Aa gen ooh bo hong tea de Feat be Oe ‘ende bebe tae dere Abb: Wee aece oe oeerees #4 MH beten dearest pen + th) bade iy setatainds saad Three be deh tipantets toys 1 98 9 » Senesseeriee tow O08 4 det na nee | Obata poke + Om sens ygnes bie bb hoe hy eres reer Uninne artes, ne ateibee aby sstahanaty WO rnugrehbe yebce #9 8 be peas) #y bl ne ot paces bas destece Ronihdirsals Ay rae . rer » ri ah te bene babe ye ee he pees ome . h riences at Pre he bee eweee | ae ” faptrtabuatet resi af clehrnttraoniney heh \ehagedd Ae tebe irasae ere ; : swig ud Ve pbelede athe aaah hens : pty r rerepua eee * pee ; art tren) Rteat ie sertetes rth 4 on bebe td soon isl gatas ode beattes supp hy ews bein: bee SH bee bee oe n bellwaebs be be eel Bie aepe begs pds gS anes 7 ; rt ig Pees Wi ee Mis) Me eke de hr ae he ba: te visheer ithe piskehstel pbabatete: : i bed pba tates tee eke 40: te be My S40 te Ot re hee Shetek aien ers Risthieries he bielee bane here teie Genes! + be eb Roeaiectan St etebe sew cahete ¥ iy eigiease ecard ae sai . 4 Nae a hy ait Cte wet Lpcbur nra! Forbes Bese beasssnetebeiate ite o he baked vibe hens oa dh rte tise ve vee Fae ha by he . Hote We be hohe by Be be bate waa eaatiesaisiehsss ths veg bel “ue sl iche ¢ Weer) i pehe “sweera! ‘ig “hs We dy he te ct rat va isa i ie sal he aaa age Vibe bt be tty bette te) Hepner dere y { a ae) shebpers eae tre ‘tence lesa bs Ms teste) , s bee D Wy ee che babel a ye sah “iin 3 “9 | by leloby ta phe ehehrer Oe hes) Rede! atiet sh seater ENigte*2: - “" nisin Pee t : iahal ari “s teh Ld, ste lgde tine tatee Lytraehaia bey Hy shes + ait i jog aly Eb hy ais “4 * —— '. r " Statist I apes rates ” v4 iq osaeee be reat feito lone ne be ryan + ae repli debehe tet) Signer sensor toes adacetahe ded rae SS L Dabs be Sabet upmierea bedebehhate tint absent iat itiinseintene iam tehe Whee ; ' ah inte wepite tien Saat free siete sspecaatte ’ Nidtene te feneee ay tH iyi ho af ns ort te te bane ba le: bebe ht be ve Sr reH tes he hoe oe he st ieee i yar O84 A te aber orld or elettiy a shi poms Hatt be a) or RY echt) rie ry eam 3 ih mt ties. + tad ” o4 Med, «jit? Lebel aves vty TN Anh bebe Sah ine cM igbes » “* Sibeve > Hitt Sd ie lek + ratiagians Bay * ‘ sched ; ai Sethe Meth at wp My i bert setelatetes, The yt it 8 9dyee bs bs ’ tie Avtar etree. pee be be, see je 8) heey te rele pee Hbylel 9 1964 Hy tedites re be bey age ne pigt rweng hs Hai hiemoves a Sekitthnnectahed pie ighed 986 Hobs be rtobmterrtird ys hed iA. Salet , Sorttte ae C8 CbeE FD miaiasieisie Rite se BLA hh leas abetebethe ket hai Iyenisatpedgea tit etre iol Y : Ae OVR LO be a helet sdthated sabeisbobetattadetasites at; 2p . dn nae ar teen itine wt bs eet. ¢ Lap a) Whe pe oege nae j\erre@ r + . Libak, ealeeded Jaihels be leiede ; a idltgte tel i? ' Peer dee tone tr v0 i an eat be 1 bbe tebe Dil dhisa, fate oy hag» "4 z vate ne fe * x abe ieinesy nse byt ah by pierre i: “ W448 ottdide iste ober s)he we ar drivyo negra te * didial ane 2 rabid eybeteh rept Ait h Lh shane pUsdalabe ebebaneeebate sahasenes rahe vet} iat #9 OF pe Meebetihubet ties i byt Hanwrer > " a vive +e OU Ca Ai he Oe oe ve dt oy ae ihe fe rete tee itera tet detragigcady UGLY , rst peeeee [hikes Hiubsor ne Metpees Labi tet shone pyererret es oboe Mo a ee hore ” abbdehslehad seated whe 30 ( piedebtaed merase eats « rh p—4 beg 848 08 eye PE UEE ELT Ree RLS De Ue phe beds he ie pa. , Hee vie. ay ae et ane ++ *) es wae 94 brats Bold “ pedi. Nee “4 pe rekebrietseete yeh beedyre iaeetattit venonte bs pikgrssteisteiest Lidetetedinl. ices votes © rt 18a 4NHO NS 10 4 aa yore ye ne ot: hare : 4 site telrte py me bene bebensess a et rerererea tener tate ee nied saesrene helahatetpravoetistoie TS) Hh 95-49 bot ie ie bo bate pee rte bebe urtengten yaar vf nigh ote Lphebe hansen SP bhaneeadenananal haben ‘ he iesehieh Pets Y maar oe “ “4 phe Y abet 4 OD) ake Lohede sobs ve ithe ts Us sele rt psbertbebeveuen bs * HY a ne ob 4 ener cielar Seo tiyoeaee oe ehanae she BS ie eeate we 5 es leniremitegeedinains * Tite pratns SddbaMieaalaii. dtmiinad dias dehehed ten i. ie ‘ ’ A steebibAs ah ahdbhived Lokal aceon et lo hy bo br hobo Ohyr be betas ped tod bleh Peek ite Ot butt ae bret re ore gebor aren ae Heetededs gt we " prbtety ~. 6 gobo NE Fhe ie Ortigee at ot et tL eos . erat eS it bern ienind boi beathedthadtn peoneeae it * he ele plphy ANS gy mpginatmas mrt pseseteas ¢ wal % perrisreresetr ty 12} inate abet ome pele: ite hte te ertenes ire Oa peal any & fe oa te ry be be de rabies eee An iY irate Teotede scorers bear yt f Wrsstiserti = Sapptasitt pha tybs bien dAplerechel msde lois hal verb per tr tri reson sta es meet sistent war ieyereesels pees bensee mae anys inbeges s+ F98960b0 rahehs $e beket -preree Y eee geet Se: ataronaicer Satacherd fritter Tete> 6 treet meat deste! ode as nee 4s 4) bebea test koe 1 bq tet a ares ore pebene) . ESE pret rrtes 4 dein be emete se on ve Seiten Neato miberby Leach hI rk, bieriee bahebabetabdedebatsrmrasecen ee nene are fede be at ” rf pare series on 2 i be hae ae Wane: betwee be ws F . naubie et bole Be od: & Ae be b+ Ge bebe 08 ee pte vot Nee t gene aned ee nied ~. 77 art |. io a OU Qe ey Ve Us Y = be ie aT 111 | AE THAT twine WARM ese. Sy TENN an fepete rie. eae Pe Mepis yyte™ tie | ay ‘Sa, SS esky SVS htee | CHT MTT SE nao at Og A and o wes ~ s4 abd or SeyDIe) yewNarnasasee: = ow 5 San” ae = - i ‘ {live b be a 4) as pe) i cn gt ai Ph e Flore " hy i} AFT teers SMA a Ec mnaggallte: Foes —_ oe ° = ss: MS eomeeecercere OO? giey.- al % mid ¥ Tlpidd wee & = ee ! 7 Ha | rd ; “at a) | | L hat} C&S be tat ' age Ny *\eees* veh G4 -s0e-- wove nit Moved - p) bie b bt if 4 Spe) ) . Ck, ; vr cour Vy ry , hd Ae ee Wn Se oe s v" Ave 4. wv Nyy eV jeans " Ane , asad axcam a TEEPE TELE AEE wf Z A fy & FS, he me , if aL ar i] J ieuaeeee Wm ar, Coes 26 ll : ‘ NCR Hy et A na tre mane s, Seine eek ee ey DWN: : nn, Wx on tar NP enaot Y Un Thane ‘he, iw ate ws a Madey si fi Natiatnd Mee A prAdlee i os ? {tT} reeves Phat Tt Ttet do is 4 aggurnmnnitl* {1} 4pd- WH, * ) *% 44 | AA 1 Py aed S Flaws, WAAL at RAYS AAA G9 nttern UG pee teen ene ET hth Tee fee =A wae «.™ Ae . ‘ Lone! x TPEL a w- wd A ‘ar pte sh test ot < 88055 +e re 8 | | of ~~ ) ~.47 ~ veath “Sy SS ‘ ; Ty | = | rar ae Ay Com la why 4 4 can VON +e, A Ns Wi v ee Nw WAY he VAR, s ‘ y Tis #- Ay ‘ PVA a) | ] if ‘ce ‘ | 4 f ay hl AAG Aap a a ‘ «As Ny a ™) | tn j a Ste. SRRIA AY aOR AYN Ge! : AA — AA. | _s We. * om q ~ xe AJ ~ & 1 5 Bees AL || Lanaae a oh, am ae esata’ 20" 4 Mw ening yr "¥~h — 2 tf : bo TTA @ gainer Cnptyey * Van i * is Ly ee ad ATERHEH ULscoveteneppeveeR MT NETL 4 + esae wy! we (ty v a uy ar a | 4s wr 4 se ae a f 4. 1 r * \ % ‘We. Mangus gy » | 1 QP wwe: : aw en eo. bea. st WE \ Ny; 5 ala (Ay us w~ aa | yw eee, nA At dF ~w #5) ‘ a r wy 7 Wy, M "y Wy a inl aif “Hy. ald ha te TINT ALA BA, . 43%@ by yy ny vs \ €\\ AS Ty e vse THE ete ua ne call Tt hhy es Ae -. ? e 3 aK ne Make Sas : r ihe th neva Wily i jt4e@ be Weed iy » a Ui’ SON Le! . ALM, AU View L Cee ue wide hi byes ‘ash wn, ‘ wl Par AAD | A WATT AD OO fet ros at pe re ¢ vs na) . we (ie , We opt” - he Se: oes rea? Oe ath : 4 Wry | a ' Mm » firey TI A. ney, «ey de 4Eleg "thoy 4a wr ad Ue UE hy c x PTL LT Tf ‘te prtang eg Se whet | Ne, Whitey, et “Oya Aud av ta |) ted or CER ay he ey Pel Bate Pat ely | PT TTT he Le ht nll | ae LW) a y : = ae tit Mt bd = Sk i \ To bikes aan Sea, ‘ Py “ i || ‘Lahey -» : VUCUTOCLUS, . “fa. wert 4}. 3 | ~ e ~: On. ) V tgs 88y, am tap ame fos qnhe n My. J ° ee aw i og ee qi is 1 re SHAN TCT LLL seer 1 Ll saphena eters anh } s SAT taster AOU pling tf, iid adh gegra tly th i TEUPTit jotha pede Aang Is i > tbe, ly | uf G fs DEPARTMENT OF MARINE BIOLOGY en SOK OF . RAPIER CARNEGIE INSTITUTION OF WASHINGTON ALFRED G. MAYER, DIRECTOR 2 PAPERS FROM THE TORTUGAS LABORATORY CARNEGIE INSTITUTION OF WASHINGTON VOLUME VI ex WASHINGTON, D. C. _ PUBLISHED BY THE CARNEGIE INSTITUTION OF WASHINGTON . IQI4 . * Lf > {/ GN g i cap dog Le S aw 2 3 AICLb pp ( " ¥. @ DEPARTMENT OF MARINE BIOLOGY 2 Qa A, £06 f OF i : ce, o CARNEGIE INSTITUTION OF WASHINGTON . ALFRED G. MAYER, DIRECTOR © PAPERS FROM THE TORTUGAS LABORATORY OF THE CARNEGIE INSTITUTION OF WASHINGTON VOLUME VI jee. PP cganian \nnil(u) x) a) Ps ‘ 2 6CorY¥] ) yf Cc. y Nations! woseur ae WASHINGTON, Be. PUBLISHED BY THE ‘CARNEGIE INSTITUTION OF IQ14 W ASHINGTON CARNEGIE INSTITUTION OF WASHINGTON PUBLICATION NO. 183 Copies of this Book were first issued i JUL 141914 PRESS OF THE NEW ERA PRINTING COMPANY LANCASTER, PA. XII. CONTENTS. . The Effects of Temperature upon Tropical Marine Animals. By A. G. IMA Tere Niche ial ajery 2 Sia syd Pe wie: ood Sip tania Stine 8 figs., 4 records . The Relation between Degree of Concentration of the Electrolytes of Sea- water and Rate of Nerve-Conduction in Cassiopea. By A. G. WV leasy tees tere Pa oyrator Seva, SHUR Gis las tuetenanchacal es ode sasevete’s ein tae ete 13 figs. . The Law Governing the Loss of Weight in Starving Cassiopea. By A. G. IMiay GmSete toy ccs cata ae heen so hs teetaate ace eae bee I plate, 21 figs. . Changes in Salinity and their Effects upon the Regeneration of Cassiopea xamacnaye, By A: J. Goldlatpea5s oi vedcdonscesc eae 4 figs. . Regeneration in the Annelid Worm, Amphinoma pacifica, after Removal of the Central Nervous System. By A. J. Goldfarb............... . Experimentally Fused Larve of Echinoderms with Special Reference to then Skeletons, © By A. J. Goldfarb... 2. 2 sinc. la cs 15 figs. . Experiments on the Permeability of Cells. By J. F. McClendon.. .3 figs. . The Relation Between the Rate of Penetration of Marine Tissues by Alkali and the Change in Functional Activity Induced by the Alkali. Bye ew LOD PIATGE Uae eras, .cvie a ts cere a.nicleur helene aisle wis I fig. . Physiological Studies on Certain Protozoan Parasites of Diadema setosum. SMI COE Be TACO PIE 555g cats Yayo ox ac, Siw BOO o Kalas Sie Oe ET 4 a wisi See . The Origin of the Electric Tissues of Gymnarchus niloticus. By Ulric Dahl- PROM ayo cstdas site (age als oldie, oc.s salad sleteletenieateleue 9 plates, 9 text-figs. . The Development of the Apyrene Spermatozoa of Strombus bituberculatus. Eee eat RV UMNO, cats 5 cute eu eiacaiys. 0x Ria. ie were im adem ats 7 plates History of the Spotted Eagle Ray, Aétobatus narinari, with a Study of its External Structures. By E. W. Gudger....... 10 plates, 19 figs. PAGE. I-24 25-54 55-82 83-94 95-102 103-121 123-130 131-146 147-157 159-194 195-239 241-323 lll i THe EFFECTS OF TEMPERATURE, UPON TROPICAL MARINE ANIMALS. By ALFRED GOLDSBOROUGH MAYER, Director of Department of Marine Biology of the Carnegie Institution of Washington. Eight figures and four records. THE EFFECTS OF TEMPERATURE UPON TROPICAL MARINE ANIMALS. By ALFRED GOLDSBOROUGH MAYER. CONCLUSIONS. Tropical marine animals commonly live within 5° C. of their temperature of maximum activity and within 10° to 15° C. of their upper death temper- ature. In marine animals of the temperate or arctic regions a considerable range of temperature above or below the normal produces relatively little difference in their activities, but in tropical forms even a few degrees of heat or cold cause a marked depression in movement. In tropical Scypho- medusz this depression of movement appears to augment about as the square of the change in temperature from that of the optimum. Thus it is more depressant than one would expect were it due to a simple chemical reaction. As was shown by Harvey (i911), the rate of nerve-conduction in the subumbrella tissue of Cassiopea varies as the increment of temperature, increasing in a linear ratio as the temperature is raised. Experiments conducted at the Murray Islands, Australian Great Bar- rier Reef, show that those corals which die at temperatures below 36.5° are killed by being buried 11 hours under the mud, but those which resist 37°, and above, are proportionally less sensitive to the smothering effects of mud and may survive being buried for 30 to 40 hours. This suggests that high temperature produces death by causing asphyxiation, the oxygen of the sea-water being insufficient to sustain the increased metabolic activity of the animal. OBSERVATIONS AND EXPERIMENTS. The literature upon the effects of temperature on animal life and activi- ties is both extensive and widely scattered, although Davenport and Castle (1895), in their paper in the “ Archiv fiir Entwicklungsmechanik der Organismen,” Bd. 2, have given a list of all previous papers treating of experiments upon the maximum temperature which can be endured by aquatic organisms; and Bachmetjew (1899), in the ‘‘ Zeitschrift fiir wissen- schaftliche Zoologie,” Bd. 66, presents a complete résumé of researches upon the subject of the temperature of insects in relation to that of their sur- roundings. Also a review of works upon the physiological effects of temper- ature is given by Demoll and Strohl (1909). Probably no single factor is a more effective barrier to the extensive geographical range of marine animals than is that of temperature. This fact has long been recognized and there are many striking examples of its S 4 Papers from the Marine Biological Laboratory at Tortugas. general truth; for example, Professor Hjort, in Murray and Hjort’s ‘‘ Depths of the Ocean”’ (1912, p. 444), concludes that the southern limit of northern boreal species of fishes from the sea-bottom everywhere coincides with the isotherm of 10° C. at a depth of 100 meters. That temperature should be so important a barrier to universal distri- bution is the more remarkable in view of the large number of observations showing how readily animals may become artificially adapted to survive in unaccustomed temperatures. Thus Davenport and Castle (1895) found that tadpoles reared at about 15° C. go into heat-rigor at 40° to 41° C., but if kept at about 24.5° C. for 4 weeks their heat-rigor temperature is raised £0.43; t0.43.8° C. The experiments of Dallinger (1887) are even more remarkable, for he kept Flagellata in a warm atmosphere beginning at 15.6° C., gradually raising the temperature for several years until it became 70° C., the colony still surviving and reproducing normally, although in nature the same species were killed at 23° C. In studying animals in a state of nature, Vernon (1899) found that at Naples the mean death-temperatures of various marine invertebrates ranged from 34° to 42.3° C., but that the death-temperatures of the same species in July and August are from 0.6° to 1.3° C. higher than in March and April, this change being associated with a rise of about 10° in the surface temper- ature of the Mediterranean. Observations of similar purport were made by King (1903) upon the effects of seasonal temperature on the development of toad eggs. It is, therefore, somewhat surprising, when we see how readily certain animals may become acclimated in a short time to abnormally high or low temperatures, that whole races of marine organisms are confined within a narrow range of temperature. For example, the rhizostomous scypho- medusz are restricted to tropical or warm seas, none of them extending into regions beyond the summer isotherm of 16°C. Also, the species of the genus Cyanea are equally well confined to cold seas, being bounded in their range from either pole by the summer isotherm of 20°. On the other hand, Aurellia aurita is eurythermal, ranging from pole to pole and giving rise to many varieties and local races. Among the 440 known species of hydromedusz only one, Solmundella bitentaculata, appears to be eurythermal, living at temperatures ranging from 30° to —1° C. and at depths between 1,500 fathoms and the surface, and even among the widely ranging Siphonophore, Bigelow (1911) is able to name only one doubtfully eurythermal species among the 90 known forms. A study of the temperature reactions of the only known eurythermal scyphomedusa, Aurellia aurita, is of interest, for it shows that the success of this species in adjusting itself to a wide range of temperature is due to a remarkable ability for acclimatization. Thus, the Aurellia at Halifax, Nova Scotia, ceases to pulsate at 29.4° C., at which temperature it is most active at Tortugas, Florida; and conversely, the Tortugas meduse cease to pulsate Effects of Temperature on Tropical Marine Animals. 5 at about 7.75° C., while those from Nova Scotia continue to move at a temperature of —1.4° C. with ice floating in the water above them. If frozen solidly into the ice the Tortugas medusz are killed, but (as was observed by Romanes, 1885), our northern Aurellia withstands this treat ment without serious injury. Aurellia aurita is more abundant in cold than in warm seas, however, perhaps on account of the more abundant zodplankton and food-supply in cold waters; yet it may be due to a better adjustment to environmental temperature in the meduse of the cold waters, for at Halifax the surface- temperature in summer is about 14° C. and the range in activity of the medusa is then about 15° C., both above and below this temperature, the animals being most active at from 18° to 23° C. On the other hand, at Tortugas the average surface-temperature in summer is about 20° C., being practically identical with that of the animal's greatest activity; but a rise of 8° C. would cause all movement to cease and is generally fatal to the medusz, while the same species at Tortugas can still pulsate even at a temperature of 21° below that of the ocean in which they live. It appears probable that Auwrellia is a boreal or arctic animal which has wandered into the tropics and become fairly well acclimated, although living in these warm waters within 9° C. of its death-temperature. Thus, in comparison with their northern relatives, these Aurellias in the tropics are poorly adjusted to their temperature environment, and a change of even 2° or 3° above their normal temperature causes a decided lassitude and loss of rate in pulsation. We see, therefore, that the medusa has become adapted to a tropical environ- ment, but this has been accomplished at the expense of its factor of safe adjustment to temperature changes. Harvey (1911, p. 34) showed that Cassiopea, which is strictly confined to the tropics, can not withstand a higher degree of heat than can Limulus at Woods Hole, Massachusetts. Thus, Cassiopea is an animal living constantly within 15° C. of its death-point, yet it is not adapted to with- stand higher temperatures than the heart of a northern animal (Limulus) living at 25° to 30° C. from its death-point. In general, one finds that tropical marine animals normally live at tem- peratures much nearer their death-point than do northern forms. For example, the most resistant reef coral, Siderastrea radians, which lives at about 28° to 30° C., is killed at 38.5° C., and the West Indian reef Madre- pores are killed at about 35.8°. The tropical sea-urchins, Diadema (Centre- chinus) setosum and Toxopneustes (Lytechinus) variegatus, and the brittle- star, Ophioderma appressa, are killed at 37.4° to 37.7°._ In these forms the temperature of the sea-water in summer is only from 7° to 10° C. below the death-temperature. By way of contrast, we may cite the cases of the arctic forms Cyanea arctica and Beroé cucumis, which are killed at about 27° and 30° C., respectively, and which live in an ocean not warmer than 14° C., thus being at least 13° to 16° C. from their death-points. This low factor of safety in tropical marine animals may at times become 6 Papers from the Marine Biological Laboratory at Tortugas. of biological significance. Thus on July 21-22, 1911, at Tortugas, Florida, after several hot, calm days, the shallow water over Bird Key Reef rose to 33° to 38° C. and Dr. L. R. Cary observed that large numbers of Diadema, Octopus, Fissurella, and other mollusks and small fishes were killed in considerable numbers over extensive areas, and corals were injured even when not exposed to the air. In order that a marine animal may live throughout the year in the shallow or surface waters of the tropics, it must be capable of surviving at 29° C. Similarly, the animals of the Arctic Ocean must survive at 0° and, indeed, Cyanea arctica continues to pulsate even when half its bell is frozen into the ice, and after being embedded solidly for several hours it revives at once, apparently uninjured, when the ice has melted. It is easy to see why such forms as the rhizostomous scyphomedusa Cassiopea and the sea-urchin Diadema setosum are confined to the tropics, for they lose all power of movement at from 10° to 12°C. Also the floating barnacles, Lepas fascicularis, can not survive in arctic waters, for they are unable to move if cooled to 4.6° C., and, conversely, Cyanea arctica can not enter the tropics, for it dies at 27° C. On the other hand, even the tropical Limulus polyphemus, from the Marquesas, Florida, survives being frozen into the ice, and near the northern limit of its range, off the northern coast of Massachusetts, it continues to move until heated to at least 40°. Judging from its marked adaptability to extremes of temperatures, one would expect this animal to be of world- wide distribution, yet it ranges only from Maine to Yucatan and is unknown from European coasts. When a boreal or arctic animal such as Aurellia becomes acclimated to the tropics, its upper death-temperature is raised, and conversely it becomes unable to withstand a degree of cold in which its northern relatives may thrive. Its optimum temperature is, however, raised even more con- spicuously than its death-points and thus its factor of safety against ab- normally high temperatures is reduced. These facts are illustrated in fig. 8, page 18, wherein the ordinates of the curve ABC represent the average rates of pulsation of Auwrellia at Halifax, Nova Scotia, the abscissas representing the respective temperatures in degrees centigrade. The curve DEF shows the same factors for the A urellia from Tortugas, Florida, and it is evident that at Halifax a wide range of temperature, from 2° to 19° C., produces but little change in the rate of the medusa, whereas at Tortug a the medusa lives constantly at the temper- ature of its maximum activity and any further elevation of temperature causes a marked decline in its rate of pulsation. The case of the brittle-stars Ophioderma brevispina and O. appressa is unlike that of Awrellia, for at Montego Bay, Jamaica, they both have the same temperature-range, their movements ceasing at 7° to 8° C. in cooled and 37.3° to 38° C. in heated sea-water; yet the former species ranges from Woods Hole, Massachusetts, to the West Indies, and the latter is restricted to the tropics. Effects of Temperature on Tropical Marine Animals. y An even more striking case is that of the tropical reef-flat coral Sider- astrea radians, which, although it normally lives in water of 29° to 31° C., can survive without apparent injury after being placed for 8 hours in water at 6.7° C., and one specimen survived with some maceration of tissue a temperature of 1.9° C. It is difficult to see why this coral does not invade the waters of the temperate regions, but it is unknown north of Bermuda. The development of the ability to withstand higher or lower temperatures is associated with physical changes in the protoplasm. For example, Dallinger (1887), Davenport and Castle (1895), Greely (1901), McGill (1908), Rautmann (1909), etc., observed that organisms when heated excrete water and thus their albumens become denser and the temperature of coagulation is raised. A decided congelation of the slime is observed in Casstopea which has been cooled to its non-active limit. In the tropical scyphomedusa Cassiopea xamachana, however, move- ments cease long before heat-rigor develops, and in fact the loss of move- ment is due to a general muscular relaxation, which develops several degrees below the temperature at which the nerves cease to transmit the pulsation stimulus. If, for example, we cut a ring (fig. 1) of subumbrella tissue of Cassiopea, leaving a radial strip projecting from the outer side and then start a pulsa- tion-wave in the ring, every time the wave B passes the point B a portion of the wave is diverted and passes out to the end of the strip BC. If, then, we place the strip BC and the ring in separate dishes of sea-water, and warm Fic. I. the ring while the strip remains at a normal temperature, the amplitude of pulsation is reduced to insensibility in the ring at about 38.5° C., and all the muscles are relaxed while the nervous im- pulse which produces pulsation still causes the strip BC to respond vigor- ously every time it passes the point B up to about 41.7° C. Thus the muscles are more profoundly affected by the heat than are the nerves. There is another reason for the belief that heat-rigor, or at least coagu- lation of albumens, is not the cause of death at high temperatures, for practically no marine animals can withstand 46° C., most of them dying below 40°, while the most readily coagulated emulsion of egg albumen does not congeal below 56° C.! Snyder (1907, Archiv fiir Anat. und Physiol., p. 113), basing his con- clusions upon Nicolai’s observations of 1901, came to the conclusion that as many physiological reactions have the same temperature coefficient as have chemical reactions, the underlying cause of the physiological reactions must be of a chemical rather than a physical nature. In 1908 Snyder published an extensive and important paper upon this subject in which he made use of the results of all previous studies bearing upon the case; and he concludes 1As a result of our recent experiments upon the corals of the great Barrier Reef of Queensland it appears that high temperature causes death through asphyxiation, the oxygen of the sea-water becoming insufficient to support the increased metabolism. 