog perfusions in the absence of serum proteins (abstract), 404. SaxTon, JOHN A., Mary M. SCHMECKE- BIER AND ROBERT W. KELLEY. Autogenous transplantations of pig- mented and unpigmented ear skin in guinea pigs, 453. SCHECHTER, VICTOR. Comparative hy- potonic cytolysis of several types of invertebrate egg cells and the in- fluence of age (abstract), 410. SCHMECKEBIER, Mary M. See Saxton, Schmeckebier and Kelley, 453. ScumittT, F. O., R. S. BEAR AND J. Z. Younc. Some physical and chem- ical properties of the axis cylinder of the giant axons of the squid, Loligo pealii (abstract), 402. SCHOEPFLE, G., AND J. Z. YouNG. The structure of the eye of Pecten (ab- stract), 403. Sculpin, excretion of inorganic phos- phate, 360. S02 Sex and environment in the oviparous oyster, Ostrea virginica, 353. Sexual phases, functional, sequence of, in Teredo, 122. SICHEL, F. J. M. See Keil and Sichel (abstract), 402. ——}, —. —. —. AND A.C. Burton. A kinetic method of studying surface forces in the egg of Arbacia (ab- stract), 397. SmitH, G. M., anp C. W. Coates. Cutaneous melanosis in lungfishes (Lepidosirenidae), 282. SMITH, J. See Grave and Smith (ab- stract), 407. Solutes, utilization by mosquito larvae, 343. SPEIDEL, C.C. Experiments on the con- tractile substance of muscle fibers (abstract), 401. Spermatogenesis, cytological phenomena of a haploid parthenogenetic hy- menopteran, Aenoplex = smithii (Packard), 375. STEINBACH, H. Burr. Effects of salts on the injury potentials of frog’s muscle (abstract), 388. Stimulating agents, free calcium in action of, on Elodea cells, 306. STUNKARD, H.W. Notes on life cycles of digenetic trematodes (abstract), 411. SuciyamMa, M. See Dan, Yanagita and Sugiyama (abstract), 395. SVEDBERG, THE. See Eriksson-Quensel and Svedberg, 498. AYLOR, J. F. See Krahl, Clowes and Taylor (abstract), 400. Teredo, sequence of functional sexual phases, 122. TRAGER, WILLIAM. The utilization of solutes by mosquito larvae, 343. Transplantations, autogenous, of pig- mented and unpigmented ear skin in guinea pigs, 453. TYLER, ALBERT. On the energetics of differentiation, III, IV, 59, 82. INDEX LTRA-centrifuge, recovery changes in transplanted anterior pituitary cells stratified in, 462. , effect on chick embryonic cells, with reference to “‘resting’”’ nucleus and mitotic spindle, 188. Uroleptus mobilis, effect of heavy water on fission rate and life cycle, 199. ATERMAN, A. J. Inhibition of gastrulation in Arbacia with nickel chloride (abstract), 411. ——, —. —. The effect of alcohol on gastrulation in Arbacia, 413. : The membranes and germinal vesicle of the egg of Sabel- laria vulgaris, 46. WICHTERMAN, R. Division and conju- gation in Nyctotherus cordiformis (Ehr.) Stein (Protozoa, Ciliata) with special reference to the nuclear phenomena (abstract), 412. ——, —. See Hunninen and Wichter- man (abstract), 409. Wo tr, E. ALFRED, AND GRACE RIETH- MILLER. Studies in calcification: III. The shell of the hen’s egg (abstract), 391. YANAGITA, T. See Dan, Yanagita and Sugiyama (abstract), 395. YounG, J. Z. See Schmitt, Bear and Young (abstract), 402. ; See Schoepfle and Young (abstract), 403. ——, RosBert T., AND Denis L. Fox. The structure and function of the gut in surf perches (Embiotocidae) with reference to their carotenoid metabolism, 217. OBELL, CiaupE FE. Observations on the multiplication of bacteria in different volumes of stored sea water and the influence of oxygen tension and solid surfaces, 324. MARINE BIOLOGICAL LABORATORY SUPPLY DEPARTMENT LIVING MARINE MATERIAL For a number of years the Supply Department has been furnishing Living Marine Materials. From experience it has been found that during the period between the first of November and the.end of February these live animals and plants can be shipped with the most success. This service has proven to be of great value to both instructor and student, particularly to those who are located at some distance from the sea coast. : MARINE AQUARIA SETS. The following marine sets will be supplied during’ November, December, January, and February. We assume responsibility for shipments going as far west