8 Papers from the Marine Biological Laboratory at Tortugas. that the temperature coefficients of all physiological reactions in which metabolism occurs are of the magnitude of the coefficients for chemical reactions, being about 2 or 3. The equation for chemical reactions at increasing temperatures may be a in the form y = ac*, where y is the rate of the reaction at any 10° C. temperature interval, a is the rate of reaction at the lowest temper- ature in the series; c is the temperature coefficient, the value of which is usually between 2 and 3; and x is the number of the 10° C. temperature intervals. Thus, if the temperature be raised 10° C., then x = 1; if the temperature be raised 30° C. above the lowest temperature in the series, then x = 3. Hence, x is the increment of increase in temperature expressed in terms of units of 10° C. Harvey (1911) tested this subject for the rate of nerve-conduction in Cassiopea by experimenting upon two subumbrella rings which were deprived of rhopalia and set into pulsation by means of an entrapped circuit-wave. In being heated from 18° to about 30° C. the rate of pulsation increased in arithmetical proportion to the increment of temperature, in other words, as a rectilinear function of the increase in temperature, not as an exponential function as demanded by Snyder’s hypothesis. As Harvey was able to obtain only one wholly satisfactory series of observations, it seemed desirable to repeat his experiments. This has been done upon five separate subumbrella rings of Cassiopea, the results being presented in table 1, which gives the rates of pulsation of five rings of sub- umbrella tissue of Cassiopea xamachana warmed in sea-water, at Tortugas, Florida. TABLE I. Number of times per minute the wave Number of times per minute the wave MAS goatee traveled around the ring. ae ae traveled around the ring. sea-water. sea-water. RingA.1/Ring B.|Ring C.|Ring D.|Ring E. RingA.1| Ring B./Ring C.|Ring D.|Ring E. AS AG 17.5 Ly, a PSO OCCA (CES onl orice BIS Oy)> viing.6 sneol lan OOCoe mors Wseosoonlecananc 20.2 ON Mg Biarerstal| bercanate cil covers erovell ieiciet ate fais 35.1 TAO! "| tere) cvetetel|(eveletctevaceil ete staters’eil(alntotel state 21.6 FAO nal alle craustatoiniliocceisteletallveiereeucrers wicgacteie Sia Diet ri cfatalatararel| biatevcjetetellcreleistarers PSO leniiclelere 24.5 eH sesgacnollAoooooulooconsdlEeaodGs Sede Uulicteleveiete si] stevetare ts /ollateldietetatalliay>lovelerare 164 26.6 nobis lagoodas|saucosolloans scdllarcooe oo AACE Wesooocallssooocallacaccod TOO. lerejeieteies 27 TOS | cher ciaicret all ore etotore elllate cele tererpllis erase eters SO lagooeeolocope no LOSe5) | aleve tee eel cteverotarate DB AM ereretie, stall shejerenereeiliciaerstetals ie ie Wel Gecrnc elorey, © \ls@orooc TZOiw | isieicravsieve| niajevatevetell iatefetebas B7eGe wallaterterecte TOO Mi\ ere tata ete ailteta etanatetel| fatove evetave 36.75 bie ellen oeecolonocoon lpoonoculrocoacc 28.9 mie Nosecnoalooopbacilodocogclassacod ZO. Geel leisvelelezerall a alelederener ayevaumtarate rey Gob ccco ZOW | Misdcatiec'| ateletets stele eiereeiel| sraeweree 140 D7ILOZ ys Mil Crete eielerellicverete clots TOS pra tevelererallieterata eters 2OiSie liveyereeteys| aieteveleute by ee i etacicchoey (Cea e ere SISOS) (ll cratatareravel| fois ekevanaell eustsveneletal ll skeiehereusts 154.5 30.9 DZS il /steievake evel] lo varclorereteltevere sioner ovmeiciete Sif eA ml ctetictetesere ey NS anane, TSO\ | inereeretace Spee = lisecabed tao anol loonmccs WEES G5 anbdc STO) Miiliateve micvese Meveierelauete EZA Sloe ce ellsleleteielats See liad Sacre Gaaance OTe Sy inus| /eteyetctcte:| (orators overs S77 Mil tate shies 1670 NEG Goppdl aopce bollonaciooo SOOM lle. dinin ceil cvoloreve ciel shareretaite ESS etciscrsec 37-9 FG/7/ | overetete ciel] (niet etatet= elle nln inus)ece 150 32.95 ESL crore are arate: wreterell enareteterera\| steve terete Bara” | Sister rete ESS i laterciaterciail stotetete avail efatetatenste Bio IRROICCICIO are arnIce TOBZiG) Wave ois, oveie|'avexe wheters 38.3 TOGie Wi Oe ob Wevese tate ateiliete's) ac ofc |(cyeseiontale BALTD. | Seeiie cle We wterercrerel| eretorrere's 7 0 i le ee | SSB o PP iiclersrerearelliesteraseiare Te ocdoe 143 SAMS) sonore is eieiell iolete) sia ie EOS © beietsecisalistaare ce | Eien SSE soc Moocedollpodo.coalaregoge 5 S4ea et ercterat erate E2G" Alehs tye tes alters a erste o dteveretere 39 DSI accrdac occillc derelete clare seve /elal| a elerwtetels 1 Ring A was first cooled from 27.3° to 17.5°, its rate becoming lower as follows: 27.3° C., 70 per minute; 26°, 68 per minute; 24.5°, 72 per minute; 22°, 67.5 per minute; 20.3°, 64 per minute; T7357) GAs per minute. (See Record 1A.) 2 This ring was finally heated to 38.75° and then upon being cooled to 28.5° it gave 89 pulsations per min- ute, being somewhat weaker than the normal. 3 Upon being cooled to 30.5° this ring gave 126 turns per minute. Pulsating normally. 4 Upon being cooled to 26.1° this ring gave 116 sper minute. Fully recovered. 5 Upon cooling to 29.05° the rate became 120.5°. a Effects of Temperature on Tropical Marine Animals. 9 The most successful experiment was upon Ring A, the wave within which traveled around the ring 70 times per minute at 27.3° C., and gradually declined to 54.5 at 17.5° C., from which temperature it was slowly heated to 39° C., its rate then becoming 173. Then, upon cooling the ring to 26.1° C., it pulsated in a normal manner at the rate of 116 per minute, showing no ill effects from the experiment (see Record 14). The obser- WAUAUACAWAUAUAUAUAWAUAUAUAUAUAUAUAUAUAUAUAUAUAUAUAUAUAUAUAUAUAU AUC aU aU a UaUaUaUaUaUatataUAUAVAVAVAVAVAUAN 26C. WVYYYV VV AAAI ANIA VAYOOWOOWOWIOWOWOWOWOWOWOWOWOWOOOWOOIOOIOIWOOOIOIYIONONNONNOWOONIONONNONNONONONINONONIONINN ei ‘ SRL NI ENS ANS NS ONS NI ONS NLA NL NS SNS ND NS NG SSN Nh NGOS 20°3 PDP PPP LLP PLL LLP SLL LLLP PLD LAL LLLP LAL PPP LIISA 26°1C.. RecorD No. IA. 10 Papers from the Marine Biological Laboratory at Tortugas. PRD PDD IP PDPPSPPPPPPPPPPD PIPPI IDI IAAI IAIN ‘27°8C 34°.3 WN NUN NUNN UNO NUNUNUAU NUN UN UDUDU)UDU DUA UAULVEULULU AV EORULULULUDUDULUNW)ULUAULUIUIVINLNLUL VLU) LUAU) Ul OEOE VEVIVAVL VEN LULU LSLULU ELV VEO LOL VL COSI] VICAVESLVLN] VLViN EU ONY oem OEvET CEE) J a BaP RSIS DIMA ERNIE AVAVAVAVAULUAUAUAURUAU ACL UAUAUAUAU AU}! 30. Recorp No. 1 B. vations are plotted in figure 2, and it appears that the increase in rate is in the form of a straight line from 17.5° C. to 35° C. In other words, it is arithmetically proportional to the increase in temperature. 170 ¢ 160 Rate per Minute 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 Degrees Centigrade Fic. 2, illustrating Record No. 1 A. ~—_ =r 7 Effects of Temperature on Tropical Marine Animals. II This result is confirmed by taking the average of all five rings; for here also the increase in rate is in the form of a straight line practically coinciding with that displayed by Ring A. This averagecurve is shown in figure 3. The 200 Rate per Minute 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 87 38 39 Degrees Centigrade Fic. 3, illustrating Table I. equation for the increase in rate with elevating temperature between 17.5° to 35° C., in these rings, is y = 4.85x + 54.5 and for Harvey’s ring it is y = 6.66x + 50; where y is the rate at any given temperature above 17.5°, but below 30° in Harvey’s and 35° on my curves. «x is the temperature- increment. Thusat25.5°C.x* = (25.5 — 17.5) =8, and at 27.5° C. x = 10, etc. The constants 54.5 and 50 are the rates of the rings at 17.5° C. Above 33° C., Harvey observed a sudden decline in rate, but my rings did not display this decline until they had been heated to 35.5° to 39°, and Ring A did not display it even at 39° C. As Harvey states, these curves bear a close, although possibly only a superficial, resemblance to those of enzyme actions. Certainly our observations are not in accord with the conditions necessi- tated were the rate of the nervous impulse a simple chemical reaction, for in this case the curve would be given by an equation of the form y = ac?, 12 Papers from the Marine Biological Laboratory at Tortugas. where a is the rate of 17.5° C., and c is the temperature coefficient. The actual equation is, however, of the form y = a + x(¢ + 1). 27.3C 32°72S°¢ INNOONNOOAAAOAAAAANOOAOAAAANNOO OAR AAAAKAN 3415 C AAA AANA AR | 35°07C 36.05 C 37026. 37°65 C mann SUI 365 Recorp No. 1c. In this connection it is interesting to observe that Knowlton and Starling (1912) find that in the excised hearts of dogs and cats the increase in the rate of pulsation between 24° and 4o° C. is in a straight line, the increase in rate being arithmetically proportional to the increment of temperature. They also comment upon the inadequacy of Snyder’s hypothesis to meet the conditions of their determinations, and this is the more remarkable because many physiological reactions do unquestionably follow Snyder’s law. A relation which may possibly be of some significance in this connection is that the kinetic energy of the sea-water increases, as does the absolute temperature, the increment of increase being arithmetically proportional to the temperature increment, thus increasing in a simple linear ratio, as does the rate of pulsation. The kinetic energy increases, however, at a much lower rate than does the rate of nerve-conduction. Thus, between 17.5° and 35° C. the rate of nerve-conduction has increased in an arith- metical ratio from 54.5 to about 140, or to 2.568 times its initial rate. The kinetic energy has in the same temperature interval increased from K to 29304235 273s ag-5 ’ or from I to 1.06. Thus the rate of nerve-conduction has increased from Effects of Temperature on Tropical Marine Animals. 13 54.5 to 140, or 85.5 per minute, the kinetic energy in the same interval having increased by 0.06. The equation for the rate of nerve-conduction at any temperature between 17.5° C. and 35° C. may therefore be expressed by 85. : y = 55) (aE Tt sas 0.06 273 + 17.5 or % 290.5 + =) mi | y = 1425 (oe I] + 54.5 or, adopting general terms, T+ y= 1425| (7-*) - 1 | +545. Wherein 7; is the absolute temperature of the lowest point of the series and x is the temperature increment corresponding to the rate y, as has been explained on page II. If we let I represent the rate of nerve-conduction and also the kinetic energy of the solution at the lowest temperature of the series, then: The increase inrate _ 2.568 — I The increase in energy 1.06 — I = 26.13. Thus the increment of increase in rate of nerve-conduction, for any elevation in temperature between 17.5° and 35° C., is 26.13 times the incre- ment of increase in kinetic energy of the solution. Hitherto we have been considering the rate at which the nerve-net conducts the stimulus which produces muscular contraction. In nature the stimuli which produce the nerve-impulse arise in the motor centers or marginal sense-organs of the scyphomedusa and, as explained by Mayer, 1908, p. 119, the pulsation-waves annul themselves after each systole; thus the rate of pulsation in the normal medusa is much lower than the rate of nerve-conduction, being, indeed, determined solely by the rate at which the motor centers can generate successive stimuli. In the paralyzed ring, on the other hand, the rate of pulsation is simply that of nerve-conduction, for no new stimuli are generated. The effect of temperature upon the perfect medusa is therefore its effect upon the stimulus-generating ability of the motor centers—in other words upon the rate of the metabolic process which produces stimulation. We have, then, in Cassiopea a unique opportunity for differentiating between the rates of production and that of the conduction of nervous impulses. Cassiopea xamachana with motor centers intact pulsates at its maximum rate at about 33° C., ceasing to pulsate if cooled to 16.6° or if heated to 38.5°. 14 Papers from the Marine Biological Laboratory at Tortugas. Table 2 gives the mean rates of pulsation derived from the study of 7 separate meduse (with marginal sense-organs intact) at various tempera- tures, the rates stated in the table being the arithmetical mean of the observed rates. Thus if 3 meduse# each pulsated at the rate of 3 per minute, and if 4 others each pulsated at the rate of Io per minute, the average rate of pulsation would be 7. TABLE 2. Tempera- | No. of pulsa-|| Tempera- |No. of pul-|} Tempera- |No. of pul- ture of tions per ture of sations per ture of sations per sea-water. minute. sea-water. minute. sea-water. minute. Ge mie 2Gy 16.6 (0) 24 32 31 17 2.5 25 33 33-9 18 5-5 26 23-4 34 33.25 19 9.6 27 25.3 35 30.25 20 ne 7 28 28.2 36 32.4 2I 15.8 29 30.8 37 23.06 22 17.4 30 31 38 13.8 23 18.2 31 30.9 38.5 5.25 Rate per Minute 0 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 338 34 35 36 37 38 Degrees Centigrade Fic. 4, illustrating Table 2. The observations recorded in table 2 are plotted in figure 4, wherein it appears that the rate at which the motor-centers engender stimuli declines if the medusa be cooled or heated from 33°. In both heating or cooling the loss of rate is slight at first, but constantly augments as we continue to cool or heat the animal. In fact, the loss of rate is about proportional to Effects of Temperature on Tropical Marine Animals. 15 \ IN [RUNG D/NE TNE _IN_ IRN Fee aS a INGE INS INGE ING PLN YR PROP RD 27.5 C, gO GR NO NO NY NEE NS EE NE NO NEI NS EET PEI CCIE, SU Som) cei collage LENG Mo 25.) : Recorp No. 2.—A Cassiopea xamachana cooled from 29.5° to 16.8° C. and then warmed to 30.9° C. the square of the departure of the temperature from its optimum of 33°. Thus a change in temperature above or below the optimum appears to reduce the rate of pulsation in proportion to the square of the change. For example, if cooling 1° C. produces a loss of rate of I per minute, a cooling of 2° from the optimum will produce a loss of 4, and for 3° the loss will be about 9, etc. Thus if x be the departure of temperature from the optimum of 33° (for example at 30° or at 36°, x = 3), and if y be the loss in rate (for example, if the rate has declined from 34 to 30, y = 4), then in cooling y = 0.122x*, and in heating y = 0.926x?. It seems that the loss of rate in heating is nearly 8 times as rapid as in cooling. Moreover, when the medusa has lost all movement through being heated, its muscles are relaxed and recovery rarely occurs, even if the medusa be then cooled to a normal temperature. If, however, the medusa be cooled to a standstill, the muscles still exhibit a decided tonus, some shrinkage occurs, and the slime congeals into a gelatinous mass. The falling-off in rate that is observed when the medusa is either cooled or heated from its optimum temperature appears not to be a simple chemical reaction, for it is more rapid than can be represented by the equation 16 Papers from the Marine Biological Laboratory at Tortugas. y = ac*, but in all these experiments the equation y = bx? fits the observed facts fairly well, but this formula is merely tentative and should not be taken as being more than an empirical attempt to express the law. If this be the law, however, it would mean that the decline in rate varies as the square of the degree of heating or cooling above or below the optimum, and it leads one to suspect either that an inhibiting enzyme is formed in a constantly accelerated ratio or a stimulating one is destroyed in the same manner. This is, however, a mere suggestion, for the observations are too few to establish it as a law. Experiments upon Cassiopea frondosa and Aurellia aurita at Tortugas appear to lend some support to the results attained from a study of C. xamachana. Thus table 3 shows the average rates of four meduse of Cassiopea frondosa at Tortugas, Florida, cooled from 28.5° to 12.5° C., and figure 5 gives a graphic representation of the results. It is evident that the rate declines somewhat more slowly than if it followed the equation y = 0.0507x?. TABLE 3. 3 115: Tempera- | Pulsations per ture. minute. 28.5 | 13 II Rate per Minute O O O O ° O O © O O ‘eo Oo O O © O—O 0 1S 4s 5 16) 17 18) 19) 20) 21 22) 423) 24 25 26) 3272s Degrees Centigrade Fia. 5, illustrating Table 3. Upon being heated, Cassiopea frondosa begins to lose in rate at 32.45° and all movement ceases at 38.3° C. Unfortunately, I did not make enough observations to determine the curve of the decline, but if it be a parabola its equation must be y = 1.05x*, and thus the decline in rate upon heating above the optimum is about 20 times as rapid as the decline on cooling below the optimum. Table 4 and figure 6 show the decline in rate of an Aurellia aurita from Tortugas, Florida, cooled from 28.8° to 11.8° C. ~~ ne ee ot a Effects of Temperature on Tropical Marine Animals. 17 TABLE 4. Temperature of| Pulsations per ||Temperature of| Pulsations per sea-water. minute. sea-water. minute. Nt Ore “Ge 28.8 25 14.6 12 20.5 18 13.2 8.5 18.1 17 13 I 16.4 14 11.8 oO 2 é s 4 z 3 g 3 a a 3 Ss [=] 12 13 14 15 16,17 18 19 2 21 22 23 2 2 2% 27 2 29 30 27 2 29 80 31 32 33 34 35 36 37 Degrees Centigrade Degrees Centigrade Fic. 6, illustrating Table 4. Fic. 7, illustrating Table 5. TABLE 5.—An Aurellia aurita at Tortugas, Florida, warmed from 27.3° to 36.4° C. Temperature | Pulsations per || Temperature | Pulsations per of sea-water. minute. of sea-water. minute. SiC aC. 27.3 29 34.1 19 28.9 29 30.15 5 | 30 28 36.4 °o 32.8 24 The parabola y = 0.0865x? fits the observed facts fairly well. Similarly, table 5 and figure 7 show the decline in rate displayed by a specimen of Aurellia aurita, from Tortugas, which was heated from 27.3° to 36.4° C., and the actual conditions are fairly well represented by the curve y = 0.515x’, showing that in heating the rate declines about 6 times as 2 18 Papers from the Marine Biological Laboratory at Tortugas. rapidly as it does on cooling; the case being in general similar to that of Cassiopea xamachana, where the rate of decline upon heating is about 8 times as rapid as in cooling. Hitherto we have considered only tropical forms or animals living in the tropics, and in these we see that a definite temperature of maximum activity is well shown in their temperature reactions. Thus for C. xama- chana the optimum is about 33° C., for C. frondosa 28.5° to 32.5°, for Aurellia aurita 29°, for the movements of the branchial arms of Lepas fascicularis it is about 32° C., and for the reef corals it ranges between 30° and 35.7° C. Thus the optimum temperature is very close to the usual temperature of the sea-water itself, this being about 28° to 31° in summer. Even a slight degree of cooling or heating beyond the optimum causes a decided falling- off in rate. A very different picture is presented by the arctic scyphomedusa Cyanea arctica, from Halifax, Nova Scotia, in September, for here the optimum Rate per Minute ee se ere Oe hi Akane Oi ae ar oe oO Degrees Centigrade Fic. 8. temperature ranges from about 2° to 21° C., the animal’s rate of pulsation increasing only slightly as the temperature rises, and with an ill-defined maximum at about 19° C. Upon cooling below 2° C., the medusa pulsates until the ice imprisons it, although its rate declines rapidly. Similarly, if we heat it above 21° C. the rate declines at about the same rate as it does upon cooling from 2° C. All movements cease at about 27° C. The same general conditions are also shown by Aurellia aurita from the cold waters of Nova Scotia, in marked contrast with its behavior at Tor- tugas. For example, at Halifax, Nova Scotia, in September, when the harbor water is about 14° C., the optimum temperature for Aurellia is anywhere from 2° to 18° C. and it ceases to pulsate at —1.4° and at 29.4° C. Effects of Temperature on Tropical Marine Animals, 19 At Tortugas, on the other hand, a very slight departure in temperature either above or below 29° C. causes a decided falling-off in rate. The contrast between the behavior of Auwrellia at Tortugas and at Halifax is shown in figure 8 wherein the ordinates represent average rates of pulsation and the abscisse temperatures, the curve ABC applying to the meduse from Halifax, while FDE shows the general reactions of the same species from the warm waters of Tortugas, Florida. The wide range in temperature in which these northern forms are about normal or only slightly increasing in activity accords with the fact that tropical marine animals are, physiologically speaking, but poorly adjusted to their temperature environment in comparison with creatures of colder waters, for the tropical forms live near their upper death-points and slight elevation in temperature affects them adversely. Nor are they capable of withstanding the same proportionate degree of cold as do northern forms, as may be seen after every severe ‘“‘norther’’ in winter, when the Florida beaches are strewn with reef fishes which have succumbed to the cold, although the temperature of the water over the reefs in these storms rarely declines below 17° C. TABLE 6. Lowest Temperature Highest Tempera-| tempera- Temperature at which tempera- Temperature ture at ture at which ten- tentacles ture at which Name of coral. which survived | tacles were |retracted and| Survived death death without most fully movements without occurs occurs. apparent expanded. ceased. apparent % injury. injury. °C, bcos Ges cG: GC. Siderastrea radians!....... 4.5 1.9 to 6.7 | 32.5 to 35.7 35-7 38.2 38.5 Siderastreea siderea?....... 5.1 Diese |Wfevekerensoevenceetere, oll tomtronerereee reel cree 37.8 38.3 Meeandra:-cereolatas: cis ccacce cus eine oes 10.2 30) to; 34.2) ||| 31.8 to 35.2 35.9 36.7 Meeandra Clivosats/.. sic cislloms 2 eerste 15.3 31.3 34.4 37.7 38.2 WietrCinian PUT OSA iP he eearet all lope ceca © eitheeey Seeavevo tks oil Covcisystete\atnexel exe tt atereporardvaicarpueiess 36.1 37.8 Porites urea tart oasis crate vam ress 3 ‘hrs, at 32.65 36.4 38.2 37.2 to 38.3 Porites clayvaria Genco) .cs'1es)sereresoeisee ee 10.2 32.8 34.5 to 34.9 36.7 37.7 Porites astreoides’........ Ba Llama ree ae 33-7 34.3 36.7 27.7 Bawick Sika pia taa Saws octets aeoeietl ls: aveteiscedarare|l ove ie aia crsrmreielltis'@ aus feaconumteneye 34.7 36.1 Bizz Orbicella annularis®........ 14.1 VOM Oo raistersiemis trcruie ella tsistarscne siemens 36.2 36.8 Orbicella cavernosa........ ELD OVC Ie Oeil oa Mee ret ec teayiel| ray ai rena evcloyel svete lteher otesoreten oie de tal te ontdae Grehererake below 37.3 PN PATICIAS LS eadterenetd oes inl aN one esi | chains etl Greate nolo aue allttra wrmemre ee Gounll, fereenee en 35.95 Missad(isophelem) (dinsaceak |v tccoat cy sraralllsvekesevs orca acall crye chetescaara revered eee te eee eee eee 36.7 WIUSSAR (SOD YHA ACIGA se |e seeyenssa eo |arercicieesaccin cl rie arate ea ieee | noitonine coe oul eal ambees 35.95 Ocala diiisas .cscias etl sree eercsneys TAS OF cwrulllicate aneuct hee) ec ahote lia nesta skooeeaueuaeens 36.1 36.4 Madrepora (Acropora) muri- Catala erin cy Walemaisieteras acta eraletaie oiccrde 14.1 33-75 34.7 36.8 35.8 to 37 Madrepora (Acropora) pal- SINAC a aro so) vreallsley -etsievsler crete llakewete cei serene llaraveSieisereis te 33-75 34.7 35.7 35.8 Eusmilia aspera (knorri)... LT Lien |\atalccateis ores ayes (netetoverarn otacrcccoket lace tae cue crores 36.1 36.4 1 Survives without apparent injury after being for 9 hours at 11.2° to 12.1° 2 Survives 13° to 14.4° C.for 9 hours; half killed by I1.2° to 12.1° for 9 hours. 3 Apparently uninjured by 13.2-15.1° for 10.5 hours. 4 Killed by 13° to 14.4° for 9 hours. 5 Nearly killed by 13.2° to 15.1° for 10.5 hours. 6 Not injured by 13.85° to 16.8° for 9 hours. 7 Killed by 14° to 15° for 9 hours. 8 Injured but survived 13.85° to 16.8° for 9 hours. § Killed by 14° to 15.8° for 9 hours. 10 Killed by 13.3° to 15° C. for 9 hours. The reef corals are interesting, for, as is well known, they are restricted to tropical seas. Accordingly, the vital limits in respect to temperature of the 18 most abundant reef species of the Florida-Bahama region were studied. The results are presented in table 6, the method pursued being the same as in all other experiments recorded in this paper: the corals were 1Experiments upon corals of the Great Barrier Reef of Australia, near Torres Straits, show that their temperature reactions are on the whole similar to those of the corresponding Atlantic genera. Thus natural selection has not aided the Florida corals in withstanding cold, or improved the resisting powers of the Pacific forms in respect to heat. 20 Papers from the Marine Biological Laboratory at Tortugas. placed in large glass aquaria and the temperature was raised slowly at the rate of about 2° C. per hour, with frequent stirring of the water. Time is an important factor in these experiments, for animals can withstand a higher degree of heat if the temperature be raised quickly than if it be raised slowly. These experiments were made during May-July 1912, at Golding Cay, Andros Island, Bahamas, and at Tortugas, Florida, when the normal temperature of the sea-water ranged from 27° to 29.5° C. as shown in table 6. It appears that the reef corals at Tortugas, Florida, live in water which is commonly within 10° C. of their upper death-temperature, and if the ocean were heated to 38° C. (100.4° F.) only one species, Siderastrea radians, could survive. This form lives on the shallow flats, often in places where the circulation is imperfect and where wide temperature ranges occur. Ac- cordingly, it is also the species most resistant to cold, withstanding 6° to 7° C. without apparent injury, but usually being killed at about 4.5° C. (40° F.), although one individual survived without apparent injury after being at a temperature of 1.9° C. (35.4° F.) for 11 hours. Next to Siderastrea radians the most resistant coral is S. siderea, although it lives in relatively deep water on the outer reefs, where the circulation is of the best and the temperature range is therefore slight. It is associated in its habitat with Orbicella annularis, one of the most sensitive of the reef corals, which is killed at 14.1° and 36.8° C. In general, however, the corals of the shallow-reef flats, such as Sider- astrea radians, Porites furcata, and Meandra areolata, are the most resistant both to heat and cold, while those of deep water, such as Madrepora palmata, Eusmilia knorrt, and Oculina diffusa, are the least resistant. Forms such as Porites clavaria, P. astreoides, Maeandra muricata, Orbi- cella annularis, O. cavernosa, and Favia fragum, which usually live in fairly shallow but freely circulating water, all show moderate powers of resistance. The air-temperature at Tortugas commonly ranges from a maximum of 98° F. in summer to about 60° F. in winter,! and the coldest “‘northers”’ in winter appear to reduce the temperature of the water over the reefs to about 17.2° C. (63° F.). In view of this fact, the more abundant reef species were maintained at about 13.9°C. (57° F.) for periods of at least 9 hours. As a result, we are led to conclude that were the water cooled by an exceptionally prolonged norther to 13.9° C. for 9 hours, Siderastrea radians, S. siderea, and Meandra areolata would survive without apparent injury, while Porites furcata, P. clavaria, Meandra clivosa, and Favia fragum would also survive, but with more or less injury, the first-named being the most resistant. On the other hand, this temperature would be fatal to Orbicella annularis, Porites astreoides, and Madrepora muricata (cervicornis). The observations cited above are upon the death-temperatures, but the temperatures at which the feeding reactions and normal metabolic processes cease are much more significant, for naturally an animal can not long survive 1In the year from June 1, 1912, to May 31, 1913, the air temperature at Tortugas ranged from 95° to 65.5° F. as determined by a self-recording thermometer. Effects of Temperature on Tropical Marine Animals. 21 in water in which it can neither move nor function. Unfortunately, how- ever, no observations were made upon this interesting point; but, crude as they are, the experiments show why it is that most of these reef corals can not enter the waters of the temperate regions and demonstrate that, in com- mon with other marine animals, they live at temperatures within about 5° of their temperature of maximum activity and within 10° of their death tem- peratures. Thus the factor of safety in respect to elevation of temperature is far less in tropical than in temperate marine animals, and they are, relatively speaking, poorly adjusted in a physiological sense to their tem- perature environment; slight differences in temperature producing a more serious effect than is observed to result from similar temperature changes with the marine animals of the temperate regions. Moreover, paradoxical as it may seem, tropical marine animals can withstand cooling better than they can survive heating above their normal life-temperature. High tem- perature appears to cause death by asphyxiation, the oxygen of the sea- water becoming insufficient to support the augmented metabolic activity of the animal. Papers from the Marine Biological Laboratory at Tortugas. 22 6z eC ejeap visieaieadeseis ey Oy, dig.s'sieicey oF 0} ot Bf Ske TW eie ley os Zf noqy PORES WA ot | 0} oIZx selaielsiekes *) Srey eZ OOp cere VI sicveineiiesry 50.0} (6°68 see ser) 68°62 0} ,§°L2 DODO OD a Wet talon wets pe ee) zz sisimieiagesaie(e'4iehsie BeCy i MEY: suse jcieiersiczene cielleyy ey. RSLS Niet ih" 3 Ce ae) “*s*I97eM UL SUTUIIOY a1 S[e4SAIO 991 VpIYM ‘yno paMey} usy} pue ONUIZUOD SJUSUIDAOUI OI JO YO] PI[Os & OjUr gI pieiecsssvenereie' esis 8 82) A ZY seus) ei sta)ecieseisiesiacy OY. er MOS “) ol I— MOP uazoly suleq SOATAING *IOJEM Ul SUIUIIOJ VOI of gieieisiainie/s:Sclmeseyr eS et OY. sheieierseisisiecisieemy LeCp Slclesecituelcierers“lsyy Ty rieieleleie)eleleseielecrny) 2gtO=te qq ro) of o— SOATAINS ‘Dov it= WIOIJ I9A0D9I SUDUTIOadS jo yyey ynOqYyY = *JoyeM oy} Ul UWIIO} sTeySAIO aor sym “Dd Qf'e— 6z ERR EIR ICC | of 2b oleksaniveny oLov 07 ov'gt sieieteary orice 0} of Of este leseseteay oS II 0} (OF pue ,v1— us9MJeq Pe wz oe ee ey Pee ee ee *9£'S— 0} BUIZIIIJ SOATAING Qz 0} zz cece eee ee eee eee eee ee sites) clenceariaie isa EVE eee eee eee eee eee wee eC rca ‘2 .9'0— ‘2 of'e— 6z ere eee eee eee weer eene eer eee eree eee eee eee erew ewer ee ere ee eeee eee eee eee eee ee eee ee eee 0. BUIZIeI} 4q Pe *‘DIjJaAny uey} WUeASISOI socces* *TOREM Ul SUL sioul suleq ‘Ainfur jua -UI1IOJ J1¥ S[eISAIO OI -redde NOY}IM 9dr OUT vr FOHOODECS Tate tet ie) 4 ssisi*isie*—y 097109 .9'Or DOOD OST.) (01 qnoqy ayy *S ov't — MOP A][prjos BUIZIeI} S9ATAING seecressy 68 SAOGE OG ISTE ACEKGIOTS aah HAN OU pozeoy jl 9ATAINS Ul SuIWI0; sje sAID “201 BY} OFUI A][pI[Os v1 Ajjensn 3") (OE 0} ,62 | *°*"" "dD o£'6z 03 (62 | °°" °°" "SD Q€Z 03 81 | aor YM “DO QvI— | Wez0Iy Suleq saATAINS “****-esnpaul JO 90ue 4s -qns snourjejes UlYyzIM ~ Seer ee eee eee #2) os BE Pceheisisey ov gt 0} oV9E CU SLL ES) 0% qnoqy sissin/ orcs) og II 0} os hl WIIO} S[e4SAIO re) P2IEST ° Saeay Bie *sind90 YyyeICq “QSB90 SJUSMIBAO/T *9SE90 SJUIUISAOJT “sind00 yzyeaqd ue 34} yom “APA 4s9} e013 Ul JazeM-eas jo aInjeIedwMaL, jo oinjzered "yeaTT “plop -W19} [EUION ‘ganqvsaquiay, 01 suoyspay—L AIAV], *II6OI A[nf ‘epriopy ‘sesnq -10 I, WOI *psopuodf DagotsspD *IIG6I “das ‘eI}ODS PAONT ‘xeyey Wolly ‘stwuNnINI g04ag ‘1161 “Sny “sseypy a) (0) 5 SpooMm mony [rTOD oy "200 “sseyy ‘wenDstuuy woz “I161 Ajnf ‘eplio,y ‘sAaxyy sesonbieyy Wolly snuayg%yog snjnutrT *II61 Ainf-ounf ‘eprop,y ‘sesny “10 J, WIOI} st4DjnI19SDf spdgaT “1161 ‘3ny ‘s}jesnyoes “SPJ ‘20H SpooM wo "yore *eoreur -ef ‘Aeg osoqUOpP, WoT “1161 Ant ‘epliop,y ‘sesnzioy wormy ID1Ja4DY) DIADUUaT *IIOI Jaquis}dag ‘eIj09G PAON ‘xXe}I[eH WlOI} 0917940 DauDaD *IIOr laquiaydag ‘e1309S PAON ‘xejyey wo "1161 A[nf—sunft ‘eplioyy ‘sesnyoy wo iDjtAND DYjaAny *jeulrue jo owen 23 Effects of Temperature on Tropical Marine Animals. ’s91d01} 9Y} 0} PpauyUod sI so1ads sIny “Zz ‘SaIPUT IS2AA OF ‘S}ESNYoesseP ‘JOH{ SpOOAA WioIy SeduLI satoads SIU], “I €Sz 03 Ezz sires; s(etwiaicisle =a ype L Le sereery (ELE 07 (eke Se Rear, UT ie KOI €'Sz 0} SE'zz eit wsisietetheleievcy LoL, IONS | ol LE 0} (ELE s/siriv ciety (OS qnoqy I'Sz 0} ghz "df LEaATAINSjOU UeD SOOO CID DRO OOR) ose sleretureiarelny o9e 0} vz 6z musemiee'2 8c) Ar 4 aaoqy BOO | ol cv 0} ollv seer see APS qnoqy “soIp [eUlTUe OS OO ON rie TERS ‘JOU Op 392j-3qn} yng ye ase Jeoj-aqn} Ajireiodula} IeA0001 1° oO LE 03 (SOE 2 Soe Oude 6z sauids $°d ,9°L£ 0} LE | aAOUI 03 asead soUIdS | aAT}OeysOUI OIE soUIdS ‘Do of sigleneie ere soy 08 03 sh es SjebsTaxens sey 08 03 oS°L SiSBSLG ie) tie hs 290 sera ee sKsis Toye aaa? say 086 03 of ‘Oo 0} UMOP 2AOUI 03 anuiquoo soulds *'D 08 OI 0} ,O1 72 BAOUI 0} aseao =. yooj-aqn |, “SUIAT] 3 aJ0M s]eUr SInd90 yee ‘asad sjUSUIBAOW] s1ue ay} yoIyM *AWAIOV 48078013 MEAs PASeee jo ainqesoduiay, jo ainjzerod -09} [CWION *98C99 S}USUISAOJT "?) L’I— ye PA OD 921 pedtaing IIO GOO TO II) ah Ys teeeeeeee egy oo MOG "D of'%— ye parry s1 IN “D oD 0} oI-+ wWoIy sIaA0DayY “sop yeullue 9y} OS ‘I9A00 -31 JOU Op 4eaj-eqn} qnq ‘aimyereduise} siyy WO1} PauUlIeM SI JeuTUe usyM All1e10dule} I3A09 -a1 soutds SY, °D 99°2+ -eoremef ‘Aeg osojU0;T WIOl} zDpnDIINIsg DiUsap0I1yGOC ‘eoremef ‘Avg 0394 sUOW sg DSsasggp DusaporygoO eoreuef Aeg oso}U0;T WOIy DUuIdssaa4q DUuLsaporygoC *saIpUy so ‘Coreulef ‘Avg OZayUO;L + WoIy ‘snipsDas “Da (snuty2ajhTJ) Sajsnaugoxo TL “1161 ‘Iz pue oz Ajnf ‘epopy ‘sesnqyoy wol1y ‘plo shep ¢ 10 (DIDINf 291uUNnNq]) WIIOM ojojed « orjuey}y jO BAIL T “1161 ‘sny—Ainf ‘epo,y ‘sesnz “10 J, WIOI} ‘uénsojas DUapDiqd *sInd90 4}e8Cd “panuyuoj—asnjosagusa, 07 suoyrvag—'L AIAV |, *jeuyue Jo aWIeN . Ph LITERATURE. BACHMETJEwW, P. 1899. Uber die Temperatur der Insecten nach Beobachtungen in Bul- garien. Zeit. fiir wissen. Zool., Bd. 66, pp. 521-604. 1901. Die Lage der anabiotischen Zustandes auf der Temperaturkurve der wechselwarmen Tiere. Biolog. Centralblatt, Bd. 21, pp. 672-675. BicELow, H. B. 1911. The Siphonophore. Mem. Museum Comp. Zool. at Harvard College, vol. 11, 401 pp., 32 plates. Boun,G. 1910. Comparison entre les réactions des Actinies de la Méditerranés et celles de la Manche. Compt. Rend. Soc. Biol., tome 68, pp. 253-255. Paris. Also: Les réactions des actinies aux basses temperatures. Jbd., pp. 964-966. DALLINGER, W. H. 1887. The President’s Address. Journal Royal Microscopical Soc., pp. 185-199. London. DAVENPORT, C. B., AND W. E. CASTLE. 1895. The acclimatization of organisms to high temperatures. Archiv fiir Entwicklungsmechanik der Organismen, Bd. 2, pp. 227-249. (List of papers.) DEMOLL, R., UND J. STROHL. 1909. Temperature, Entwickelung und Lebensdauer. Biologisches Centralblatt, Bd. 29, pp. 427-441. GREELEY, A. W. 1901. On the analogy between the effects of loss of water and lowering of temperature. Amer. Journ. Physiol., vol. 6, pp. 122-128. Harvey, E.N. tg11. The effect of different temperatures on the medusa Cassiopea, with special reference to the rate of conduction of the nerve impulse. Papers from the Tortugas Laboratory of the Carnegie Institution of Washington, No. 132, vol. 3, pp. 29-39. Jotty, J. 1904. Influence de la température sur la durée des phases de la division in- directe. Compt. Rend. Acad. Sci., tome 138, pp. 387-389. Paris. Kinc, H. 1903. Effects of heat on the development of the toad’s egg. Biol. Bull., Woods Hole, vol. 5, pp. 218-231, 4 figs. KNOWLTON, F. P., AND E.H. STARLING. 1912. The influence of variations in temperature and blood pressure on the performance of the isolated mammalian heart. Journal of Physiology, vol. 44, No. 3, pp. 206-219. Cambridge. LoeB, J. 1908. Uber den Temperaturkoeffizienten fiir die Lebensdauer kaltbliitiger Tiere und iiber de Ursache des naturlichen Todes. Pfliiger’s Archiv ges. Physiol., Bd. 124, pp. 411-426. Matisse, G. 1910. Action de la chaleur et du froid sur l’activité motrice et la sensibilité des quelques invertébrés marins. Bull. Inst. gen. psychol., Paris, Ann. 10, pp. 247-269. MavreEL, E. ET LoGRIFFE. 1899. Determination et action des plus hautes et des plus basses temperatures compatibles avec la vie de certaines poissons. Bull. Soc. Hist. Nat. Toulouse, tome 32, pp. 199-226, 2 fig. McGILL, C. 1908. The effect of low temperature on Hydra. Biol. Bull, Woods Hole, vol. 14, pp. 78-88, 1 plate, 2 text figs. Moore, A. R. 1910. The temperature coefficient for the process of regeneration in Tubularia crocea. Archiv fiir Entwicklungsmech. der Organismen, Bd. 29, pp. 146-149. Also the temperature coefficient of the duration of life in Tubularia crocea. Ibid., pp. 287-289. Morse, M. 1906. Notes on the behavior of Gonionemus. Journ. Comp. Neurol. and Physiol., vol. 16, pp. 450-456. Murray, J., AND J. Hyort. 1912. The depths of the ocean. 821 pp. PaRKER, G.H. 1908. The sensory reactions of Amphioxus. Proc. Amer. Acad. Arts Sci., vol. 43, pp. 415-455. PETER, K. 1905. Der Grad der Beschleunigung tierischer Entwicklung durch erhdhte Temperatur. Archiv Entwicklungs. der Organismen, Bd. 20, pp. 130-154. RoMANES, G. J. 1885. Jelly-fish, star-fish, and sea-urchins, being a research on primitive nervous systems. International Scientific Series, vol. 49. New York. ScuPIN, E. 1906. Etwas iiber Anabiose (Wiederaufleben) bei eingefrorenen Wassertieren. Wochenschr. Aquar.-Terrar-Kunde, Jahrg. 3, pp. 333-336. SNYDER, C.D. 1908. A comparative study of the temperature coefficient of the velocities of various physiological actions. Amer. Jour. Phys., vol. 22, pp. 309-334. VERNON, H. M. 1899. Heat-rigor in cold-blooded animals. Journal of Physiology, vol. 24, pp. 239-287. Cambridge. 1899. The death-temperature of certain marine organisms. Journal of Physi- ology, vol. 25, pp. 131-136. Cambridge. WaitnEy, D. D. 1907. The influence of external factors in causing the development of sexual organs in Hydra viridis. Archiv Entwicklungsmech. der Organis- men, Bd. 24, pp. 524-537. Witiiamson, H.C. 1910. Experiments to show the influence of cold in retarding the development of the eggs of the herring, plaice, and haddock. 27th Annual Report Fisheries Board of Scotland, part 3, pp. 100-128, 1 plate. WINTERSTEIN, H. 1905. Warmelahmung und Narkose. Zeitschrift fiir Allgem. Physiol., Bd. 5, pp. 323-350. 24 I. BHECRELATION BETWEEN THE DEGREE. OF GCONCEN- TRATION OF -ELECTROLYTES OF-SEA-WATER AND THE RATE OF NERVE-CONDUCTION IN CASSIOPEA. BY ALFRED GOLDSBOROUGH MAYER, Director of Depart ment of Marine Biology of the Carnegie Institution of Washington. Thirteen figures. 25 “ha "Gne aN y | ty fi ia 21 y Rata eae a PV SG een PET hee: oy Se Sd Ke EVAR AUS ih Nie Ang a sae acy vy ee PARA AS We ead: | k? tate Pen ta eae pA eh parti ANP AY RHEE a etal foe eo eee ne peachy ais, 1 inte i vag Af oe a ee ee Pe “Sp ; op geet: en a et As BPs cath Pty ao ov ids Sa a ra Lai iy Canes owt a ee ae THE RELATION BETWEEN THE DEGREE OF CONCENTRATION OF THE ELECTROLYTES OF SEA-WATER AND THE RATE OF NERVE-CONDUCTION IN CASSIOPEA. By ALFRED GOLDSBOROUGH MAYER. CONCLUSIONS. If sea-water be diluted with distilled water, or with a 0.9 molecular solution of dextrose, thus preserving its normal osmotic pressure but reducing the concentration of the cations of sodium, magnesium, calcium, and potassium, the rate of nerve-conduction increases as dilution proceeds, becoming most rapid in 90 per cent sea-water + 10 per cent distilled water or dextrose. In 80 per cent sea-water it is again about normal in rate, and in successively lowered concentrations it declines in a right-line ratio. The curve for rate of nerve-conduction in sea-water diluted with distilled water is practically identical with that for sea-water diluted with 0.9 molecular dextrose, thus showing that the changes in rate are due to changes in concen- tration of the electrolytes and not to osmotic pressure. (See fig. 6, page 40.) If sea-water be diluted with a solution of 0.487 molecular sodium chloride dissolved in 0.075 molecular dextrose, thus preserving the normal concentration of sodium in sea-water and the normal osmotic pressure of the sea-water, the rate of nerve-conduction is most rapid in about 90 per cent sea-water + Io per cent of this solution and then declines becoming about normal in rate in 80 per cent sea-water + 20 per cent of the (NaCl + dextrose) solution, after which it declines in a right-line ratio upon further dilution, but more slowly than if the sea-water had been diluted with dis- tilled water or dextrose. Thus the sodium cation is an active stimulant for nerve-conduction. (See fig. 7, page 43.) Similar experiments with the magnesium cation show that it is not a stimulant for nerve-conduction, being no more effective in this respect than is distilled water. Thus it is inert and nontoxic, and exhibits only neg- ative properties and probably it should not be classed as an active inhibitor. Speaking crudely, its rdle in respect to sodium in sea-water is comparable to that of the nitrogen of the air in relation to oxygen. (See fig.9, page 47.) In very slight excess the potassium cation produces a permanently stimulating effect, as does sodium, but in denser concentration it produces momentary stimulation of the rate of nerve-conduction followed by de- pression. In all essential respects the effects of potassium are similar in kind, but more marked in degree, to those of its close chemical ally, sodium. 27 28 Papers from the Marine Biological Laboratory at Tortugas. I can not determine the individual effects of the calcium cation owing to its habit of associating itself with the sodium, thus enabling the sodium to offset the effects of magnesium. The laws stated above for the rate of nerve-conduction apply also with modifications of detail to the rate at which the motor centers (the rhopalia) generate the stimuli which produce the nerve-impulse. The rhopalia are, however, more readily affected by osmotic and by concentration changes than are the nerves, and I have not observed any increase in rate upon diluting the sea-water with distilled water or dextrose. Any change in concentration either above or below that of normal sea-water produces depression in the action of the rhopalia. Attention may be directed to an explanation of the apparent converse relation between the activities of muscles and cilia in trochophores, cteno- phores, and other forms having well-differentiated cilia which move in a coordinated manner. In all these cases the normal muscular tonus of the animal produces a state of tension over the outer skin, thus pressing upon the cilia-bearing cells and reducing or even stopping their movement. When this tonus is relieved, however, the cilia beat rapidly. Thus magnesium reduces the muscular tonus, but its depressant effect upon cilia is not so marked as upon the muscles, and the cilia-bearing cells being relieved of pressure beat with abnormally great activity in such a solution. Sodium, on the other hand, contracts the muscles, thus increasing the pressure upon the cilia-bearing cells and stopping the movement of their cilia. Hence this converse relation between neuro-muscular and ciliary movement is a mechanical, not a chemical, matter.! THE GENERAL NATURE OF THE PULSATION-WAVE. Casstopea xamachana is a very favorable animal for physiological experi- mentation, for it normally lives in semistagnant lagoons and is, therefore, relatively insensitive to changes in osmotic pressure due to concentration or dilution of the sea-water. Moreover, as there is probably a considerable amount of dissolved CO, in its normal environment, and as the medusa is infested with symbiotic plant-cells, it thrives remarkably well in aquaria, there being practically no deaths due to confinement under laboratory conditions. In this series of experiments, whenever the animals were placed in a new solution they were permitted to remain in a large quantity of it for at least an hour with several changes before the effect of the new solution was determined by a kymograph record. In this manner the ultimate condition of penetration and balance between the animal and the surrounding solu- tion was secured. The relative concentrations of the solutions were deter- mined by titration with AgNO; and potassium chromate. The effects described are all practically reversible, and fatal or seriously injurious degrees of dilution of ions were avoided, so that when the animals 1 Mayer (1912), Ctenophores of the Atlantic Coast of North America, p. 25. Carn. Inst. Wash. Pub. No. 162. Papers from the Marine Biological Laboratory at Tortugas. 29 were returned to sea-water they soon regained most of their normal rate and amplitude of pulsation. The stimuli which cause pulsation normally arise in the motor centers called rhopalia, or marginal sense-organs, of which there are usually about 16 in Cassiopea xamachana, although they may vary in number from 9 to 23. It appears that this normal stimulation is due to the constant formation of sodium oxalate in the rhopalia, thus precipitating calcium oxalate to form the crystals of the sense-club and setting free ionic sodium, which is a most powerful stimulant. The pulsation-rate of the normal medusa is therefore the rate at which its motor centers can produce successive stages of stimulation. This is always much slower than the rate at which the nerves can conduct the pulsa- tion-wave around the subumbrella of the medusa, and there is a long pause after each contraction, during which the animal remains unstimulated and hence inactive. Casstopea is, however, an excellent animal for physiological studies upon nerve-conduction, on account of the ease with which we may dis- pense with the motor centers and yet maintain a continuous neurogenic pulsation-wave in the subumbrella tissue. This method was devised and described by Mayer, 1906, p. 22, and it consists in cutting off the rhopalia, thus paralyzing the nervous network of the subumbrella. Then the sub- umbrella tissue is cut into the form of a ring-shaped strip, and a pulsation- wave is started in one direction in this ring, so that it must continue to travel around the ring, being in effect entrapped within the circuit of tissue through which it must continually progress in one direction, a single stimulus often maintaining itself in this manner for days at a time. Such entrapped circuit-waves are of course not peculiar to Cassiopea, for they may be started and maintained in ring-shaped strips of the ventricle of the heart of the sea turtle, Caretta caretta, in the hearts of fishes, or in other scyphomedusze. We may thus study the operations of the nervous network as a continuous transmitter of a single pulsation-stimulus. As has been said, the nerves transmit this stimulus around the bell much more rapidly than the motor centers can engender it. For example, a medusa, which normally gave 45 pulsations per minute, was deprived of its rhopalia and cut into the form of a ring, and a pulsation-wave then traveled constantly around this ring at the rate of 129 turns per minute. This ring also responded without missing a pulsation to 152 Faradaic shocks per minute, and it followed 190 shocks, responding by a separate con- traction to each one for about 3 minutes, after which it occasionally failed to respond, following alternate stimuli. Many experiments of this sort were made, always with the same general result, and it is evident that the nerves and muscles can react to successive stimuli with greater frequency and at shorter time-intervals than they are called upon to respond in nature. It is evident, then, that there is a considerable ‘‘factor of safety’’ in the normal medusa, and that the motor centers initiate pulsation-stimuli at arate 1 Mayer, 1908, p. 130. 