as the Mississippi and as far south as Georgia, and guarantee to deliver the animals in good condition. Although we have been successful in shipping live Marine Aquaria Sets to New Mexico, Texas, Missouri, North Dakota, and Jowa, we do not guarantee safe delivery except as stated above. Set I. Live Marine Material for five-gallon Aquarium. Includes the following: 5 gallons Sea water 1 small Thyone 2 small Metridium (Sargartia may be substituted) 2 small Mytilus 2 small Starfish 2 Littorina - 1 Hermit Crab Green algae (Ulva when 2 small Sea Urchins available) Rricespensets 53a PME AR Cc Suhr koe) sores cata Racha PEE aces SNe ie ANAT $5.00 Set II. Live Marine Material for ten-gallon Aquarium. Includes the following: 10 gallons Sea water 3 small Sea Urchins Littorina or Urosalpinx 4 small Metridium 2 small Thyone 1 Nereis 2 small Starfish 2 small Mytilus 1 small Limulus 1 Red Starfish (Henricia) Balanus Astrangia Green algae (Ulva when 2 Hermit Crabs 1 Brittle Starfish available) ; EURIGE SPOT SO bri iy nyz ee eo er aR Ne IN a arte AG NCC IN or Oe Rete $10.00 Other marine animals not listed can be supplied on short notice. Substitutions can be made on any of the above sets. Prices on larger sets will be gladly given upon request. All shipping containers furnished at cost. Credit of $1.50 will be given for each carboy returned by prepaid express. ~The sea water for these sets is shipped a day or so in anvenes of the materials, together with instructions for the care of the living forms, in order that the aquarium may be set up and ready for the specimens when they arrive. If there is any further information desired Fegarding the above sets, we are very glad to furnish same. MARINE BIOLOGICAL LABORATORY WOODS HOLE, MASS. CONTENTS WATERMAN, A. J. : : The Effect of Alcohol on Gastrulation in Arbacia........... 413 COONFIELD, B. R. Regeneration in Mnemiopsis leidyi, Agassiz............... 421 BOWEN, R. BS } Sensory Hairs of the Dogfish Ear... ... LAR aa Pe ie CLA 429 GIER, HERSCHEL T. . The Morphology and Behavior of the Intracellular Bacterioids OPiROAGHES. ee eh are NT Ng ad Oe jets oma 433 SAXTON, JOHN A., JR., MARY M. SCHMECKEBIER AND ROBERT W. KELLEY Autogenous Transplantations of Pigmented and Unpigmented Bar Skin in Guinea Pigs... 3+...) ee ay 453 GUYER, M. F., AND PEARL E. CLAUS Recovery Changes in Transplanted Anterior Pituitary Cells Stratified in the Ultra-Centrifuge.......... ERO aE: 462 CAROTHERS, E.' ELEANOR Components of the Mitotic Spindle with Especial Reference to the Chromosomal and Interzonal Fibers in the Acridide.... 469 LACKEY, JAMES B. ‘Some Fresh Water Protozoa with Blue Chromatophores. -. 2. 492 ERIKSSON-QUENSEL, INGA-BRITTA, AND THE SVEDBERG The Molecular Weights and pH-Stability Regions of the EVEMOCYVANINS 15 2h a IG on ec enn ae 498 ¥ ne (OMe 4 | ie Z ] aur ty ie 4 are “eh ae AnARA nwa & CRALAAL fy . . ‘ NAY \ S| ~ AT \ a 2 el la | | \ alae Y | wr mA a ARAAG - A ea eae AAAARAAARAARARAARAA AA Beene ARAB AA Al’ aaRARARA AAR RRAAARARAAA AA AAR AAA \AAAAAAAR AAR ARAAAAAA AA 2 AAAS AAAAA AAAMAARAAAn a TAAY\ \¢ \ | A oO JARRAMARB ARS \] f \ | "| Val - \ | \l U ) o> ' g —~\/ ay H = \ - x y. - } | \ mi f_\ 1 ile We a We a a / TON.EN if . \ q ' “PS! i . y \ \ W \ am Vala lanl, V1 On! | \ ENNIS AS ¥, lanl 1 fn Fan FONT 1 7 , y . ~ » 2 \ Ve gex\ FSS . \ 1 1 | ' A \ Zam \AN\ \ BX NNN! \f A Gaeane y r | Va) | VA AT \ I AA | AAA au ~ RARRAA, pAannnanannenn RARE aa y AIA AR AARAR = \ ’ \ | | AA aan a Anan YY ln nanny A oS [> ‘a YY iN} ~~ an) FAAAAAA Ap AAANAA An AAA 333 MARAARAAAAYY EET EET YY AAAAAAAAAAAA- AARAAA AB ANAARARAAA ARAAA ARAARAR INA AAAAal (Al A AAAAAARYA nae ACARANAANAN |. aaaAAan A | A AL an RARE | A PAAR ARP JAAABAAAAAA A ar A —~ AAA A ANIA agape et ete AAAARRa WAAR AAAR AA AAAaAa r ANNAN A AY ARR aaa AA RAAANNBARn A AA AA eee anatgaasasseossonnncananaen nin AA ) MAMMA AR ARRaZRBRSC ROT AarAaan: A ORARAI AAA ANA ARPARBRARRAAR AC noe ag Aan aa AR aA rw ww 2'A's A \ f } } WF} | Wf r W PAE AAARARARA ARAN A AAT a LAA AAAR ‘ea ‘aaa Ag | ay \ \. a | ) @ | x 4 | Jal pAAaanrans AAAA nan AWAAA NAA AA / “WOON 3 9088 01228 ia