30 Papers from the Marine Biological Laboratory at Tortugas. much slower than that to which the nervous network and the underlying muscles can respond, and thus the danger of muscular fatigue is avoided. As has been pointed out by Romanes (1885), Bethe (1903), and Harvey (1911), there must be some mechanism of adjustment by virtue of which the wave travels at a faster rate around the rim than at points nearer the center of the subumbrella, the rate being directly proportional to the length of the radius of the zone of pulsation. In order to determine the average rate of transmission of the pulsation- impulse by the nervous network, eight separate rings were cut, one from each of eight Cassiopea xamachana. ‘These rings were each 90 mm. in outside diameter, 67 mm. in inside diameter, and 11.5 mm. wide. Thus the mean diameter of each ring was 78.5 mm. and its mean circumference 247 mm. TABLE I. Diameter | Times per . ; of medusa minute Distance Ring No. on record No. I. | from which| wave went wave Tempera- the ring around traveled ture, was cut. ring. per second. mm mm e Wsscick etecsl akesorevs(alerets ave reta sieve 98 II7 481 30.45 C OE cc 5 Sete eee CHEE CICUC DIT I10 116 478 2 erisihiatct ncatiayer eat ubetet eve tere taterets 109 113 464 20.7 BONY aie siete giclee ile suntele ts III 106 436 29.8 Sidaddoombénonspcsecocks 110 104.5 429.5 29.8 Biaraisuavote asetetere ates teetener rete 97 103 422 29.5 Seer cisra ane eraienetah charelereve tects 108 103 422 30.3 Ase React lovee wiete eievevere eietalerets II5 93-5 384 28.5 Avia CONGILION sa everest clelerele 107 107 440 28.9 * This ring exhibits pulsus alternans. For an explanation of this phenomenon, see G. A. Mines, 1912, Proc. Cambridge Philosophical Society, vol. 16, p. 617. The results obtained from the kymograph record, figure I, are stated in table 1, which shows the rate of nerve transmission, in millimeters per second, determined in natural sea-water, ranging in temperature from 28° te 20:45° C. These 8 meduse ranged from 97 to I15 mm. in diameter, and their pulsation-waves traveled at rates varying from 384 to 481 mm. per second, the average being 440 at an average temperature of 28.9° C. In these experiments, the rate of transmission bears no relation to the size of the medusa. All experiments show that, as in other invertebrates, strong stimuli travel faster than weak ones. Thus should a ring stop, we can not start it again and continue the same series of experiments with it, for the new stimulus is almost certain to differ in strength from the former one; or, if all care be taken to make the successive electrical stimuli as nearly as possible identical in strength, they will be received differently by the tissue and the resulting pulsation-waves will differ in rate. Thus a ring must remain in pulsation throughout the entire series of experiments that one performs upon it or no conclusions can be drawn respecting the relative rates of its pulsation-stimulus in the various solutions. Papers from the Marine Biological Laboratory at Tortugas. 31 On the other hand, temperature, purity of water, etc., being unchanged, the rate remains nearly constant for any single stimulus. Thus Harvey (1911, p. 131) maintained such a ring in pulsation for 11 days at an average velocity of 774 mm. per second. I once had a ring which remained in pulsation for 6 days, the wave traveling from 200 to 206 times per minute around the ring. The rings live for two or more weeks and exhibit abortive efforts at regeneration, but being without means of obtaining food they gradually decline in size and die. “Ring No.1 30°45 C DIVAN nensnrrsprwrry~r“~rr_ ~wrryr~r~r“~“r SINISE NII IIIS NINN IIIS 2 Baie 3 2%°3 NYAVANI AVIVA VAVAYAVAVAAT AQAA ACQQ AAA 2° AA. ADA 00D AAD AAA 4 BSE S EIS I ri ll is it Ra ER ie a5. 5 29.5 _ 6 29°% 7 29°7 % 30° 3 % rings each 10 mw outcide, mene! © (oy) mm. inside dtameter, FIG. I. These pulsating rings are usually almost machine-like in their uni- formity of rate under unchanged environmental conditions; but while the rate of nerve-conduction remains quite constant, the muscles become fatigued and thus the amplitude of pulsation steadily declines, becoming much reduced at the end of a few days. Thus the muscles fatigue more readily than do the nerves. Kymographic records enable us to study not only the rate, but the form of the pulsation-wave, and this is important; for in order that comparisons may be made it is necessary that the character of the wave should not change throughout the series of tests to which the ring may be subjected in the course of any one set of experiments. For example, the series of lines on the record (fig. 2), which were obtained from a single medusa, can not safely be compared one with another and the record is therefore worthless, on account of the tendency of the pulsation- stimulus to break up into two waves, the one following the other at the same or at different rates as in lines 1 and 3 from the top of the figure; or, the wave may break up into a complex series of strong and weak wavelets due to spontaneous excitement, as in line 5. as Papers from the Marine Biological Laboratory at Tortugas. When a simple contraction-wave breaks up into two or more wavelets, the one following the other, the rate declines, for strong stimuli travel faster than weak ones; and if an originally unitary wave divides into two, the stimulus represented by each component wavelet is naturally less than that of the wave asa whole. Thus a ring which gave 168 pulsations per minute EiGs2. with a unitary wave gave only 132 per minute when this wave broke into two component wavelets, the one following the other around the ring. Moreover, if two waves follow one behind the other, the tissue does not enjoy so long a resting-period as if one simple wave were in the circuit, and a certain time must be permitted to pass before the muscles which have responded by a contraction can again fully respond to another stimulus. In some rings the tissue may give a strong response only to every alternate or even every third return of the wave, showing that the muscles which have very recently contracted can not again give a maximal response until a certain resting-time has elapsed. This was first observed in the muscles of Scyphomedusze by Romanes, 1885, and has been studied in detail by Bethe (1903 and 1908). (See line 6 of fig. 1.) For these reasons, a double wave caused by the breaking up of an originally unitary impulse moves slower than did the parent wave. Gener- ally speaking, when a ring is first set into pulsation by electrical stimulation, there are a number of secondary wavelets, but these soon fuse into a single simple wave, and this unitary wave then rarely breaks into component parts. In Cassiopea, as in Limulus, the stimulus which produces pulsation is neurogenic, the muscles being incapable of spontaneous pulsation if the nervous system be eliminated. Thus a 0.01 molecular solution of oxalic acid in sea-water is extremely toxic to the motor centers and nerves of Cassiopea, a very short immersion preventing the rhopalia from ever again giving rise to the pulsation-stimulus. The muscles, however, still remain capable of contracting when stimulated directly. In this connection, Harvey found that in most of the amines and in inorganic hydroxides the nerves cease to conduct before the muscles lose —wetes Papers from the Marine Biological Laboratory at Tortugas. 33 their power of contractibility, and under these conditions the muscles respond only when directly stimulated. If the subumbrella tissue be heated, however, the nerves continue to function for several degrees beyond the point whereat the muscles cease to respond by contractions. The subject of the neurogenic nature of the pulsation-stimulus in Cassiopea is discussed by Mayer (1906, p. 19); and, as is well known, Carlson has presented convincing proof that the pulsation-stimulus in the heart of Limulus is neurogenic. The neurogenic nature of pulsation in these invertebrates is the more remarkable because recent work has decidedly strengthened the myogenic theory for the beating of the vertebrate heart. Thus Paton (1907) shows that both the heart and the axial muscular system of Pristiurus functions spontaneously before the development of definite nerve-fibrils from the central nervous system, although protoplasmic connectives are seen ex- tending from cell to cell, and these may transmit impulses. Also, Hooker (1911) finds that the heart differentiates and functions in a normal manner in frog embryos which have developed without a central nervous system; and a crucial experiment has been made by Burrows (1912), who observed that single, isolated muscle-cells of chick hearts, if placed in proper media, pulsate as does the heart itself. As the experiments of which we are about to speak involve alterations in the composition of the sea-water, it is important that we should under- stand the composition of the Tortugas sea-water in respect to the cations of sodium, magnesium, calcium, and potassium. Accordingly the follow- ing analysis of such of its components as may affect the experiments is presented. COMPOSITION OF TORTUGAS SEA-WATER. As a deduction from George Steiger’s analysis,| made under the direc- tion of Professor F. W. Clarke, it appears that in so far as its ingredients (sodium, magnesium, calcium, and potassium) are concerned, we may make up Tortugas sea-water by mixing the solutions shown under A, each being practically isotonic with the others, and with the sea-water as a whole: A. B. 81.1 volumes of 0.6006 molecular NaCl. 0.487 molecular NaCl. 14.3 n “ 0.398 + MgCl. 0.0571 S MgCh. 2.84 i“ ** 0.389 * CaCl. 0.0110 re CaCh. ie Ae “" 0.597 “s KCl. 0.0101 " KCl. 100.00 Thus any given volume of Tortugas sea-water may be considered to con- sist of an atmosphere composed as in B, all occupying oneand the same space. At 25° C., however, this being near the usual temperature of the tropical sea-water, the NaCl is dissociated to about 75 per cent; and this is also true of approximately 80 per cent of the MgCls, 88 per cent of the CaCle, and 95 per cent of the KCl. 1 Carn. Inst., Year Book, No.9, 1910, p. 122. 34 Papers from the Marine Biological Laboratory at Tortugas. Thus among every 4,870 molecules of NaCl, about 1,217 are at any one instant undissociated, while 3,653 are dissociated; or, expressed in tabular form for the sea-water as a whole: 4,870 molecules of NaCl 1,217 fixed + 3,653 dissociated. 571 : MgCl = 114 aaa 10 x CaCkh = 13 Py 97 IOI 7 KCI = 5 + 96 The NaCl and KCI dissociate each into 2 ions, while the MgCl. and CaCl, break up into 3 each, thus: 3,653 dissociated molecules of NaCl give 3,653 Na and 3,653 Cl’. 457 MgCl. 457 Mg “ = 914 Cl’. 97 ‘c “ CaCle 97 Ca ce 194 Cl’. 96 i 7 KCl oe Ss 96 Cl’. Totalecace Gate meio 4,303 4,857 In other words, in Tortugas sea-water for every 1,000 Na cations there are 125 Mg, 26 Ca, 26 K; and the anions may be represented by 1,330 Cl’. As is well known, at least one-fifth of the Cl’ anions may be replaced by the SO,” ion without appreciable effect, this being accomplished when the sea-water is made up of sulphates of magnesium, calcium, and potassium instead of chlorides. The cations, on the other hand, play a far more defin- ite role in the control of movements; for sustained pulsation is impossible in the absence of even the least concentrated cations, such as Ca or K. If the sea-water be concentrated to 1.69, by dap) mec Naen evaporating 100 volumes of natural sea-water to 59 0.0965 “4 MgCh. volumes, the molecular concentrations of the various 09-0186 {| CaCl. 4 0.0171 KCl. constituents becomes: This results in a relative preponderance of the Mg ions, for the degree of dissociation of the NaCl declines in a greater ratio than that of the MgCl; for in this concentrated sea-water about 70 per cent of the NaCl is dissociated, as are also about 77 per cent of the MgCle, 86 per cent of the CaCle, and 94 per cent of the KCl. Thus: 8,230 molecules of NaC! give 2,469 fixed and 5,761 dissociated. 965 “é “ec MegCle “cc 222 “ce 743 186 “ ae CaCl, “e 28 ae e 158 “ 7 ae ae KCl “ce Io “ce “e I61 “a And of the dissociated molecules: 5,761 dissociated molecules of NaCl give 5,761 Na and 5,761 Cl’. 743 kg “MeCh “ 749 Me“ 7,486,.Cl 158 Zi - real” “ISS ca 316 Cl: 161 e ‘ * KCl “SG. 9 Oe Gr Gs Motal arias as pie a See ase 6,823 7,724 Hence, we see that when the molecular concentration of the sodium, mag- nesium, calcium, and potassium is 1.69 times that of natural sea-water, the concentration of the cations of these salts is only 6323 or 1.585, and, therefore, the concentration of the cations has not quite kept pace with that of the sea-water as a whole. Moreover, in this concentrated sea-water, for every 1,000 Na cations there are 129 Mg, 27 C4, 28 K, and 1,340 Cl’ anions. Thus the propor- “err ep © Papers from the Marine Biological Laboratory at Tortugas. 35 tionate concentration of Na has decreased with respect to the other cations, and as Na is the most potent stimulant while Mg is the depressant, we might expect the stimulating power of this concentrated sea-water to have increased less rapidly than the ratio of concentration, for in natural sea- water the ratio of the Mg to Na is as I to 8, and in sea-water concentrated to 1.69 it is as I to 7.75. If, on the other hand, the sea-water be diluted to 50 per cent of its natural concentration by mixing it with an equal volume of distilled water, a calculation, similar to the one given in detail above, indicates that while the molecular concentration of the dissolved salts has declined to 0.5 (that of natural sea-water being 1), the concentration of the cations is 0.528; and for every 100 Na cations there are 124 Mg, 26 C4, and 25 K cations, and 1,376 Cl’ anions. The results of these several calculations are sum- marized in table 2. TABLE 2. Composition of solution. Gram-molecular concentration Degree of dissociation. of the inorganic salts. Ss. W. =sea-water. d. w. =distilled water. NaCl. | MgCk | CaCle.| KCl. | NaCl. | MgCk.| CaCh. | KCl. per cent|per cent/per cent per cent 9 100 c.c. 8. w. +100 c.c. of d. w.......- 0.243 | 0.0285 |! 0.0055 | 0.005 79.6 84 92 INatiical cea=waters oc.ciciccle sicicic slr ele cle -487 (05 710|) .OLEEO!|.OLOD |) 75 80 88 95 Ioo volumes of natural sea-water eva- porated to 59 volumes.......... .823 .0965 | .0186 | .or7I1|] 70 17 86 94 Composition of solution. oe romnate eee eee Relative con- |Relative con- K cations for every 1,000 centration of | centration of S. w. =sea-water. Na cations. cations of sea-water as d. w. =distilled water. Mis hn = loka. sea-water. a whole. Mg. Car K. 100 c.c. Ss. w.-100 c.c. of d. w........ 124 26 25 0.528 INatiiral ‘Sea-water «5 .7ec s:ciercie1e ie elelems si 125 26 26 1.000 100 volumes of natural sea-water eva- porated to 59 volumes.......... 129 27 28 1.585 As the osmotic pressure of these solutions is proportional to the respec- tive sums of their fixed molecules and dissociated ions, we find that if the osmotic pressure of Tortugas sea-water be 1, that of 50 per cent sea-water is 0.5004, and that of 1.69 sea-water is 1.64. Thus, knowing that the osmotic pressure of Tortugas sea-water is about 24.8 atmospheres, that of sea-water diluted with an equal volume of distilled water is 12.4, and that of 1.69 sea-water is 40.7 atmospheres. Table 3 shows the relations between the osmotic pressures and concentrations of the various solutions experi- mented with in this research, and also the molecular concentration of a solution of NaCl which is practically isotonic with each solution. We must remember, however, that pure NaCl solutions, such as those mentioned in this table, are not, strictly speaking, comparable with those found in sea-water. Thus Osterhout (1911) shows that a pure NaCl solution penetrates the mem- branes of plant-cells more rapidly than if the NaCl is mixed with CaCl in the proportions found in sea-water. There is also other evidence of a chemical combination involving the NaCl and CaCl, of sea-water in their mutual influence upon animals. For example, Mayer (1906, p. 49) states 36 Papers from the Marine Biological Laboratory at Tortugas. that in the pulsation of Cassiopea the calcium of the sea-water assists the NaCl! to counteract the inhibiting influence of magnesium, and later Meltzer and Auer, in an important series of studies upon vertebrates, found that solutions containing both calcium and sodium counteracted the inhibiting tendencies of magnesium. Experiments upon Cassiopea show that cal- cium alone can not produce tetanus and does not offset the effects of magnesium if sodium be absent, and thus it appears that the so-called “calcium tetanus’”’ is actually a result of the association of sodium with calcium. Mines (1911 A) believes that calcium enters into combination with cardiac tissue, but magnesium does not, and our results support this view. Fortunately for physiological experimentation, however, Cassiopea is but little affected by considerable changes in osmotic pressure, and table 3 is valid for all practical purposes. TABLE 3.—Changes in osmotic pressure and in concentration of dissolved salts in diluted and in concentrated sea-water from Tortugas, Florida. Relative Gram- Gram-molecular concentration of Composition of solution. concentra- | Osmotic | Molecular NaCl, MgCl, CaCl, and KCl of sea- tion, that | pressure | Concentra- |water solution, regarding sea-water as s. w. = Sea-water. of sea- in atmo. | .tion of an being a solution of chlorides. d. w. = distilled water. water spheres. ee aner being 1. pure NaCl.| NaCl. |MgCk.| CaCk. | KCl. 85 GiGi Sawe-t-O51G.C. Ge Weristecins 0.35 8.7 0.206 0.170 | 0.02 0.0038 0.0035 40 607 2 AR rete 4 9.9 -236 az .023 .0044 004 50 SO ieee: Neltpyei aes 5 12.4 -295 243 .028 -0055 005 58 Aaa wane Ct ecto ett. 58 14.4 -342 -285 $033)" || -0064 0058 60 AO! > Da G ars'ceeettes 6 14.9 +354 -295 -034 | -0066 006 66.6 ch ee Boe CeO 66 16.35 +393 325 -036 .007 0066 75 25 Le) MP tacen: 75 18.6 -442 375 -043 -00825 0075 Pureisea=waters jcc eneocios. ie: I 24.8 “59 487 .0571 OIL OIror 100 c.c. Ss. w.evap'd to 89.3 .... I.II9 27.4 -66 «545 -064 0123 O102 86:8. ..56< I.152 28.3 .68 -50 -066 0126 OII5 B4E3) Skie,s 1.186 20.2 -70 -58 .068 O13 Ors 800 Some 1.25 30.6 -737 -61 .O71 0137 O125 shay eee. 1.28 Biss -75 62 .073 O14 0128 7 A RRO 1.322 32.2 -78 64 .075 O145 0132 Pas acta 1.407 34.2 83 68 .08 OI54 Or4 6052)... 1.444 35 85 aT 082 0159 O147 O25 eee: 1.6 38.6 94 -78 Oot OI75 O16 BOSS Ste 1.68 40.5 -99 -819 .005 Orgs 0168 59 1.69 40.7 I.00 823 -096 0186 0169 RELATION BETWEEN THE CONCENTRATION OF THE CATIONS OF SEA- WATER AND THE RATE OF NERVE-CONDUCTION. If we dilute sea-water by mixing it with distilled water, or concentrate it by slow evaporation at ordinary temperatures, we not only decrease or increase the concentrations of the electrolytes, but we change the osmotic pressure in a nearly equal ratio. We may, however, avoid lowered osmotic changes by diluting the sea-water with a 0.9 molecular solution of pure dextrose,’ which is practically isotonic with sea-water, and in this manner we may reduce the concentrations of the electrolytes and at the same time maintain the normal osmotic pressure. Thus we may determine the effect of changes in concentration of the electrolytes independent of osmotic pressure. The results of such experiments are shown in tables 4, 5, and 6, while table 7 shows the arithmetical averages derived from the results recorded 1 The dextrose used in these experiments was the best that could be obtained from Merck and from Kahlbaum. The accepted samples were tested for sodium and free acids and found to contain inappreciable quantities of these impurities. The dextrose solutions were made up afresh each day. Papers from the Marine Biological Laboratory at T orlugas. 37 in tables 4 to 6. In all these tables the rate is reduced to 100 per minute in natural sea-water in order to facilitate comparisons. Thus, if on the average the medusz actually pulsated at the rate of 80 per minute in natural sea-water, and at 40 per minute in some other solution, these numbers would become 100 and 50, respectively, in the tables. Al Sea water 29°C. 75 sea watert 25 distilled water 27246 UG: i 8 Swi ss . 20:5) — 2in So sea water+ Sodistrilled water 27 .%).C SO Ss€a water + Sodistilled water 2G Dey, 75 s@a water. +. 25 distiltted water — 30° 75 Sea water + 25 distilled water DPLLLPLLPLLL LLLP LLILL LIF LLLLLLDLDLLDLDDLDPYJDLLL PIII DaBBBEYYYYIYYIYPYIYPYYLIPPUPADRAPDDDDALRARADRARAAR, Sea water . PDL PD PDL PLL LLLLLLLLP LLLP LPP LLLP PPD PLPP PD PRADPRDALRARRRALAA, DPLDP LDP LP PPLPLLPLPLPDPDLPLAPPLPLAPPPPPPPPPLPIDLS sea water 29°95 7a Sea water 29.5 Bel ei ee i ere A we SI NS ed la a SNS IIR 75 sea water+o5 distilled water 29.9C Pie, 2° Table 4 shows the relative rates of nerve-conduction in 20 rings, and of pulsation in 4 normal medusz of Casstopea xamachana in sea-water and sea- water diluted with distilled water at 29° to 2 Ban Gee TABLE 4. “Ae “ Z Relative rates of pulsation, the rate in normal sea-water ag ars hs the solu Osmotic Bein 30h: : pressure in ae S. W. =sea-water. ce Ring Ring Ring Ring | Ring | Ring | Ring | Ring | Ring d. w. =distilled water. No. 1.1} No. 2. | No.3. | No. 4. No. 5.|No. 6.|No. 7. No. 8.|No.9, Natural sea-water........ 24.8 100 100 100 100 100 | 100 | 100 | 100 | 100 95 p.ct.S.w.+ 5p.ct.d.w. 23.6 [foreterats stall eceratetereyel|fetaneveta sail occretryatalllonecsescllle teres IPSikeee I caida 90 10 Bey Ne Set: HRacaiionavarecsillzacacieverscelltevaeansi seo fie tcl bo a orem ew ee [ie eel ees 85 15 PRES MGemieemol Arien sere) eee || rea ae Sana OE een Hs a [seats fotei| Sees 80 20 BOO n de torr asiiatatass KM OA ce Pesce cats ov sl saou ele Sodhe tcl oewke 75 25 18.6 87 86 88 86 81 86 ive 86 89 70 30 E-ESe A fararniasioie\ffea/ara0 2% sl|or+1S ste etal llesats iim Sail alaietste:| erste: alee ise loka eee 60 40 TATAG Merton eee ene a 7G | \aarstslexecs 65 5 Su ilhsectecalice sie 81 | 50 50 12.4 GF Wwe OAV OER Rroreessel ters stra ova ci oreo ree lane ae 62 40 60 OD sea | im hip eves hail ieYs edeh ches AILWe vo. ches el| ic, wiahors ove, | GuasSeethinmiccte'lloscheeens Nerorerualitaretaors 35 65 Soe aml eictaataite Ipoaoke yale tel fe patscakey esi |[alcuet oye\ et [rtteeleceselerses[eeees kaveretete Relative rates of pulsation, the rate in normal sea-water Composition of the solu- Osmotic being roo. tion. pressure in — - ae atmo- Ring | Ring | Ring | Ring | Ring| Ring Ring | Ring | Ring - a Ea ae spheres. No. No. No. No. | No. | No. | No. { No. | No. Io. Eq rae rs. I4. I5. 16. 7s 18. Natural sea-water......... 24.8 100 100 100 100 I00 | 100 | 100 | 100 100 95 p.ct.s.w.+ 5p.ct.d.w. ASO a alicten tic. e ails ae eal eer MES by | tea: ae 105 | 108 | 107.5] ror 90 Io CI el il are tent He cer enell Me bite Le be a IOI II3 | 106 105 85 15 it aed Feat SR Op ee eee Ee 20). Ee ee 99 | II2 | 107 105 80 20 ZOOM) “er joatee seet eee ELENA. =. 94 | 104 96.5} 108 75 25 18.6 Co) Saal ear QAP Psislsisic ais SAP Warslons:aillecsccrs oi re. cree eons 70 30 Tie ete |ecasevorcte: olin ect ttn mere Sa ih ene 8r S37) 7) 205 60 40 14.4 50 OA Weteunetenn letlewsi We 0 Py aAeiate 50 57 58 56.5 1See record, fig. 4. 2 Six hours after being restored to sea-water, rate was nearly normal with amplitude lower than normal 3 When restored to sea-water, this medusa pulsated at about two-thirds normal rate. See record, fig. 5 Papers from the Marine Biological Laboratory at Tortugas. 39 PDDLYrYYrYLYIYIYYLYIIYYYYLYPLPPPILPPPPPPADPP LPP PPP BPP LLLP PPLL PLL LAA LPL Sea water i, hs PDP LDILLPSILLLPLPLPPPDDLDSDIARRALDARRARLRA a 75 Seawater + 25 dextrose TEBE ODADAS 3O sea water + 50 dextrose 2%: .¢ (cee Se SS a a Ne aes Pe iPS ag aS ie Se rem ee eo S50 S€a water + 50 dextrose ML hag ove i a YPSPSYIIDSIN NINES ALA SIC NN selon ey) 23°S$ C Fic. 4. So clist:lLled water + 50 Sea water Pla ie SS ye a aa a a Ce | | 6 Sea water 2 Sal S 33.33 dextrose + 6666 sea water 31,15. © ——— — —_ 42 dextrose + S% Sea water 24..C. sea water July 8, Sh 5Sém PM 2.75.6. Sen water Sata: 9! Gh. 14m AM Ba le: Fic. 5. Table 5 shows the relative rates of pulsation of 2 medusz and of nerve- conduction in 20 rings of subumbrella tissue of Cassiopea in sea-water, and in sea-water mixed with 0.9 molecular dextrose. Table 6! shows the relative rates of nerve-conduction in subumbrella tissue of Casstopea xamachana in sea-water diluted with distilled water, and also in sea-water diluted with 0.9 molecular dextrose. (See tables 4 and 5.) The practical coincidence of these two curves indicates that the changes in rate of nerve-conduction are due to corresponding changes in the con- centration of the electrolytes and not to changes in osmotic pressure. _ 1See fig. 6, which shows that these two curves are apparently identical, the differences between them being probably due to errors of experimental nature and individual variation. 40 In figure 6 the full line and open circles represent the rates in sea-water diluted with 0.9 molecular dextrose, and the dotted line and black circles Papers from the Marine Biological Laboratory at Tortugas. show the rates in sea-water diluted with distilled water. TABLE 6. Composition of the solution. S. Ww. =Ssea-water. d. w. =distilled water. Relative rates of nerve-conduc- tion, that in natural sea- water being rIoo. Sea-water di- luted with dis- tilled water, thus corre- spondingly reducing osmo- tic pressure. Sea-water di- luted with 0.9 molecular dex- trose, thus maintaining normal osmotic pressure of 24.8 atmospheres. Natural sea-water, salinity 36.15 p.ct..... 100 100 95 p. ct. Ss. w.+5 p. ct. d. w. or dextrose... I03 104 90 10 Pid 105 106 85 15 IOI 104 80 20 100.5 100 75 25 86 92.5 71 29 aistete 87 70 30 82.5 ete 60 40 70 73 50 50 50 56.5 Figure 6 gives a graphic representation of the results stated in table 6, the ordinates giving the average rates of the pulsating rings, and the abscissz giving the relative concentration of the dissolved salts of the sea-water, that of normal sea-water being 100. Thus the ordinates of the full-line curve show the relative rates of conduction of the nerve-impulse when the sea-water is diluted with 0.9 molecular dextrose, and the dotted curve shows the same for sea-water diluted with distilled water so as to reduce its 100 90 80 B 8 70 60 50 50 60 70 80 80 100 Relative Concentration of the Sea-water Solutions. Fic. 6. osmotic pressure in nearly the same ratio as the reduced concentration of the electrolytes. The practical coincidence of the curves for sea-water, diluted with distilled water and with 0.9 molecular dextrose, shows that the phenomena in question are not due to variations in osmotic pressure, but to the varying concentrations of the electrolytes. Moreover, both these curves show that the most rapid rate of nerve- conduction is in diluted sea-water, wherein the electrolytes are only about Papers from the Marine Biological Laboratory at Tortugas. 41 gO per cent as concentrated as in natural sea-water. Below this the rate declines, becoming normal in 80 per cent sea-water and from this point down- ward it declines in a straight-line ratio, at a somewhat more rapid rate than the corresponding decline in concentration of the dissociated electrolytes. The curve as a whole reminds one of that of an enzyme reaction having its maximum in 90 per cent sea-water and rising in a straight line from 50 to 80 per cent. This does not prove, however, that it has anything to do with enzyme action. It will be recalled that Loeb (1891) discovered that growth and regen- eration in hydroids proceeds at a somewhat more rapid rate in sea-water diluted with 15 to 20 per cent distilled water than in normal sea-water. The experiments of Goldfarb (1907) upon the regeneration and growth rate in hydroids and his more recent observations upon Cassiopea! confirm this fact. Loeb ascribed the increase in rate of regeneration to the effects of potassium and magnesium, but Goldfarb’s observations cast doubt upon this simple explanation. It is clear that the initial increase in the rate of nerve-conduction in Cassiopea, which accompanies the decline in con- centration of the electrolytes below the normal, is not due to lowered os- motic pressure, for it occurs when the sea-water is diluted with 0.9 molecular dextrose, thus maintaining the osmotic pressure found in normal sea-water. A possible but purely hypothetical explanation may be that if the rate of nerve-conduction is due to an excess of ionic sodium in the tissues over and above its concentration in the surrounding sea-water, a slight lowering of the concentration of the sodium in the sea-water causes a corresponding increase in its ratio of concentration within the tissues and may thus cause an augmentation in the stimulus. There is some evidence to support this view, for when the sea-water is diluted with 0.487 molecular sodium chloride, the rate does not increase as rapidly as it does when the sea-water is diluted with distilled water or with dextrose. The slight initial increase in rate when the sea-water is diluted with 0.487 molecular NaCl may be due to the reduction of the potassium ion which in such concentration is a stimulant, as is sodium. THE RELATION BETWEEN THE RATE OF NERVE-CONDUCTION AND THE CONCENTRATION OF THE SODIUM ION. In order to test the effect of the sodium cation upon the rate of nerve- transmission, we may add increments of a solution of 0.6 molecular NaCl to the sea-water, thus maintaining a constant and normal osmotic pressure, but increasing the concentration of Na while the concentrations of Mg, Ca, and K are decreased. The physical effects of such a procedure are illustrated in table 7, showing that if we add 0.6 molecular NaCl up to 100 sea-water + 800 NaCl we have changed the concentration of the sodium ion from I to 1.205; while at the same time, in comparison with the Mg, Ca, and K, the Na ion has become relatively 10.85 times as concentrated as it was in natural sea-water. ; 1See Dr. Goldfarb’s paper upon the rate of regeneration of Cassiopea in concentrated, normal, and diluted sea-water; page 88 of this volume. 42 Papers from the Marine Biological Laboratory at Tortugas. TABLE 7. Proportionate es mposition of the solu- | Relative dilu- | Relative |of.Nacompared|) Gram-molecular concentrations of— poe tion. ltion of the Mé,| concentra- |With that of the Ca, and K of|tionof Naof/Mg, Ca, and K Ss. w. = sea-water. sea-water. sea-water. |(proportion in seacwater! being|= 111. as Sot Be? Wee I.00). NaCl. | MgCh. CaCh. KCl. Natural sea-water..... 1.00 I.00 H I.00 0.487 | 0.057 0.0110 0.010L 100 s. w. + 10 NaCl.... -909 1.008 | I.I 491 .052 -0099 .0091 I00 20 Sect 833 1.03 I.23 506 -047 -00016 .0084 100 30 -7615 1.04 I.4 -5I -043 -00837 .0077 100 40 -714 1.06 1.48 +52 -0407 .00785 100 50 .666 1.07 1.6 525 .038 .00739 100 60 625 1.08 Ey, +529 .0356 .00687 100 70 588 I.09 1.85 +533 .034 .0064 100 80 +555 2 A 1.98 537 -0316 .0061 100 90 -526 I.108 735 ¢ +54 -03 .0058 100 100 -50 I.II 2.22 +543 .0285 .0055 100 200 +333 1.15 3-45 -562 .019 .0036 100 700 +125 1.203 9.62 -586 .007 .0013 100 800 angie 1.205 10.85 -587 .006 .0OI2 Table 8 shows the relative rates of nerve-conduction in four rings of Cassiopea xamachana in sea-water to which increments of 0.6 molecular NaCl are added. TABLE 8. ‘ Average Cc f th eee conde omposition of the * es ‘ , F reduced to eater. tion of Ca,| Ring | Ring | Ring 100 pulsa- Mg, and K No.1. | No.2. | No. 3. tions per Ss. Ww. =sea-water. of sea- minute in water. Natural sea-water...... 1.00 100 100 100 100 s.w.+ 10 NaCl.... .909 HOEY iosacoallsoeraoallolbocs 100 20 arelere 833 TO eS oer Allansmoca load arac 100 30 BOGE -7615 TOS Ml \cve-s ehabealiors, Weasavaray slatevalsi Ne 100 40 Sie «714 MOG s|laratecaletetelicta cpesere cil crete initetete 100 50 Free. .666 Coy WSs acacniloaccoanllaogoado 100 60 ene 625 OPP? Wetereretects\|fatere oratavall'aceVerorstars 100 70 Siete -588 TOD aalinvecevetars cil elevetevetenal|ececsueneners 100 80 serene 555 7, ereterevavatelliotwictataterall levenstoteters 100 90 Goan -5206 ee Wend csllgoadcod lecracaar roo 100 sate -50 IOI BS || stotetatete,0¢ PRSLNWN : 9pyNuim sad Suoiqesind Ls “1TDEN L + 4a4 7M P25 I w) Gus ov Oyu. spem PSHPpauw BUS 341 “LIEN L + sapem eas | ur paorid = Gurag Lazhr sanoy % 0! 5 $08 19VCN 2+ sazem vas ur pacsrya Gwaq 4995e Saami, 9 sanoy & i oS a eee ie ee eh ae Syav = Hivd L + sLayemras yard | Ur paovyd Gureq saqyfe saynuru © ANAS ok ' “AIEPMA POS rsnpow vf 2D tbe ran 46 Papers from the Marine Biological Laboratory at Tortugas. their behavior is one of degree rather than of kind. In other words, the motor centers are merely a more sensitive part of the nervous system than is the nerve-net of the subumbrella. For example, a medusa of Cassiopea xamachana (record, fig. 8) with motor centers intact, and pulsating at the rate of 33 per minute in natural sea-water, pulsated very irregularly at the rate of about 8 per minute after having been for 10.5 hours in a solution of 100 volumes of sea-water + 700 volumes of 0.6 molecular NaCl. Here we see the decline in pulsation-rate is greater than in the nervous network without motor centers. If, however, the sea-water had been diluted with 0.9 dextrose, the medusa would have ceased to pulsate when the sea-water had been mixed with about an equal volume of dextrose; hence the effect of the lowering of concentration of Mg, Ca, and K and proportionately increasing the concentration of the sodium ion is to stimulate and maintain pulsation. The initial effect of an increased concentration of NaCl upon the normal medusa is to increase its rate of pulsation. Thus the medusa which pul- sated at the rate of 33 per minute in sea-water, immediately pulsated at the rate of about 116 per minute when placed in I volume of sea-water plus 7 volumes of 0.6 molecular NaCl, although after half an hour its rate was about normal and after 10.5 hours it was only about 8 per minute. Thus the initial effect of the relatively increased concentration of sodium was to stimulate, but later it produced depression in a manner resembling the effect of its close chemical ally, potassium. It is interesting to see that Magowan (1908) and Osterhout (1909) find that sodium and potassium are closely similar both in their toxic and protective action upon plants.1 THE INERT NATURE OF THE MAGNESIUM ION. In order to determine the effects of the magnesium ion, we may add magnesium to sea-water in the form of a 0.057 molecular solution of MgCle dissolved in 0.78 molecular dextrose. This decreases the concentration of the sodium, calcium, and potassium cations, but maintains the normal con- centration of the magnesium ion and also the normal osmotic pressure of the sea-water. The effect must, therefore, be due to the decline in con- centration of the sodium, calcium, and potassium. Table 10 and figure 9 (dotted curve with open circles) show the results of this experiment upon 7 rings of subumbrella tissue of Cassiopea, and it indicates that the magnesium ion, in the concentration found in sea- water, is not in any sense a stimulant; for the rate of nerve-conduction declines in practically the same manner is it does when the sea-water is diluted with distilled water or with 0.9 molecular dextrose. The non-toxic nature of magnesium is clearly shown by the good and rapid recovery of the rate of pulsation when the rings are replaced in natural sea-water. Figure 9 shows the effects of magnesium on the rate of nerve-conduction. The full-line curve shows the effect of adding 0.4 molecular MgCl. to sea- 1 Botanical Gazette, vol. 45, p. 45; vol. 48, p. 98. Papers from the Marine Biological Laboratory at Tortugas. 47 water. The dotted curve applies to 0.057 molecular MgCle. The crosses show the effects of adding 0.9 molecular dextrose to sea-water, and the black circles show the effects of increments of distilled water added to sea-water. Rate. e & 3 6 ave & 9 as a Relative Concentration of Na, Ca, and K. FIG. 9. Table 10 shows the decline in rate of nerve-conduction in 7 subumbrella rings of Cassiopea in sea-water diluted with a solution composed of 0.057 molecular MgCl. + 0.78 molecular dextrose. TABLE 10. Relative rate of nerve-conduction, the rate in normal sea-water Composition of the solution. being roo. ! S. W. =Sea-water. Ring | Ring | Ring | Ring | Ring | Ring | Ring No. 1./ No. 2.| No.3.| No. 4.| No. 5.| No.6.| No. 7. WNattrall sea-watersis'e; ss is/e1e 0s.» sisvereravcinils 100 100 100 100 100 100 100 Haprcte ds: We} 20 Dice We Cla -|-CEXtEOSEL| © crecrsiciai|(srere/eve.s =|[\ererere oreiellinrorere eteter| aie everetara llarctete saci locie cate ellie we sxere 75 25 97 99 or 97 100 92 90 60 40 79 80 82 89 80 79 54 50 ro Po ye ee . laperoer Cho ilScomoac 56 49.5 55 47-5 Rate an hour after being returned to IGEN Ea Watelareicieiclaie alelcieiniaielelcieraie]|(areteverslsie Dol layetets fevers) orelererateye 93 OOW Pilnieinvretore||lereseceteiate 1 Only two rings were tested in this solution and they both gave relative rates of 94. This rate is prob- ably too low to be typical, and therefore I have rejected the results obtained from these two rings and omitted them from the table. By another method, which was tried at Tortugas in 1911 and 1912, the normal osmotic pressure was maintained by adding to the sea-water a 0.4 molecular solution of MgCl, in distilled water. This causes a dilution of TABLE II. _ Propor- Relative Relati tionate con- }Gram-molecular concentration of | concentra- elative | centration NaCl, MgCl, CaCle, and KCI. | tion of the aaa coe of Mg com- Na, Cd, and| #10m of the | pared with K, their | M& their | that of Na, Composition of the solution. concen- = = n. Ss. w. = natural sea-water. concen- | tration in | Ca, and K, tration in | natural sea-| the propor- natural sea-| water _| tion in nat- KCl. water being 1. | Ural sea- being r. water being 1. Batelsteriete ae ete ere'e © : 0.0101 1.000 1.000 I.0 Niviasatatore 44 Ft 2 .009018 .909 1.546 bf sieawete F 4 .0084I1 8333 2 2.4 o\neiels ie . e A .0077 -7615 2.402 3.15 osavee e. . : : . 2.714 3.801 sie a\nele F . x A : 3 4.504 ae wieis F : 3 ci 3.284 5.25 48 the sodium, calcium, and potassium ions and an augmentation in the con- centration of the magnesium, as is shown in table 11. Table 12 gives the decline in rate of pulsation shown by 2 rings and 9 normal meduse of Cassiopea xamachana in sea-water to which increments of 0.4 molecular MgCl, have been added. Papers from the Marine Biological Laboratory at Tortugas. TABLE 12. : Relative rates of pulsation. Composition of the solution. Relative Reape ete Ring |Medusa/Av- of me-| Av. of me- 8. w. =bea-water. Carne K Nee i |awor a2 No.at dusz Nos. 4) duse Nos. |dusz Nos.9, and 5.4 |6,7,and 8.5 Natural sea-water............ I.00 7100 100 8100 100 I00 C.c. S. W.+ 2-5 c.c. MgCl... QZ! © lx ae Rote, ie mesrecene elllereonctoicval| ousveveinys otsrevas tere ayscetere eters 5 40 BOR 5 dle wveracaretellicscteceterevel |acetetsrecens 62 7.5 OSB. 1S hevclausac terallleretenerenaeel oreretece rare leer Relies chereiin | tetermtchatetewen ste 10 -909 107 97.5 51 28 12.5 SOB8. | lhdctesdcareeilietetsvete sil cie sicuevatel llaierecateyalereie\ell Custe eto aici 15 2h Sell (CRSP (AOC tac oG 23 17.5 Ell cers ssyonenelfleyetarctet vel here: enavace el] wuecevele eleneheneil teveteniee echoes 20 833 94 94 23 15 22.5 POLIT.” Millarereresetorelltatcheter eveves|(otel cverminiall (eraect ake teerovavelt sxataratete texters 25 1p SSS sotto Sra trctey ate fac arenes II 27.5 BIB | Clie se cochianvel|ianevere exe «ll ere ate chain inite: e ckatisre-etersll eta aticte terre tate 30 -761 88.5 GEO Wes oder fo) 40 -714 75 FR Wk retake etdve lrgcehactocerearelavetl rats Gueleretar overall tele iectonete tae 50 -666 68 SOW,” 4 ciate tele lfexatene ies -owrerel|| ares avererete; «Jail aetedetsieaerene 60 625 71 72 O fee cee eee eeleeeeeeeeeelecereeeees 115.5 hours after being replaced in sea-water this ring pulsated at the normal rate. 23 hours after being replaced in sea-water this ring pulsated at its normal rate. 8 After being replaced in sea-water this medusa recovered perfectly and pulsated at its normal rate. 4 After being replaced in sea-water they soon recovered and pulsated normally. 5 These meduse recovered completely and pulsated normally after being replaced in sea-water. 6 Upon replacing these medusz in sea-water, they pulsated at normal rates. 7 A partial record of this ring is given in record, fig. 19. 8 See record, fig. II. Table 13, which is derived from table 12, shows the average rates of nerve-conduction in two rings of Cassiopea xamachana in sea-water and in sea-water mixed with 0.4 molecular MgCle. TABLE 13 100 20 100 30 100 40 100 50 100 60 Composition of the solution. Ss. W. =sea-water. eee eer eseee eee ewe eecee a ee a Table 14 shows the average rates of pulsation of 9 meduse of Cassiopea xamachana in sea-water and in sea-water mixed with 0.4 molecular MgCl. | Av. rate per two rings. Av. rate re- duced to min. of the,100 per min.| and K of in normal sea-water. 1 See fig. 9, full-line curve. Dilution of the Na, Ca, the sea- water. Relative concentra- tion of Mg, its concen- tration in sea-water being 1. The rates are reduced to a scale of 100 in natural sea-water. Tables 12 and 13 show the results of experiments upon two pulsating rings in sea-water at 28° to 31° C. to which had been added increments of 0.4 molecular MgCl, dissolved in distilled water. were permitted to pulsate in any new concentration of magnesium for at least I and usually 3 hours before a record of the pulsation-rate was taken, The rings or meduse Papers from the Marine Biological Laboratory at Tortugas. 49 and in the interval the solution was changed several times. In this manner an attempt was made to permit the animal to become thoroughly permeated by a solution before its effect was tested. TABLE 14. Composition of the solution. |Relative rate of Na, Ca, | Mg (concen- 8. W. =sea-water. . land K of the|tration in sea- sea-water, |waterbeing!). Natural sea-water 1.000 100 c.c. s. w. +10 MgCle -909 100 20 -8333 100 8 100 -7615 Tables 13 and 14 show the average conditions; table 13 referring to the rate of nerve-conduction in the rings without motor centers, and table 14 to the pulsation of medusze with motor centers intact. It is evident that any excess of magnesium causes a rapid decline in rate in the perfect medusz. In rings of subumbrella tissue 0.4 molecular MgCl. causes a decline in rate of nerve-conduction which is but little more pronounced than that observed when the sea-water is diluted with distilled water or with 0.9 molecular dextrose. This leads one to conclude that the magnesium ion is inert rather than an active repressor of nerve-conduction. If this be the case all concen- trations of MgCl, from 0.4 molecular downward should behave alike and should exert the same influence as distilled water or 0.9 molecular dextrose. PSI LIINIIIIND DAIS ID IIS SLID SII ILI LL INI DILL LLL DPI III INI INI IID INI IIIS III IIL IIS Sea water 23°C, LOL PLP PPL POL LLL PLP looSea water +10 MacCt, -em. 30! EVV CCCC ~~ PLN PIL PLD LDS LLNS ANA ENN NEN INNA NPD OLD PP 1oo " " + BO ” ‘ rs 29° foo) | " As ee 27:3 1oo “ bad + 44-0 " " “u EM: Se 100 ” ue + 50 " “ “ 27.7 Fic. 10. Figure 9 appears to show that this expectation is largely realized, the effects of 0.057 molecular and of 0.4 molecular MgCl, being closely similar to those of distilled water, and while 0.4 molecular MgCl: may exert a slight specific depressant effect, it is evident that the concentration of Mg found in sea- water (0.057 molecular) is not more depressant than is distilled water. 4 50 Papers from the Marine Biological Laboratory at Tortugas. Sea water PU OE iO se a: NNN COM, btn xO MgCl, 4 om 320°4 BAO Sea WWaALeT* =es ap Ma Cl, Benue loo Sea water + GO Mg on 29°7C Sea watyer ETC EG 11: Thus apart from osmotic effects the degree of concentration of mag- nesium makes but little difference in stupefying marine animals by Tull- berg’s method of adding MgCl. or MgSO, to the sea-water. The com- parative unimportance of the magnesium ion may also account for the fact that while the relative concentrations of the sodium, calcium, and potassium in the blood of mammals remain pretty much as they are in sea-water, the magnesium has declined from 3.7 to 0.4 or to about one-ninth its former con- centration (see ‘‘The fitness of the environment,’ p. 187, by Laurence J. Henderson, New York, 1913). Normal medusz soon cease to pulsate if placed in a solution of 100 c.c. sea-water + 30 c.c. of 0.4 molecular MgCle, and they may remain in this solution for 8 days without further pulsation, the bell, however, slowly disintegrating, due to bacterial action which the weakened concentration of sodium is unable to check. Whenever these magnesiumized medusze are replaced in normal sea-water they at once go into clonic tetanus and then pulsate normally. This initial tetanus resembles that which develops in sea-water lacking magnesium, and is there due to the combination of sodium and calcium unchecked by magnesium. It seems, then, that an excess of magnesium acts as a shield or ‘‘buffer” to prevent the calcium from pro- ducing tetanus, and when this shield is suddenly removed the medusa is unable to withstand the effects of even the normal concentration of calcium.! It will be recalled that Mines (1911 B, p. 185) arrested the pulsation of the heart of a ray by magnesium and then restored pulsation by simply raising 1 S6rensen, 1909, Comptes Rendus Lab. Carlsberg, vol. 8, p. 1, calls attention to the action of various sorts of ‘‘buffer’’ substances in physiological solutions. Papers from the Marine Biological Laboratory at Tortugas. 51 the hydrogen ion concentration of the solution from about 6.5 to 9. Thus the inhibiting effect of the Mg was offset by the addition of an alkali. It appears from my experiments of 1906 that calcium can not produce tetanus unless sodium be present, for in the absence of sodium even pure solutions of 0.4 molecular CaCl, do not produce tetanus in Cassiopea, and if the medusz be stupefied in magnesium they can not be temporarily re- vived by calcium unless sodium be present. These observations are in accord with those of Meltzer and Auer on mammals, wherein they found that the inhibiting tendency of magnesium is offset by a calcium-sodium combination, not by calcium alone. As in all experiments of this series, normal meduse are much more sensitive to the effects of magnesium than are rings. In other words, as is always the case, the motor centers are more readily affected than are the peripheral nerves. EFFECTS OF THE POTASSIUM ION. We may maintain the normal osmotic pressure and increase the con- centration of the potassium ion by adding to the sea-water a 0.64 molecular solution of KCI. Even so small an amount as 3.75 c.c. of this solution added to 100 c.c. of sea-water stops all movement in normal medusz after an initial excitement which soon subsides. The peripheral nerves, on the other hand, are not so readily affected as the motor centers, for a ring still pulsated at about two-thirds its normal rate when placed in 100 c.c. sea- water + 7.5 c.c. of 0.64 molecular KCl. A Rate per Minute 1 2 3 D 4 5 Relative Concentration of K FIG. 12. Table 15 and figure 12 illustrate the results of experiments with potassium. The rates stated in figure 12 are the actual rates assumed by animals after being at least 45 minutes in the solutions mentioned, the volume of solution 52 Papers from the Marine Biological Laboratory at Tortugas. being very great in comparison with the size of the animals and the solutions being changed frequently. It appears that a very slight increase over and above the normal con- centration of the potassium ion is a permanent stimulus. Table 15 shows the relative rates of 3 normal meduse and 1 subumbrella ring of Cassiopea xamachana in sea-water to which increments of a 0.64 molecular solution of KCl have been added. TABLE I5. Relative con- centration of Composition of the solution. potassium ca- | Relative rate of | Relative rate of| Relative rate of tion, its concen-| 3 normal me- | anormal me- |a ring of sub- Ss. w. =Sea-water. tration in nat- dusz.1 dusa.? umbrella tissue. ural sea-water being I. Natiiral Gea=Wwater., crasrefclssls aleisleis ereteleers I.00 3100 4100 5100 TOOC.Cs 85 We-1-0.25) 6:0. ING) So crcrcieste cr svers I.16 LOO)" Allciereraratercteve forera'cieil | sisrevele et oreketetorentre 100 Oi” Eee eae 1.28 LOS ey wid | Se ciarete cts erie score, (orators ona ete ere etats 100 Ost Oe Stisreauscletes 1.44 T2Q, Ws: steraravevatevave\e,evevet| elem e¥eleve avelere ener 100 Legh Aisopoodponad 1.6 ra Nea cioucdonadudinoocemoas cmoae 100 Di2GQual MRR Se aerteterte 1.76 GQ? “TONE cfetctervetave ovale eters Meietate ce asare mre create 100 Tee te os ec iaversiowia ait I.92 Als ws aerate orateveiatecolsl | eletorecottelemaneerae 100 Degen MY. Sie crctereicstuts 2.07 G7 FY cise Giie wea cists [ieiatsvers ere etelorctorete 100 (Pa oy WES Adaasconboe 2.22 ah i a al SACS OOe Aocco leet eonodono oboe 100 BiZGOe dole, Bik cvasbaachonetal 2.37 MN iaiele\ersve are ale'a elec || siatolerate aterekeertane 100 Ci, oe © So cecSrcer 2.53 81 33 IOI 100 BIAS PONE ied cretcee cite 2.69 G7. NN Nskese reer ahe: ciclevelnic’s Veanjege srevotetoenorere 100 oe Oe AUS oto 2.84 OF FP ileivenenssyec niente eo aaios ein 100 B25 1 PPT SS artearccelct oh 2.98 STP) PR Wavectararetetarsyavavatatal |leve evaverevevel eroponstete 100 Cy ie ee Ine SAE Bis 22 = |haretshs\dtare seyoreyujpvexan eetece: nee tetovoxeue tele 100 Bets IT Nand as haere BERT TNs Sai elaie eecardieiareies| siace eva raja ete .e auc ean bie tame cate meee 100 ee a) Mee oot aor ten ARP e yl UT Hevelatetel alee; n% lets elarelepateletaretarectee 85 100 Vice, Wis Bebrt Ceota Maia Se SRT |) Ubeie(ata ete euazatateterel| aterieteve: crete, etnies 62 1 These meduse recovered their normal rates after being returned to sea-water. 2 Stopped in less than one hour. 3 See curve CD, fig. 12. 4 See curve AB, fig. 12, and record, fig. 13. 5 See curve EF, fig. 12. This medusa recovered its normal pulsation when returned to sea-water. Sea water 21,25, WWW 160 s€a water + 25 KCY .64 mm Sie25 2 LONI SEE WM STIDT DEI IE URSA BELA EUR SB AYIA VR LI 3° ' after wards 30.3 AED VV VAIWINIUNIIIININN/NIIINUINAISINIININIINAIIINIINAI RRNA TE after One hour 31° AWWW SA AVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVATAVAVAVAVAVAVAVAVAVAVAV, VAVAVIVAVAVAVAVAVIVAVAVIVAVAVAVAVAVAVAV VOOESea Water —-. oS EL ar A VAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVAVA ” we n 100 " sex water + 7% KCL Se mrnutes PII IIL LLIN ALN WP A A AP ALA SNS DAD SS SDS DD ADD ALD DISD SP DID DP DDD DDMNW DD ADDADANANANANANADDASS a ,teroa nels 312) 31.2 ee ee ee eh ah ata aft terwards Oe Dee eee Shs n ” ” F “d ay 1 hour Fic. 13. Papers from the Marine Biological Laboratory at Tortugas. 53 If we place normal medusz in a solution having the proportions of 100 c.c. sea-water + I c.c. of 0.64 molecular KCl, their rates are increased and maintained thus for more than 24 hours. This solution gives the potassium ion a concentration 1.6 times as great as in natural sea-water. Any further increase in the concentration of the potassium ion causes an initial excite- ment, but this is soon followed by a loss in rate. Rings are not so sensitive and the nerves may maintain a normal rate in a solution having 2.5 times the concentration of the potassium ion in sea-water, but a still further concen- tration causes a decline in rate after temporary, initial excitement. It has long been known that weak concentrations of potassium cause initial excitement followed by loss of rate. It is, however, interesting to see that a very weak concentration may be a permanent stimulant, for it shows that potassium acts in a manner similar to that of sodium, its close associate in Group I of the periodic system. Like potassium, sodium itself produces sustained increase in rate in weak excess, but depression in stronger concentrations. Also in stronger concentrations its initial effect is to stimulate, but depression soon follows. Sodium also is slightly toxic, potassium even more so. Thus in all these respects potassium resembles sodium, but its metallic properties are more decided, as indeed we would expect from its higher atomic weight. EFFECTS OF CALCIUM. It is difficult to determine the effects of the calcium ion per se, for (as is well known) it associates itself or combines with the sodium ion, possibly forming a sodium-potassium ion proteid, and it enables the sodium to counteract the negative effect of magnesium. If Cassiopea be placed in a van’t Hoff’s solution resembling sea-water, but lacking calcium, all movement soon ceases, but pulsation begins to be revived when we add a concentration of calcium equal to 0.4 or 0.5 that found in natural sea-water. Pulsation is most rapid, however, when the concentration of the calcium ion is about 2.5 times that found in natural sea-water. Further additions of calcium cause increasing depression and movements usually cease when the solution contains 7 times the concen- tration of calcium found in sea-water. Such a solution is not toxic, however, for complete recovery takes place almost immediately if the medusz be restored to sea-water. It is interesting to see that half the normal concentration of calcium suffices to restore pulsation, for half the normal concentration of sodium will accomplish the same end, provided magnesium, calcium, and potassium be present in normal concentration. PAPERS CITED. BETHE, A. 1903. Allgemeine Anatomie und Physiologie des Nervensystems. Leipzig, 487 pp. Also Archiv fiir Ges. Physiol., 1908, Bd. 124, p. 541; 1909, Bd. 127, p. 219. Burrows, M. T. 1912. Rhythmical activity of isolated heart muscle-cells in vitro. Science, vol. 36, pp. 90-92. Carson, A. J. 1904. The nervous origin of the heart-beat in Limulus, and the nervous nature of the coordination or conduction in the heart. Amer. Jour. Physiol., vol. 12, pp. 67-74. GOLDFARB, A. J. 1907. Factors in the regeneration of a compound hydroid, Eudendrium ramosum. Jour. Exper. Zool., vol. 4, pp. 317-356. Harvey, E.N. tg11. Effect of different temperatures on the medusa Cassiopea, with special reference to the rate of conduction of the nerve impulse. Papers from Tortugas Laboratory of Carnegie Institution of Washington, Pub. No. 132, vol. 3, pp. 29-39. Hooker, D. 1911. The development and function of voluntary and cardiac muscle in embryos without nerves. Jour. Exper. Zool., vol. 11, pp. 159-196. Loges, J. 1891. Untersuchungen zur physiologischen Morphologie der Thiere. 2. Organ- bildung und Wachsthum. Wiirzburg, 82 pp. Mayer, A.G. 1906. Rhythmical pulsation in Scyphomeduse. Carn. Inst. Pub., No. 47, 62 pp. 1908. Carn. Inst. Pub., No. 102, pp. 113-131. 1912. Ctenophores of the Atlantic coast of North America. Carn. Inst. Pub., No. 162, 58 pp., 17 plates. Mines, G. R. 1911A. On the replacement of calcium in certain neuro-muscular mechan- isms by allied substances. Jour. Physiol. Cambridge, vol. 42, pp. 251-266. 191IB. The relation of the heart-beat to electrolytes and its bearing on com- parative physiology. Journ. Marine Biol. Association, vol. 9, pp. 17I- 190. 1912. Some observations on electrograms of the frog’s heart. Proc. Cambridge Phil. Soc., vol. 16, pp. 615-620. OsTERHOUT, W. J. V. 1909. On the similarity in the behavior of sodium and potassium. Botanical Gazette, vol. 48, pp. 98-104. 1913. The permeability of living cells to salts in pure and balanced solutions. Science, vol. 34, pp. 187-189. Also Plant World, vol. 16, pp. 129-144. Paton, S. 1907. The reactions of the vertebrate embryo to stimulation and the associ- ated changes in the nervous system. Naples Mittheil., Bd. 18, pp. 535- 581. = RoMANES, G. J. 1885. Jelly-fish, star-fish, and sea-urchins, being a research on primitive nervous systems. International Scientific Series, vol. 49. 54 BE THE LAW GOVERNING THE LOSS OF WEIGHT IN STARVING CASSIOPEA. BY ALFRED GOLDSBOROUGH MAYER, Director of Department of Marine Biology of the Carnegie Institution of Washington. One plate and twenty-one figures, 55 THE LAW GOVERNING THE LOSS OF WEIGHT IN STARVING CASSIOPEA. By ALFRED GOLDSBOROUGH MAYER. METHODS. Before being weighed each medusa was dried upon filter paper so as to remove the water adherent to its surface. It was then weighed in 25 c.c. of sea-water on a delicate balance, the weight of the medusa being deter- mined by subtracting the weight of the water and the glass beaker serving as its container. The medusz were starved in sea-water which had been passed through two glass funnels, each containing a double layer of the best quality of Chardin’s filter paper, thus removing the zodplankton upon which Cassiopea feeds. In some of the experiments the sea-water after having been passed through Chardin’s filters was refiltered through porcelain or it was rendered sterile in so far as animal life was concerned by heating it to 71° C. and then restoring the evaporation by distilled water. These additional pre- cautions were, however, found to be unnecessary for the Chardin filters removed all or practically all food from the water. The medusz were always starved in the purest sea-water which was either dipped from the ocean in glass or canvas buckets or pumped into glass reservoir tanks through hard-rubber pipes by means of a hard-rubber pump. In running comparative series care was taken that the sea-water in which each medusa was starved had had the same history; 7. e., was gathered in the same way, stored in the same tank, etc., for every medusa in the series. Whenever possible the medusz were starved side by side in one and the same glass aquarium, but when this was impossible the aquaria were of similar size and form and were placed side by side, so as to be sub- jected to similar environmental changes. EXPERIMENTS. It is a pleasure to acknowledge my indebtedness to Dr. Francis Gano Benedict for important advice and highly appreciated kindness, and to Mr. Joseph C. Bock, of the Nutrition Laboratory of the Carnegie Institution of Washington, for having determined the nitrogen content of the meduse. In common with other scyphomeduse, Cassiopea xamachana consists of a large central mass of gelatinous substance, which is covered with a thin layer of cellular tissue, the weight of the gelatinous substance being very great in proportion to that of the cellular elements of the medusa’s body. 57 58 Papers from the Marine Biological Laboratory at Tortugas. When the animal starves, this gelatinous substance decreases in volume and apparently serves as the chief store of food for the starving animal. If W be the weight of the medusa when starving begins, aW may repre- sent the decline in weight due to loss of body-substance and of water at the end of the first day, so that at the end of the first day the weight of the medusa is W — aW = W(1 — a). Similarly, at the end of the second day the weight is W(1 — a) — aW(1 — a) = W(i — a). Hence the weight y after starving x days is y = W(1 — a)* where a may be called the index of catabolism, the rate of starvation increasing as a increases. This formula accords fairly well with the facts of observation, as will be apparent from an inspection of tables 4 to 27 and the curves accompanying them. At times the observed weights are above and at other times below the theoretical curve, but the general average in table 1 for 30 normal medusz starved in the diffuse daylight of the laboratory is practically coincident with the curve y = W(1 — 0.567)? fora TABLE I. period of 19 days starving in filtered sea-water. Daya | Observed |Catcuiated Table 1 and figure 1 show the average decline in elapsed.) SGiont, | Weight | weight deduced from six series of experiments upon 3 = es 30 medusz of Cassiopea xamachana starved 19 days 7 SS. er in diffuse daylight in doubly filtered sea-water. 3 URE 70-43 The weights are reduced to a scale of 100 grams x2 oe goer | at the beginning of the experiment: Thus, if a 19 32.9 32.9 medusa weighed Io grams at the beginning and 5 grams at the end of a certain number of days, 1Calculated according to the for- P mula y = 100(1 — 0.0567). these weights would appear as 100 and as 50, re- spectively, in the accompanying table: This simple law governing the loss of weight in Cassiopea leads to the conjecture that the chemical constitution of the animal does not change, but that one and the same class of substances serves to maintain the medusa throughout the period of its starvation. It is well known that in verte- brates a very different condition pertains, for the starving animal at first mainly consumes its store of glycogen, oxidizing it into carbohydrate; then the fats are chiefly drawn upon, so that a starving vertebrate contains relatively less fat and more proteid and water than a normal creature. Thus there is a progressive change in the chemical constitution of the body of the animal. This appears not to be the case in Cassiopea, and this supposition receives further support from an analysis of the percentage of nitrogen in the body, which appears to range from 2.121 to 2.997 per cent of the dried weight of the animal, the variation being independent of the number of days the animal has starved. Hence it seems probable that starving neither increases nor decreases the percentage of nitrogen. The results are shown in table 2, wherein it appears that there is no codrdination Law Governing Loss of Weight in Starving Cassiopea. 59 between the number of days the medusa has been starved and the per- centage of nitrogen it contains. The medusz were desiccated at 100° C. at Tortugas, Florida, and the dried substance was then sent to Boston for analysis. TABLE 2. Live weight Dried Per cent of | Per cent of Days jin grams just| weight in | dried to | nitrogen to starved. | before being | grams. living dried killed. weight. weight. oO 109.055 5.401 4.9 2.566 . oO 109.641 5.43 4.9 2.147 3 91.9 4.425 4.8 2.476 4 157.71 6.91 4.3 2.48 6 So0c 5.972 sees 2.997 10 94.380 4.38 4.6 2.121 12 86.72 4.362 5.0 2.226 14 40.288 1.93 4.8 2.451 16 83.232 3.081 4.8 2.543 Av. 4.762 |Av. 2.445 1 Not weighed. Weight y bog) 84) 66.798: 9 10 11 12 13 14 15 1617 18 19 20 Days x ExG: 1: It appears also that the dried weight is about 4.76 per cent! of the living weight, and that this ratio does not change as the animal starves. Thus the living weight bears a practically constant ratio to the weight of the solid substance of the animal, and the law governing the loss of rate may be 10f this 4.7 per cent, 3.5 per cent consists of the salts of sea-water; and thus the actual body-substance of the medusa is only 1.2 per cent, the animal containing 98.8 per cent of water. 60 Papers from the Marine Biological Laboratory at Tortugas. determined either from the one or the other, the living weight being prefer- able on account of its greater magnitude and the consequent lessening of the degree of error in its determination. Hence the reason for the validity of the simple law that governs the loss of weight in the starving jellyfish is that no appreciable chemical change occurs in the composition of its body-substances, and there is no appreciable selective use of different substances at different times during the progress of starvation. Sections show that as a physical result of starvation, the cells become reduced in size, many degenerate and disappear, and the cell boundaries tend to become indistinct; moreover, the gelatinous substance becomes vacuolated and the muscular tonus is largely lost, so that the bell-rim bends upward and inward ina balloon-like manner, as is shown in plate I, figure A, which represents 6 meduse that have been starved 41 days and whose combined weights have declined from 74.95 to 2.78 gm., all being still alive. In fact, throughout these series of experiments no medusa has been lost through natural death, although some of the series of observations upon meduse starved in the dark were terminated on account of the develop- ment of local maceration due possibly to ineffectually checked bacterial activity. Plate 1, figure B, gives an enlarged view of the smallest of these medusz shown in plate 1, figure A, and one may observe its crumpled, shrunken bell, the loss through coalescence of the mouths, and the absorption of the mouth-arm appendages. Indeed, the first organs to disappear are usually the mouths and their appendages, and thus after being starved for about 3 weeks recovery may become impossible through the inability of the animal to capture the zodplankton upon which the medusa feeds. Experi- ments indicate that they do not feed upon diatoms, but in common with the corals, as determined by Vaughan, their food appears to be exclusively of an animal nature, the smaller forms of the zodplankton being taken into the numerous mouths on the mouth-arms. It is known that much of the nannoplankton can pass through the best filters, but one was unable to discover any specimens of the zodplankton in the doubly filtered sea-water used in these experiments. Nevertheless, medusz with their stomachs removed starve more rapidly than do animals with their stomachs intact but regenerating their bell-rims. The medusz which possess stomachs are, however, not called upon to regenerate any of their parts and are thus subjected to less drain upon their resources. The value of a in the formula y = W(1 — a)* gives us a fair measure of the rate of loss of weight. It may be called the ‘‘coefficient of negative metabolism,’’ for as a increases the loss of weight increases in like ratio. In the experiments recorded in this paper a ranges as shown in table 3. It appears from table 3 that Cassiopea thrives best and starves at its slowest rate if maintained in large quantities of sea-water changed only once in each 24 hours, thus resembling the conditions of the semistagnant lagoons in which the medusa lives. Law Governing Loss of Weight in Starving Cassiopea. 61 TABLE 3. Medium in which medusz were starved | Value of a when medusz are | value of a when meduse are Tables re- and conditions of the experiment. starved in diffuse day- starved in dark. ferred to. light of laboratory. Normal medusz each in 6 liters or more | 0.046 to 0.075; av. 0.0567. .| 0.0639 to 0.075; av.0.0695 | 4 to 10 and of doubly filtered sea-water changed 14 once in every 24 hours. Medusz regenerating bell-rim but with | 0.0681 to 0.0689; av. 0.0685 | 0.0511 to 0.0538; av.0.052| 19 to 25 stomach and mouth-arms intact, each in 6 liters or more of doubly filtered water changed once in each 24 hours. Regenerating disks with stomach and | 0.06 to 0.1034; av. 0.0815. .| 0.074 to 0.0937; av.0.0892/ 18 to 25 mouth-arms removed, each in 6 liters or more of doubly filtered water changed once in each 24 hours. Normal medusez in 400 c.c. of doubly | 0.1145..........+.00000- Oi ayateateraleyspsioverete: «eleisiercts Ir and 12 filtered water changed once in each 24 hours. Normal medusz in large quantities of | 0.1015 to 0.15; av. 0.125...| 0.165 to 0.185; av. 0.175.. | 15 and 16 doubly filtered constantly changing sea-water. Normal med nsaiiniGlitersiof sea=wate@r||0;05'5) «isleieic-ereve +'o/<1s/ cele of (o)e'|| siiavetm iain ol/alayeefels/e tysiolelel sfels)s 17 sterilized by heating it to 71° C., and then cooling it to normal tempera- tures, The water being changed once in each 24 hours, The relative rates of starvation both in darkness and in diffuse daylight, shown by medusz under other conditions, are as follows, the results being derived from table 3. Normal meduse in large amounts of stagnant water.............. 1.0 Meduse regenerating their bell rims, but with stomach and mouth- CMB ASM IETLECYG Gh Py D ROROO OID UO Ome Biot 0 Cin Oc RO Ae De Oo Ce 0.96 Meduse-starvine without stomachs: 2). occas. cee sce ee ee ne 27 Normal medusz in small aquaria of 400 c.c. capacity............. rg Normal medusze in constantly running water...............2+05. 2.4 Normal medusz in water sterilized by heating it to 71° C......... 1.0 Although for the first week or more the regenerating meduse generally lose weight more rapidly than do the normal animals the final rate of starvation appears to be practically the same in normal and in regenerating medusz provided their stomachs be present, and this is in accord with the well-known converse fact that the rate of regeneration is independent of the food-supply, starving animals regenerating quite as rapidly as well-fed ones. In this connection it appears from an inspection of the curves accom- panying tables 4 to 27, that if the starving medusz lose but little weight for a day or two, a succeeding period of abnormally rapid loss is apt to ensue and thus the animals tend to follow the theoretical curve y = W(1 — a)’. Cassiopea lives in shallow lagoons of semistagnant water, where it lies fully exposed to the hot sun and, as table 3 shows, it loses weight slowly when kept in quantities of water of about 6 liters per animal in aquaria the water of which is changed only once in each 24 hours. If the water be changed constantly, however, or the medusz be kept in running water, they starve more rapidly than if maintained in stagnant water. It will be recalled that Allen and Nelson! found that if sea-water be heated to 70° C. practically all life within it is killed, and the water remains sterile unless reinfected. 1 Allen, E. J., and E. W. Nelson (1910). Quarterly Journal of Microscopical Science, London, vol. 55, Pp. 361-431; also, Journal of Marine Biology Association, vol. 8. 62 Papers from the Marine Biological Laboratory at Tortugas. Bearing this in mind, the sea-water was doubly filtered and then heated to 71° C. after which it was allowed to cool in sterilized glass vessels pro- tected from dust. Medusz kept in this sterilized water, whether in the dark or in daylight, lost weight no more rapidly than would have been the case had they been starved in sea-water that had been merely passed through four Chardin filter papers without being sterilized by heating. When sea-water is heated to 71° C., however, some of its dissolved air is expelled, and in order fully to compensate for this medusz were starved in a control sea-water solution which had been placed in a vacuum under the receiver of an air pump, thus expelling far more air than was lost by the sea-water which had been heated to 71°C. It was found, however, that the commensal plant-cells of the medusa render the animal comparatively independent of the oxygen of the sea-water, provided an excess of CO: be absent, and thus the medusa starves at practically the same rate in heat- sterilized, air-exhausted, or normal filtered sea-water. In fact, it starves at this same rate in sea-water which has been doubly filtered through the Chardin filters and then refiltered through porcelain to extract bacteria and nannoplankton. Putter’s! “dissolved food’’ of sea-water may, there- fore, still be available as food even after all organisms have been taken out of the sea-water, and it appears that such organisms are not necessary to initiate a process of assimilation. If the stomach be intact, the animals starve less rapidly than if it be removed. Animals whose stomachs have been cut off can regenerate new gastro-vascular centers, but this is rarely done, and it so happened that none of the disks studied by me succeeded in replacing their stomachs, although all made more or less progress in regeneration. In order to determine the effect of the regeneration, we compared the starving rate of these agastric meduse with others having stomachs intact but regenerating their bell-rims, an annulus of the bell being cut off close to the outer edge of the stomach, as is shown in plate 1, figure C. Table 3 and tables 18 to 25 show that disks without stomachs lose weight about 1.25 times as rapidly as do regenerating medusz with stomachs intact. Carbon dioxide plays a conspicuous rdéle in controlling the activities of the medusa, for Harvey? found that during the day the velocity of nerve- conduction is greater than during the night, and this fact is possibly due to the variation in the supply of oxygen conditioned by symbiotic algz in the subumbrella tissue of the medusa. It was found that if we make a weak solution of rosolic acid in sea-water and divide it into two equal volumes in two separate containers and then cut a medusa into halves and place each half in a volume of the rosolic-acid solution, it will be observed that if one half of the medusa be kept in diffuse daylight and the other half in darkness, the rosolic acid surrounding the 1 The most complete presentation of Putter’ s views is given in his recent work ‘“‘Die Ernahrung der Wasserthiere und der Stoffansalt der gewasser’’ (Jena, 1910). 2 Harvey, E. Newton. Carnegie Institution of Weahinntal, Year Book No. 10, I9II, p. I31. Law Governing Loss of Weight in Starving Cassiopea. 63 darkened half of the medusa loses color and fades rapidly, while the fluid surrounding the half of the medusa which is in daylight fades more slowly. This and the lime-water test in magnesium-free sea-water show that the medusa gives off COs, and that in diffuse daylight less COs is given off than in darkness; the difference being apparently due to photosynthesis on the part of the commensal plant-cells or zodxanthellz which infest the gela- tinuous substance immediately under the external cell-layers of the medusa. Tables 3, 11, and 12 show that when the medusz are confined in a small volume of water they starve more rapidly than if kept in large aquaria. They also pulsate slower when confined in small aquaria, and a slight excess of CO, brings them to rest, but even when not moving they lose weight far more rapidly than if pulsating in normal sea-water, the toxic effect of CO, being very noticeable. If a few bubbles of CO, be passed through sea-water the medusz cease to pulsate; plate 1, figure A, is derived from a photograph of animals treated in this manner. Upon replacing the medusze in normal sea-water pulsation is immediately renewed, no permanent toxic effects appearing until they have been subjected to the CO, for about 24 hours. The influence of CO: is probably counteracted by the well-known photo- synthesis due to the commensal plant-cells or zodxanthelle. When the medusz are starved in diffuse daylight these plant-cells become more and more crowded as the medusa declines in bulk, and after several days of starving they begin to escape early in the morning, when the swarm- ing zodxanthellz penetrate the ectoderm and pass out into the ocean, where they swim actively, and finally settle upon those lighted parts of the cylindrical glass aquarium which are away from the source of light, the swarming cells being entrapped in the lighted meniscus of the aquarium in accordance with the Jennings reaction. When, on the other hand, the medusz are starved in the dark, the algal cells are rendered inactive and do not escape into the water. Meduse starved in the light become dark brown in color after the first 10 days, for the plant-cells remain alive, al- though apparently unhealthy and probably inactive. When the medusz are starved in darkness many of the plant-cells die and dissolve in situ in the gelatinous substance. Not all of these cells die, however, for a few still remain alive and apparently healthy after being more than a month in the dark, and if the medusa be replaced in the light these cells propagate and again provide the animal with a normal supply of zooxanthellz. Sections show that when the medusz starve in the dark the gelatinous substance becomes vacuolated and edemic, due possibly to the unreduced excess of CO; in the tissues. Thus the meduse when starved in the dark appear to suffer from lack of oxygen, being in effect partially suffocated. Yet even when the plant-cells have been greatly reduced through 26 days of starvation in the dark, the medusz can still pulsate for 2 or more hours at nearly a normal rate in sea-water that has been boiled to expel its oxygen. It at one time seemed that these plant-cells played 64 Papers from the Marine Biological Laboratory at Tortugas. a significant rdle in maintaining the weight of the medusa when starved in daylight, but further experiments have shown that this is not the case, for the medusze sometimes starve at a more rapid rate in diffuse daylight than in the dark, although the reverse is usually the case. It is not the object of this paper to present an exhaustive review of all previous researches upon the metabolism of inanition in invertebrates, because previous studies do not touch upon the chief subject of this study— the law of the decline of weight. A brief mention of certain important sources of general information should, however, be presented. The earliest work appears to be that of Dumas and Milne-Edwards (1820, Annales Chim. et Physique, tome 14; and also in the Comptes Rendus, Paris, 1843, tome 17, p. 531) upon the metabolism of bees. Peligot (1865, Comptes Rendus, tome 61, p. 866; and 1867, Annales Chim.et Physique, tome 12, p.445) studied silkworms in the same connection. Especially worthy of mention is the work of Slowtzoff! (1904), who found that when starved the vineyard slug consumed about 25.74 per cent of its total weight and about 28.41 per cent of its store of energy. The relation between the organic and inorganic substance remains the same in the normal and in the starved animals. The loss of vital substances consists of fat, carbohydrates, and water in the proportions of carbohydrate 93.98 per cent, fat 78.51 per cent, and water about 30.02 per cent, while the loss of albumen was computed to be 23.70 per cent. Asa result of starvation the soft parts of the animal acquire an increase of insoluble salts to about 35.9 per cent. During the period of starvation the slugs consume 4.84 calories of energy per kilogram of their weight per 24 hours. The phos- phoric albumens were largely retained, only about 19 per cent of these being parted with. In 1901 the same author? starved the dung beetle, Geotrypes stercorarius and found that in from 5 to 11 days it lost about 21 per cent of its original weight, the loss appearing to be chiefly water and fat. Schulz? starved Planaria lactea and states that after 6 months without food some of the worms were reduced to about one-tenth their original size. There were, however, considerable individual differences, some of the animals being three times as large as the smallest after all had starved for 6 months. Schulz made a histological study of the starved worms and found that some of the cells were killed as a result of the starvation, while others had degenerated or become reduced in size. The epithelium tends to fuse and its cells to degenerate into a syncytium, very much as we have observed in Cassiopea. Perkins‘ observed that the cells of the attached hydroid larve of Gont- onemus murbachi degenerated in aquaria, where presumably they were 1 Slowtzoff, B. (1904). Beitrage zur vergleichenden Physiologie der Hungerstoffwechsels, Mittheil. 2, Der Hungerstofiwechsel der Weimbergschnecke (Hofmeister’s Beitrage). 2 Slowtzoff, B. (1909). Idem., pb 5, Der Hiinmpetstoireechsel der Mistkafer (Geolrypes stercorarius). Biochemischer Zeitschrift, Bd. Io, 04. 8 Schulz, E. (1904). Ueber Parductioden, , Ueber Hungerscheinungen bei Planaria lactea. Archiv fiir Entwicklungsmechanik der Organismen, Bd. 4 PP. 555-577, Laf. 34, 11 fign. 4 Perkins, H. F. (1902). Proc. Acad. Natural Sci., Philadelphia, vol. 54, p. 765;also: 1902. Biol. Bulletin Woods Hole, vol. 3, Dp. 172-180. Law Governing Loss of Weight in Starving Cassiopea. 65 unfed or poorly nourished, and that the larvae became amceboid and changed into creeping plasmodia without cell boundaries. Thacher’ also observed a process of degeneration in hydroid colonies of Eudendrium, Pennaria, and Campanularia kept in aquaria. The polypites of these forms are absorbed, the degenerating cells of the entoderm and finally the ectoderm of the hydranths being turned into the digestive tract of the stem. By contrast with the few studies upon invertebrates many researches have been conducted upon the metabolism of matter and energy in the body of starving vertebrates. To properly review this literature would carry us too far afield and indeed a text-book would be required for the purpose, but the literature of the subject, in so far as it relates to man and the higher vertebrates, has been exhaustively reviewed by Atwater and Langworthy (1897) in their ‘‘ Digest of Metabolism Experiments” (434 pp., Bulletin No. 45, United States Department of Agriculture, Office of Experiment Sta- tions). One should also consult such later works upon vertebrates as At- water and Benedict (1903), Benedict (1907), Weber (1902), Ergebnisse der Physiologie; Schaefer (1898), Text Book of Physiology, vol. 1, p. 891; Manca (1902), Pembrey and Spriggs (1904), Abderhalden, Bergell, and Dorpinghaus (1904), Hatai (1904), Cathcart and Fawsitt (1907), and Maignon (1908), and others referred to by these authors, the list of papers cited by Pembrey and Spriggs in the Journal of Physiology (Cambridge, vol. 31, pp. 320-345), and by Benedict (1907), being especially enlightening. EXPLANATION OF THE TABLES. In the figures accompanying tables 4 to 25, when the meduse are starved in darkness their observed weights are indicated by black dots; when starved in the diffuse light of the laboratory their weights are shown by circles. Tables 4 to 25 refer to Cassiopea xamachana and table 26 to the allied species C. frondosa, which follows a law in its starvation similar to that of C. xamachana. Table 4 shows the decline in weight of 6 normal meduse of Cassiopea xamachana starved 41 days in filtered sea-water, in the diffuse light of the laboratory, at temperatures ranging from 27.3° to 30.1° C., from June 18 to July 29, 1911, at Tortugas, Florida. The medusz were kept in a 6-liter glass aquarium, the water being changed once in every 24 hours. Compare this with table 10 to see the effect of darkness upon medusz subjected to conditions as nearly as possible identical with the medusz starved in day- light. The curve y = 75(1 — 0.075)” does not accord well with the facts, but the curve represented by y = 75(1 — 0.058)? accords fairly well with the. observed decline in weight up to the sixteenth day, after which it departs more and more widely from the observed weights and is unsatis- factory. (See table 5.) 1 Thacher, H. F. (1903). Biol. Bulletin Woods Hole, vol. 4, pp. 96-98. 66 Papers from the Marine Biological Laboratory at Tortugas. TABLE 4. Observed see weight in grams, °o 74.95 I 73.36 2 68.315 3 65.345 4 62.075 5 59.065 6 55-37 7 51.62 8 47.955 9 45.16 Io 42.69 II 40.225 12 36.28 13 33.825 14 31.185 I5 29.11 16 28.065 17 27.77 18 24.78 19 23.15 20 21.44 21 18.95 22 18.18 23 17.01 15.67 Calculated weight! 75 64.13 54.9 1 According to formula ¥Y = 75(I — 0.075). TABLE 5. Days _ | Observed) Calculated elapsed. weight. 75 55.65 41.7 30.6 22.72 16.8 12.45 9.22 6.87 1 According to formula y = 75(I — 0.058)”. 33 34 35 36 37 38 39 40 41 6 Cassiopea medusae starved 41 days in diffuse day-light ° 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 1051) 12> 137 14 15 ; ei ae: a ee S Le © 4 Sous Diagram illustrating Table 4. Days x Law Governing Loss of Weight in Starving Cassiopea. 67 Table 6 shows the decline in weight of Io normal medusz of Cassiopea xamachana starved 25 days in filtered sea-water in the diffuse light of the laboratory from June 22 to July 17, I910, at Tortugas, Florida. The meduse were kept in a 6-liter glass aquarium, the water being changed once in each 24 hours. Grammes y 40 30 20 weight, in grams. 130.75 128.06 119.58 115.34 108.907 106.58 103.55 97.60 97.22 1 According to formula y = 131(1 — 0.046)”, Observed | Calculated weight.! 131 124.97 II9.21 113.71 108.6 103.49 98.77 94.19 89.87 8 9 TABLE 6. Observed weight in grams. 92.14 88.71 86.24 79.54 72.30 69.27 66.81 61.24 Calculated weight.! 10 11 12 13 14 15 16 17 18 19 20 21 22 23 2 2 Days x Diagram illustrating Table 6. Days | Observed | Calculated] elapsed. | Weight in | “Weight.1 grams. ee 2No weighing. 10 medusae of Cassiopea starved 25 days in diffuse day-light Table 7 shows the decline in weight of 11 normal medusz of Cassiopea xamachana starved 22 days in filtered sea-water in the diffuse light of the 68 Papers from the Marine Biological Laboratory at Tortugas. laboratory, at temperatures ranging from 27.1° to 30.1° C., from June 19 to July 11, 1911, at Tortugas, Florida. The medusz were kept in a 6-liter glass aquarium, the water being changed once in each 24 hours. TABLE 7. eo ee Days Observed | Calculated.| Days | Observed | Calculated || Days | Observed | Calculated weight in . 1 weight in A 1 weight in A elapsed. aan weight. elapsed. pine weight. elapsed. pena! weight.! ° 496.9 497 8 328 321.56 16 223.2 208.24 I 457 470.66 9 315 304.66 17 206.9 197.31 2 432.1 445.8 10 306 288.86 18 191.2 186.87 3 399.69 421.95 II 290.99 273.35 19 179.2 176.93 4 385.1 399.59 Ta 278.95 258.04 20 172.2 167.49 5 369.3 377.22 13 266.3 245.02 | 21 161.63 158.54 6 365.5 358.83 14 249.5 232.10 22 150.09 150.25 7 345 339.45 15 240.8 BEOQIOT > |AIl'e. Fieveitenararn/el| (ae keteteyevevava tell ve rorotetehenerete 1 According to formula y = 497(1 — 0.0520)”. EMARKS.—Free-swimming algal cells were given off from these meduse every morning soon after day- light for the first week. These plant-cells collected on the side of the cylindrical glass aquarium which was away from the light. 11 medusae of Cassiopea starved 22 days in diffuse. day-light Grammes y Om Meo SedAeers. 1G) Tees” Snel Deell 912,913,142) Ibe Gel? 918p 919) cD ol ae Days x Diagram illustrating Table 7. Table 8 shows the decline in weight of a single normal medusa of Cassi- opea xamachana starved 17 days in the diffuse daylight of the laboratory in filtered sed-water at temperatures ranging from 26.5° to 30.1° C., from June 22 to July 9, 1911, at Tortugas, Florida. The medusa was kept in a 6-liter glass aquarium, the water being changed once in each 24 hours. TABLE 8. Days | Observed | Calculated || Days | Observed | Calculated elapsed.| Weight in | “Weight ||elapsed.| weight in | weight. grams. grams. (0) 261.05 262 bike) 133.1 132.31 I 248.5 244.7 II 128 123.66 2 243.9 288.46 I2 I21.95 115.28 3 231.5 213.53 - 13 Ii2 108.21 4 215.8 199.38 I4 I1I0.2 102.61 5 192.75 186.28 15 101.8 94.06 6 185.7 173-97 16 93.2 87.77 7 166.7 162.44 17 81.9 82 8 158.3 I51.7 9 144.7 141.74 1 According to the formula y = 262(I — 0.66)”. Law Governing Loss of Weight in Starving Cassiopea. 69 300 Fe) 250 A medusa of Cassiopea starved 17 days in diffuse day-light 150 9 Grammes y 100 ¢ 50 9 OPP 2063 Mpe5 6) ee Se 9) 10nd 12 ise 14) 15, los Days x Diagram illustrating Table 8. Table 9 shows the decline in weight of two normal medusz of Cassiopea xamachana starved 25 days in diffuse daylight, in filtered sea-water, at 27.5° to 30° C., from June 28 to July 23, 1912, at Tortugas, Florida. The meduse were placed in a glass aquarium holding 6 liters, the water being changed once in each 24 hours. TABLE 9. 55 Medusa A, Medusa B. Days zs Observed Observed 50 elapsed. | Seight in | Weight. | weightin | Weight.? grams. grams. 45 re) 42.62 42.62 54.9 54.9 4 31.73 32.00 36.48 43.2 7 26.04 25.76 34.94 36.18 10 20.52 20.80 29.08 30.19 40 13 16.48 16.79 25.51 25.25 17 13.25 12.57 19.41 19.87 21 10.19 9.46 14.99 15.70 95 25 7 i) 12.35 12.35 1 Calculated from formula y= 42.62(1 — 0.0692)”. 2 Calculated from the formula y = 54.9(1 — 0.0579)”. Weight y 10 11 12 13 14 15 1617 18 19 20 21 22 23 24 25 Days x Diagram illustrating Table 9. 70 Papers from the Marine Biological Laboratory at Tortugas. Table 10 shows the decline in weight of 6 normal medusz of Cassiopea xamachana starved 26 days in filtered sea-water in the dark at temperatures ranging from 27.3° to 30.1° C., from June 18 to July 14, 1911, at Tortugas, Florida.! The medusz were placed in a 6-liter glass aquarium, the water being changed once in each 24 hours. The temperature conditions were similar to those maintained for the medusze whose rates of decline are shown in table 4. TABLE I0. Days Observed | Calculated Days | Observed | Calculated ays teed ved | Calculated elapsed. | weightin | “Weight.i || elapsed. | Weightin | “weight.t . | weight in grams. grams. grams. 85.6 9 41.03 42.47 23.24 79.18 10 40.96 39.20 20.965 73.27 II 39.4 36.20 18.85 67.71 12 33-535 33.56 17.56 62.66 I3 28.51 31.07 15.31 57.87 14 27.23 28.76 14.38 53.59 Is 27.12 26.54 13.07 49.56 16 26.49 24.57 10.78 45.88 77 24.87 22.77 8.99 ° I 2 3 4 5 6 7 8 1 According to the formula y = 85.6(1 — 0.075). REMARKS.—These 6 meduse were captured in the moat of Fort Jefferson, Tortugas, on June 16 and weighed for the first time on June 18, 1911. The smallest was about 25 mm. wide on June 18. On June 22 the smallest was only 8.5 mm. wide, and by June 25 and 26 the medusz had begun to macerate. 90 re) 80 70 6 medusae starved 26 > 3 E 50 ra) 10 11 12 18 14 15 16 17 18 19 2 21 22 23 2A 25 2G Days x a eRe Gy 2 er ae Diagram illustrating Table 10. 1 For the first 14 days the medus# remained in complete darkness, not being exposed for an instant to the light. On the fourteenth day they were observed in diffuse light, and it appeared that they were all blue- translucent, not greenish gray, as are the normal meduse. The algal cells were then greatly reduced in number, The meduse were then replaced in the dark as soon as possible, and remained in darkness until the end of the experiment, excepting that from the fourteenth to the twenty-sixth day they were weighed in the diffuse light of the laboratory but were replaced in the dark room immediately after each weighing. Law Governing Loss of Weight in Starving Cassiopea. 71 Table 11 shows the decline in weight of a single medusa of Cassiopea xamachana starved 19 days in filtered sea-water, in the dark, at temperatures ranging between 27.3° and 30.1° C., from June 30 to July 19, I911, at Tortugas, Florida. During the first 10 days this medusa was placed in A medusa of Cassiopea starved 19 days in the dark Grammes y OL (2) 3h 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Days x Diagram illustrating Table 11. 400 c.c. of sea-water which was changed once each day. After this it was placed in a glass vessel holding about 6 liters and the water was changed constantly by mechanical means, thus giving an abundant supply of fresh sea-water. Compare this with table 12 which shows the starving rate of a medusa maintained in daylight. TABLE II. | | Days re a ae Days ey a Sa pee Bes Calculated ° i 1 elapsed. eiccaas weight. elapsed. aa weight elapsed. aaud. weight. te) 44.51 44.5 7 23.66 21.27 14 10.45 10.19 I 36.69 40.05 8 20.49 19.13 15 9.45 9.17 2 31.44 36.24 9 18.99 17.24 16 7.70 8.25 3 30.59 32.44 bas) 16.05 15.53 17 6.30 7.39 4 28.825 29.19 II 15.03 13.97 18 5.23 6.68 5 24.485 26.25 I2 I4.0I 12.57 19 4.23 6 23.77 23.63 13 13.12 11.303 ae a eee 1 According to the formula y = 44.5(1 — 0.1)?. REMARKS.—On July 19 maceration began and the experiment was discontinued. Table 12 shows the decline in weight of a single medusa of Cassiopea xamachana starved 20 days in the diffuse daylight of the laboratory, and then for I5 more days in the dark, in filtered sea-water at temperatures ranging between 27.3° and 30.1° C., from June 30 to August 4, IgII, at Tortugas, Florida. During the first 10 days this medusa was placed in 400 ‘c.c. of sea-water which was changed once each day. After this it was placed in a glass vessel holding about 6 liters and the water was changed constantly by mechanical means, thus giving an abundant supply of sea- water. The conditions of the experiment are comparable with those shown on table 11, excepting that table 12 applies to a medusa starved in daylight, while table 11 shows the decline in weight of one starved in darkness. v2 Papers from the Marine Biological Laboratory at Tortugas. Immediately after the twentieth weighing the medusa was placed in the dark and thenceforth maintained in darkness until the end of theexperiment. The results are given in table 13. When the records ceased, maceration had begun. For the first 20 days the medusa was in diffuse day-light, but from the 20th to the 35th day it was kept 3.2(1~9.15)0 s ¥ 3 = ic] wD oD + z 3 ges & Fa 2 re as a 3 & S3e2 5 S 3 a TABLE 12. 3 2 S £ S$ > < A Days Observed | Calculated : elapsed. | Weight in | “weight! grams. ° 41.32 41.3 I 33-92 36.55 2 30.52 32.38 3 20.18 28.66 4 26.66 25.4 5 25.09 22.47 6 24.03 19.01 a 22.03 17.04 8 20.77 15.61 9 17.63 13.84 pas) 16.41 12.22 II 13.80 10.82 12 12.02 9.58 13 9.86 8.51 14 9.66 7.52 15 8.03 6.65 1 Ae 16 7-34 5-91 17 5.64 | 5.20 face 18 4.875 4.63 } 19 4.27 4.09 = 20 3.66 3.63 Y= 1 According to formula had y = 41.3(1 — 0.1145)”. re ~~ —) S 8 3 8 & ES S we & souuely Diagram illustrating Tables 12 and 13. TABLE 13. | PPh fess - le ese Sees Play Observed | Calculated Peace me Observed | Calculated A ; s ‘ ; s A 7 Gnvieces. weight. weight. Aatenese weight. weight. | daxkuean. weight. weight. 21 I 3.2 3.2 26 6 1.542 1.42 | SE lecr 0.68 0.64 22 2 2.83 opie 27 if 1.205 1.206 32 | x2 0.54 0.54 23 3 2.53 2.31 28 8 I.0r 1.02 33 13 0.54 0.45 4 1.98 1.96 20 9 0.87 0.84 34 | 14 0.45 0.39 5 1.52 1.64 30) | r0 0.75 0.74 35 ulets 0.29 0.34 1 According to the formula y = 3.2(1 — 0.15)*. 2An error in weighing. Law Governing Loss of Weight in Starving Cassiopea. 73 Table 14 shows the decline in weight of a normal medusa of Cassiopea xamachana starved 25 days in the dark, in filtered sea-water, at 27.5° to 30° C., from June 28 to July 23, 1912, at Tortugas, Florida. The medusa was placed in a glass aquarium holding 6 liters, the water being changed once in each 24 hours. TABLE 14. Observed | Observed Days jweight of} Calculated. Days j|weight of| Calculated 30 § elapsed. |medusain|} weight.1 elapsed. |medusain| weight.1 grams, grams. 25 o 32.77 32.77 13 15.09 13.93 4 21.58 2hl7 7 12.52 10.65 a 19.54 20.61 21 9.04 8.12 10 17.34 16.91 25 6.28 6.28 20 1 According to formula y = 32.77(1 — 0.0639)*. 15 4 Weight y 10 ¢ 0 123 4 5 6 7 8 9 1011 12 13 14 15 1617 18 19 20 21 22 23 24 25 Days x Diagram illustrating Table 14. Table 15 shows the decline in weight of two meduse of Cassiopea xamachana starved in diffuse daylight in constantly changing, filtered sea- water at temperatures ranging from 27.3° to 30.1° C., at Tortugas, Florida, from July 10 to August 4, 1911. Compare this with table 16. These medusz were starved side by side in the same 6-liter glass vessel and were subjected to the same changes of temperature, water, light, etc. Being in running water, they lost weight more rapidly than one would expect had they been in semi-stagnant water. TABLE I5. Medusa A. Medusa B. Medusa A. Medusa B. Days Days elapsed.| Ob- | Expected | Observed | Expected elapsed.| _Ob-| | Expected | Observed | Expected served | ‘weight.! | weight. | weight.? served | weight.! | weight. | weight.? weight. weight. —— ee was — oO 40.58 40.6 37.16 afer! 13 10.26 10.11 8.20 9.18 I 40.26 36.48 33-5 33-37 14 9.34 9.05 7.607 8.18 2 37.08 | 32.76 26.88 20.05 15 8.56 8.16 7.38 7.39 3 31.6 29.43 23.53 26.90 16 8.00 7.31 7.00 6.65 4 27.03 | 26.47 20.03 24.15 17 7.21 6.58 6.515 5.95 5 23.475 23-79 18.35 21.70 || ‘~=18 6.09 5-93 5.39 | » §.35 6 19.79 21.36 16.53 19.47 || 19 5.41 5.32 4-73 | 4.79 7 18.15 19.20 13.78 £7.50: 5} 20 4.73 4.75 3.99 4.31 8 17.41 17.25 13.48 15.72 21 4.17 P. | 0M Beta ee 9 14.17 15.47 12.43 14.12 22 3.09 Fe ho eM | gO PECICE TE PREC CEE 10 12.79 13.93 II.19 E2167 al 23 3.48 BUA); P paon < cle aa) cis aietherercahd II II.92 12.50 10.30 11.37 24 3.08 S500 Ad ho) oe ORCI occ oer 12 10.96 11.25 8.80 10.22 | 25 2.52 2.79 MERE ands ER eed PANG Co eRe 1 Calculated from formula y = 40.6(1 — 0.1015)”. 2 Calculated from formula y = 37.16(1 — 0.102)*. 74 Papers from the Marine Biological Laboratory at Tortugas. 2 medusae of Cassiopea starved in running water in diffuse day-light i) 1 Ls) oo 4 > ao n % 8 39) 10) 41 12) 13) 145 46 17 18 19 20 21 22 23 24 25 Days x Diagram illustrating Table 15. Table 16 shows the decline in weight of two meduse of Cassiopea xamachana starved 9 days in the dark in constantly changing, filtered sea- water at temperatures ranging from 27.3° to 30.1° C., at Tortugas, Florida, from July 10 to 19, 1911. Compare this with table 15. TABLE 16. Medusa C. Medusa D. 50 8 Days 2 medusae of Cassiopea starved elapsed.| Ob-_ | Expected | Observed | Expected in the dark in running water serve weight.1 weight. weight.2 weight. (0) Ey. | laneteks, ofaiavevate 36.37 36.37 I 351.78 51I 34.17 29.64 > 2 44.09 42.58 24.88 24.15 2 3 38.86 35-55 20.83 19.68 = 4 30.5 209.68 16.88 16.04 = 5 28.28 24.79 15.33 13.06 FH 6 24.13 20.71 I4.19 10.66 oS 7 19.25 17.29 9.62 8.69 8 17.61 14.43 7.49 7.06 9 I1.97 12.03 5.79 5.78 1 According to formula y = 51(1 — 0.165)7. 2 According to formula y = 36.37(1 — 0.185)”. 3 Medusa C did not begin to lose weight until after it had been confined 24 hours in the dark. The com- mensal plant-cells seem to be of no aid to the medusa Days x when in pure, constantly changing sea-water. The me- x s dusz live, in nature, only in semi-stagnant salt lagoons. Diagram illustrating Table 16. TABLE 17. Medusa A. Medusa B. Medusa C. Days Weights in Weights in Weights in elapsed. | water which | Calculated water which Calculated normal fil- | Calculated had been tested) weights.1 had been par-| weights.? tered sea- weights.® to 70° C. tially exhausted water. of air. ° 49.567 50 | 40.45 aertrs 35.83 35.8 2 46.61 44.62 40.69 40.5 29.04 31.82 4 45.13 39.9 | 438.57 36.08 28.44 28.32 6 38.39 35.6 32.18 32.19 24.61 25.16 8 33.09 31.8 31.64 28.7 22.04 22.37 Io 29.7 28.3 29.28 25.55 20.93 19.05 I2 26.19 25.25 25.73 22.76 19.18 17.61 I4 23.7 22.7 21.82 20.29 16.77 15.65 16 19.83 20.2 | 18.28 18.1 13.93 13.99 18 18.18 17.8 16.51 16.51 I2.45 12.45 20 15.7 15.4 14.36 14.75 Teer 11.06 1 According to the formula y = 50(rI — 0.0553)”. 3 According to the formula y = 35.8 (I — 0.057)”. 2 According to the formula y = 40.5(I — 0.0559)”. 4 Probably an error. Law Governing Loss of Weight in Starving Cassiopea. 75 Table 17 shows the decline in weight of three normal specimens of Cassi- opea xamachana starved for 20 days at Tortugas, Florida, in filtered sea- water. One medusa (A) was maintained in water which had been heated to 71° C. and then cooled to normal temperatures. Another (B) was kept in sea-water from which most of the dissolved air had been exhausted by means of an air pump. The third medusa (C) was kept in normal filtered sea-water. 50 40 30 20 10 2 C 5 0 a 45 6 8 10 I2 14 16 18 20 Diagram illustrating Table 17. Table 18 shows the decline in weight of three half disks of Cassiopea xamachana each starved in 6 liters of water in diffuse daylight, together with TABLE 18. Disk A. Disk B. Disk C. : Weights of ecru ot half-disk Half of disk starved insea-| Weights _||Starved in sea-| Weight starved in Weight water which | calculated.! water which | calculated. || normal fil- | calculated. had been heated had been par- tered sea- to 71° C. tially exhausted water. { of air. 23.05 23.05 20.035 20.03 18.535 18.5 19.46 20.33 17.07 17.68 15.68 16.27 17.815 17.96 14.72 15.62 13.83 14.27 15.55 15.89 12.33 13.82 II.84 12.52 13.82 I4 11.38 12.2 II.04 II 12.16 12.4 10.17 10.8 9.04 9.65 10.63 Err 9.31 9.66 vier) 8.57 9.68 9.69 8.39 ; 8.42 7.21 7.43 8.79 8.55 7.63 7.44 6.4 6.52 7.01 7.50 6.84 6.58 5.65 5.72 6.71 6.71 6.2 5.82 5.03 5.03 1 According to formula y = 23(1 — 0.06)”. 2 According to formula y = 20(1 — 0.06)”. 8 According to formula y = 18.5(1 — 0.063)*. 76 Papers from the Marine Biological Laboratory at Tortugas. fe) re oe 6 8 10 l2 14 16 18 20 Diagram illustrating Table 18. the normal medusz described on table 17. The stomach and mouth-arms were cut off and the disk cut into halves. One half (Disk A) was starved in doubly filtered sea-water which had been heated to 71° C.; the other half of the same disk (Disk B) was starved in doubly filtered sea-water which had been placed in a vacuum to expel most of its dissolved air. The third half disk (Disk C) was taken from another medusa and was starved in normal doubly filtered sea-water. Table 19 shows the decline in weight of two regenerating specimens of Cassiopea xamachana starved 19 days in diffuse daylight, in filtered sea-water ; TABLE IQ. 40 ¢ Observed 2 Weight of re-| Weight of Days |weight of re-|Weight | generating regenerating 35 4 elapsed.) generating {of disk.1| medusa in normal disk in grams. grams. medusa.? (o) 25.48 25.48 41.87 41.87 30 ¢ 2 20.15 20.48 33.98 36.2 6 9.66 13.24 24.30 27.3 9 8 9.55 18.97 22.02 I I2 6.52 6.88 14.23 17.79 25 9 15 5.06 4.97 13.87 14.36 19 4.21 3.25 10.81 10.81 20°¢ Weight y “ey 1 Calculated according to formula y = 25.48(I — 0.1034). Ye 2 Calculated according to formula y = 41.87(1 — 0.0680)*. 15 9 10 9 0123 4 56 7 8 9 1011 12 13 14 1516 17 18 19 Days x Diagram illustrating Table 19. Law Governing Loss of Weight in Starving Cassiopea. 79 at 27.5° to 30° C., from June 26 to July 15, 1912, at Tortugas, Florida. One of these specimens was a disk from which the stomach and mouth-arms had been cut off, and the other was a medusa with stomach and mouth- arms intact but with its bell-rim cut off. The two were maintained in one and the same 6-liter glass aquarium jar throughout the experiment, and were thus subjected to identical environmental conditions, The water was changed once in each 24 hours. Table 20 shows the decline in weight of two re- generating species of Cas- siopea xamachana starved 19 days in diffuse daylight, in. filtered sea-water at 27.5° to 30°C., from June 26 to, July. °15,'° 1922). at Tortugas, Florida. One of these specimens was a disk from which the stom- ach and mouth-arms had been removed, and the other specimen was a me- dusa with stomach and mouth-arms intact but with the bell-rim removed. The two were maintained in one and the same 6-liter glass aquarium throughout the experiment, and were thus subjected to identical environmental conditions. The water was changed once in every 24 hours. Weight y 0123 4 56 6 7 8 9 1011 12 18 14 15 16 17 18 19 Days x Diagram illustrating Table 20. TABLE 20. Observed ‘ Days setae sof = F disk. weight of re-/ Weight of ns elapsed.) generating |Weisht of disk. pees antag : idisk in grams. medusa i | grams. | ee APR Sey em | ee HF ._e fe) 35.08 35.08 56.75 56.75 2 33.28 29.11 49.11 49.26 6 14.86 20.06 31.76 37-17 9 II.04 15.19 23.63 30.07 12 8.08 II.51 20.43 24.34 15 7.9 8.7 18.45 19.69 19 6 6 14.88 14.88 1 Calculated according to formula y = 35.08(1 — 0.0888)”. 2 Calculated according to formula y = 56.75(1 — 0.0681)”. 78 Papers from the Marine Biological Laboratory at Tortugas. Table 21 shows the decline in weight of the two halves of a disk of Cassiopea xamachana starved in the diffuse daylight of the laboratory at Montego Bay, Jamaica, from February 27 to March 18, I912, in water ranging 100 ¢ from 24.5° to 28°C. The 94 h and th- a stomach and mouth-arms of the medusa were re- g 4] = 2S moved and the disk cut ? 50$ ; = 404 into two nearly equal ‘ : 30 ¢ halves, both of which were on maintained in one and the jo! same glass aquarium hold- © OO OOOO Onno : ‘ 0123 4 56 7 8 9 1011 12 13 14 15 1617 18 19 ing about 6 liters. The ge water was changed once in Diagram illustrating Table 21. every 24 hours. TABLE 21. Medusa A. Medusa B. tepeed Tie ae i Calculated i Calculated Weight weiahet Weight. eights 125.93 125.93 106.13 114.33 95.61 103.76 85.33 94.32 73.96 85.62 67.51 77.69 59.81 70.52 53.06 64.09 48.86 58.18 42.72 47.97 34.56 39.53 28.65 32.61 25.08 26.81 21.61 22.16 18.26 18.27 1 According to the formula y = 111.25(I — 0.089)”. 2 According to the formula y = 125.93(I — 0.092)”. Weights y 0 Days x 123 4 567 8 9 101112 13 14 15 16 17 18 19 Diagram illustrating Table 22. Table 22 shows decline in weight of two regenerat- ing specimens of Cassiopea xamachana starved 19 days in the dark, in filtered sea- water, at temperatures ranging from 27.5° to 30°C., from June 26 to July 15, 1912, at Tortugas, Florida. One was a disk deprived of its stomach and mouth. arms, and the other was a medusa regenerating its bell-rim, but with stomach and mouth-arms intact, Days elapsed. Law Governing Loss of Weight in Starving Cassiopea. 79 TABLE 22. Observed BS ee ae weight of re-| Weight of re- generating |Weight of disk.! generating generating disk in grams. medusa in medusa.? grams. 19.76 19.76 30.11 30.11 17.8 16.22 25.89 26.95 10.8 10.95 20.21 21.59 8.54 8.14 19.22 18.31 5.04 6.07 14.76 15.5 4-79 4.62 penis / 13.13 3.05 3.05 10.53 10.53 1 Calculated according to formula y = 19.76(I — 0.0937)”. 2 Calculated according to formula y = 30.11(I — 0.0538)”. The two specimens were maintained in one and the same 6-liter aquarium jar throughout the experiment, and were thus subjected to identical en- vironmental conditions. ©1283 4 5 6 7 8 9 1011 12 13 14 15 1617 18:19 Days x Diagram illustrating Table 23. The water was changed once in every 24 hours. Table 23 shows the decline in weight of two regenerating specimens of Cassiopea xama- chana, one being a disk with- out stomach or mouth-arms, and the other a medusa with stomach and mouth-arms in- tact but with the rim of the bell cut off. Thus each me- dusa was regenerating, the disk being without and the medusa with its stomach. These two specimens were kept side by side in one and the same 6- liter glass aquarium, in con- stant darkness, in filtered sea- water at 27.5° to 30°C. for 19 days from June 26 to July 15, 1912. The water was changed once in every 24 hours. 1 Calculated according to formula y = 2 Calculated according to formula y = TABLE 23. | Observed rv : ; Bei of = Weight of re- | weight of re- | Weight of re- generating Pee generating generating disk in grams. disk. medusa in medusa.2 grams. 21.53 21.53 40.7 40.7 19.59 18.02 38.8 36.65 II.1I2 12.83 27.12 20.72 8.56 9.74 26.33 25.4 7-07 7.46 21.06 21.71 5-76 5-74 19.48 18.55 4.04 4.04 15.05 15.05 21.53(I — 0.0843)7. 40.7(I — 0.0511)”, 80 Papers from the Marine Biological Laboratory at Tortugas. Table 24 shows the decline in weight of two regenerating specimens of Cassiopea xamachana starved I9 days in the dark, in filtered sea-water, aM at temperatures ranging from 27.5° to 30° C., from = June 26 to July 15, 1912, at Tortugas, Florida. One 35 of the specimens was a disk deprived of both 30 stomach and mouth-arms, e and the other was a me- %& 2 dusa with the rim of its =* bell cut off but with stom- 20 9 ach and mouth-arms intact; thus both were regenerat- 15} ing new tissue. They were kept side by side in one and the same 6-liter aquarium jar, and were thus _ sub- jected to similar condi- tions. The water was 0123 4 5 6 7 8 9 1011 12 13 14 15 16 17 18 19 changed once in every 24 Days x hours. Diagram illustrating Table 24. TABLE 24. Observed | Observed : Days | weight of re- | Weight of ; Weight of re- | Weight of re- elapsed.| generating disk. generating generating disk in grams. medusa in medusa.? grams. te) 28.31 28.31 49.64 49.64 Be 26.91 24.25 44.61 44.65 6 17.23 17.83 31.97 36.19 9 13.57 14.16 25.71 30.88 12 II.64 I1I.24 23.61 26.36 I5 8.8 8.01 20.62 22.51 19 6.58 6.58 18.24 18.24 1 Calculated according to formula y = 28.31 (1 — 0.074)*. 2 Calculated according to formula y = 49.64(I — 0.0512)”. Table 25 shows the decline in weight of the two halves of a disk of Cassiopea xamachana without stomach or mouth-arms. One of these halves, A, was starved in the diffuse daylight of the laboratory and the other, B, in constant darkness, for 18 days, from February 28 to March 18, 1912, at Montego Bay, Jamaica, at temperatures ranging from 24.5° to 28° C. The two halves were maintained side by side in one and the same 6-liter glass aquarium and the water was changed once in each 24 hours. Law Governing Loss of Weight in Starving Cassiopea. 81 Ome ea 56" 7 78, 9 10 11 12 13 14 15 16'17 18 Days x Diagram illustrating Table 25. TABLE 25. Medusa A. Medusa B. Days | Weight ? Weight, elapsed.) starved in Weight.1 | starved in | Weight.? diffuse day- darkness. light. ts) 64.11 64.11 63.46 63.46 I 52.18 57.96 52.53 58.16 2 44.28 52.38 45.91 | » 53:12 3 37.88 47.38 40.93 48.61 4 31.08 42.83 37.06 44.48 5 29.9 38.72 praia Se 6 e heke Bee 32.27 37.25 7 25.07 31.61 29.91 34.08 9 20.9 25.83 25.60 28.56 II 17.69 21.09 22.49 23-35 13 15.83 17.24 20.6 20. 15 13.45 14.11 16.73 16.75 17 11.86 II.54 15.46 14.02 18 10.40 10.45 12.83 12.83 1 According to formula y = 64.11(I — 0.096)”. 2 According to formula y = 63.46(1 — 0.085)”. TABLE 26. Days | Half-disk starved in : 2 elapsed. deep liches weight. ° 73.01 66.11 I 63.13 ; 58.83 2 53.06 3 41.49 4 33-91 5 31.01 hd SRA = Boer eee acca 7 26.51 Br Ae csetatiatets oisisa'| vis spicersta aie 9 22.44 TOW |e cia ak crcrosare oroi|lnvercta ate aie gereretes| chetare cisterace cep le II 18.56 12 stat cus tates elesetelterete Sclevele a. SI oie Cane 13 16.22 Ee IO rere iecate eiaiaceteitiataya stosia"aee al ehstale lever alevenvelata’'er'sl| scare eia\eleieial verb ate 15 13.47 de aly [See A el Pe 17 II.17 18 9.40 1 Calculated in accordance with formula y = 73(1 — 0.108)”. 2 Calculated according to formula y = 66.11(1 — 0.11)”. 82 Papers from the Marine Biological Laboratory at Tortugas. Table 26 shows the decline in weight of two halves of a disk of Cassiopea frondosa deprived of its stomach. One of these halves was starved in diffuse daylight and the other in the dark from February 28 to March 18, 1912, at Montego Bay, Jamaica. TABLE 27. Medusa A. | Medusa B. | Days elapsed.| Observed Weights. Observed Weights 2 weights. weights. fe) 60.96 61 57-54 57-5 2 51.17 50.08 49.3 49.45 4 42.26 42.27 45.36 42.26 6 34.01 35.19 39.06 36.51 8 29.33 29.33 31.41 31.41 Io 23.59 24.77 27.6 27.25 12 21.2 20.37 24.72 23.23 1Calculated from the formula y = 61(1—0.1673)”. 2Calculated from the formula y = 57.5(I —0.14)”. Table 27 shows the decline in weight of two normal meduse of Cassiopea xamachana starved each in one liter of sea-water, changed once in each 24 hours, and kept in the diffuse daylight of the laboratory at Tortugas, Florida, from June 8 to 20, 1913. One medusa, A, was starved in sea- water which had been passed through two glass funnels each holding two sheets of Chardin filter paper. The other medusa, B, was starved in sea- water, which, in addition to having been filtered through the aforesaid Chardin filters, was also filtered through a bacteria-proof porcelain filter. It appears that all food had been removed from the water by the Chardin filters and the medusa in the bacteria-free sea-water starved somewhat more slowly than the one in the sea-water which had not been passed through the porcelain filter. DESCRIPTION OF PLATE. Fic. A. Cassiopea xamachana. The 6 meduse have been starved for 41 days in diffuse daylight (see table 4). The recently caught, unstarved medusa in the center of the aquarium is of nearly the same size that the smallest (shown by the arrow) was when starving began. The dark-brown color of the starving medusz is due to their densely crowded commensal plant-cells. Fic. B. Cassiopea xamachana which has been starved for 41 days in darkness. Origin- ally its bell was 25 mm. wide; now it is only 4.5 mm. in diameter, and the rim is turned upward and inward. The mouth-arms are each 3 mm. long, and the mouths have disappeared by coalescence of the lips. The medusa is still pulsating, a single rhopalium functioning. Fic. C;. Normal medusa of Cassiopea xamachana. Fic. Cs. Disk without stomach or mouth-arms. Fic. C;. The stomach and mouth-arms of the medusa, the disk having been cut away. MAYER PLATE 1 IV. CHANGES IN SALINITY AND THEIR EFFECTS UPON THE REGENERATION OF CASSIOPEA XAMACHANA. BY A. J; GOLDFARB; Of the College of the City of New York. Four figures. 83 ee Sern Ne be me 7 © z ci HOt) 2Toaa Fide OMA WAS. Aiea AERTA AMAR MORRO OF AS A t BAA Lap ae he Bg Prat 0s Ve 2h wea 7 —— = ae ee. — As Ar ey ee has pamaibye.' sel ee | CHANGES IN SALINITY AND THEIR EFFECTS UPON THE REGENERA- TION OF CASSIOPEA XAMACHANA. By A. J. GOLDFARB. Cassiopea xamachana, a large scyphomedusa, is very abundant in the very shallow waters of the moat at Fort Jefferson, Dry Tortugas, Florida. The conditions in this moat are such that the meduse are in all probability subjected to extreme changes in salinity, due on the one hand to evaporation in the subtropical sun, to the shallow water and inclosed nature of the moat, to the single tide, etc., and on the other hand to extreme dilution due to the subtropical rains. Doctors A. G. Mayer and T. W. Vaughan have under- taken a study of these changes in salinity in and outside of the moat, under various conditions of tide and season. Until their report is forthcoming, this phase of the problem must be left in abeyance. The present report considers to what extent changes in salinity influence regeneration in Casstopea, and the results of the investigation are compared with those previously obtained with the hydroid Eudendrium ramosum of Woods Hole, Massachusetts, and with the observations of Loeb with the hydroid Tubularia of Serino Bay, Italy. Though certain variable factors are known to accelerate, retard, or modify the regeneration of an organism, these factors have not heretofore been sufficiently taken into account. In the present investigation much care was taken to exclude or make the following factors uniform: 1. Size of medus@. Size roughly indicates age, and as the rate of growth varies with age the medusez finally chosen were approximately of the same size. Each solution contained one of medium size, 100 to 105 mm. in diameter, and one smaller medusa, 85 to 90 mm. in diameter. 2. Volume, surface, and depth of solution. To insure uniformity of respira- tion and evaporation, and to prevent excessive concentration, the medusz were kept in jars of equal size and shape, each jar containing 2,000 c.c. of the solution. 3. Aeration and change of solution. All solutions were shaken thoroughly to aerate them, and were changed daily to avoid excess COs, excess evapora- tion, and bacteria. 4. Degree of injury. It has been repeatedly urged that the rate of regen- eration is influenced by the extent of injury. Though recent evidence is opposed to this view, its possibility was taken into account by removing the same number of arms from each pair of meduse. 85 86 Papers from the Marine Biological Laboratory at Tortugas. 5. Level of amputation has repeatedly been shown to affect considerably the rate and nature of the regenerated organ. In these experiments all arms were amputated at the same level, namely, at their respective bases. 6. Temperature. By avoiding local heat and local drafts the temperature was made uniform for the series. 7. Preparation of the solutions. To avoid disturbing the “ protective rela- tion” of the salts and to avoid introducing any salt not normally present in sea-water, the solutions were made as follows: The dilute solutions were made by adding enough (rain) water to make solutions containing 95, 90, 85, 80, 75, 70, 65, 60, 55, 50, 45, and 40 per cent of sea-water. The rain- water used in this series was tested by Dr. A. G. Mayer, who, upon the basis of the chlorine content, calculated that it contained 0.001 part of sea- water, an amount entirely negligible, as shown by these and other experi- ments. The concentrated series was made by slow evaporation in the sun’s heat. In this manner, solutions containing 5, II, 17, 25, 33, 42, and 53 per cent more than the normal salts in sea-water were obtained. 8. The number of arms amputated totaled 323. 9. Three sets of measurements were recorded, 14, 24, and 30 days after amputation. These measurements are given in table I. TABLE 1.—The influence of changes in salinity on regeneration. Per cent of | No. of am- Average regeneration in mm. sea-water putated | —AA Remarks. in solution. arms. 14 days. | 24 days. | 30 days. 153 15 ts) oO ° 142 I5 (0) ° 0.12 | (0) to) : vee a 0.16 = 0.66 { Regeneration abnormal. 117 I5 0.21 0.62 0.43 | III 15 4.32 4.23 1.92 2 105 I4 ayes 3.00 2.19 100 43 a4F 6.44 pe : 8.41 4 = = a are nee aa Regeneration normal. 85 15 4.60 7.17 5.05 80 13 3.02 5-33 3-45 75 15 1.87 6.09 3-74 70 15 2.22 5.41 3.63 AB = Bae 308 a8 Regeneration abnormal. 55 IS 0.44 1.57 2.87 590 IS 0.50 (0) (9) 45 15 (0) fo) (0) 40 15 (0) to) te) The data are plotted in figure 1, in which the abscissas represent the gradient solutions and the ordinates represent the average regeneration for each solution. From table 1 and figure 1, it will be observed that Cassiopea lived in solutions in which the salt content varied from a lower limit of 40 to an upper limit of 153 per cent. In the extreme solutions the medusz decreased in size, ceased to pulsate, and the organs degenerated. This occurred in the 4o and later in the 45 per cent as well as the 153 per cent solutions. Within these limits the medusz lived and pulsated, but not all regener- ated. No regeneration occurred in solutions containing 50 per cent or less Changes tn Salinity and Effects on Regeneration, etc. 87 sea-water, nor on the other hand was there any regeneration in solutions 133 or more per cent, with one exception to be noted later. Though regener- ation occurred within these limitations, the character of the resulting regenerated arms was not the same. Normal pyramid-shaped arms with characteristic mouths and oral filaments were developed in the median solutions only, 7. e., 75 to 105 per cent inclusive. Inthe 70 per cent solution 6 Fig. 1.—Thin line represents regen- 5 eration after 14 days: thick line 24 days; broken line 30 days. 153 142 133. 125 117 111 105 100 95 90 8 80 75 70 65 60 55 60 45 40 the mouths were considerably reduced in number, while the number of oral filaments was scarcely affected. In the 65 per cent solution there were very few mouths differentiated. In the 60 and 55 per cent solutions practically no mouths were formed and the oral filaments tended to clump together. Similarly, beginning with the I11I per cent solution, of the concentrated series, the arms were increasingly atypic. In this solution some arms were normal, others clumped as in the diluted solutions. In the 117 per cent solution, fewer mouths were differentiated and greater crowding of the oral filaments occurred. In the 125 per cent solution no mouths and only very few filaments were differentiated. It is then clear that while regeneration took place in a wide range of solutions, normal regeneration occurred within much narrower limits, namely, 75 to 105 per cent solutions. Corroborative results were obtained from measurement of regenerated arms. Figure 1 and table 1 clearly show that when the average regeneration in sea-water is taken as the norm, then the average regeneration in III to 80 per cent solutions is norm or supernorm, varying from 3.4 to 5.4 mm. On either side of these limits there is a rapid decline, sharper in the more concentrated solutions, far less rapid in the more dilute series. For ex- ample, the average regeneration in the 117 per cent solution is less than in the 50 per cent solution. The second series of measurements was taken Io days after the first, 1. €., 24 days after removal of the arms. During these 10 days practically no regeneration occurred in the concentrated solutions, while retrogression setin. In all the dilute series, however, considerable regeneration occurred. This is seen in figure I. Perhaps the facts are made clearer by table 2, which 88 Papers from the Marine Biological Laboratory at Tortugas. gives the difference in the average regeneration between the fourteenth and twenty-fourth days. Table 2 shows that in the concentrated solutions, 105 to 125 per cent, there was no additional regeneration; in fact, three out of four solutions showed a decrease and in one solution the increase was only 0.63 mm. In all the dilute solutions, however, there was an increase, in sea-water 3.03 mm., in 95 per cent solution 2.96 mm., 90 per cent solution 2.63 mm.,etc. Itshould be noted that the increment was unequal, a greater relative and absolute increase occurring not in the optimum solutions, but in 75, 70, and 65 per cent solutions. TABLE 2.—Extent of regeneration during the second period, 1. e., between the fourteenth and twenty-fourth days after amputation. Average gain or loss be- Solutions. tween 14th and 24th day. mm. 1.33 fo) 1.25 —0.33 DeE7, +0.63 I.II —0.09 1.05 —0.72 I.00 norm +3.03 0.95 +2.06 0.90 +2.63 0.85 +2.590 0.80 +2.31 0.75 +4.22 0.70 +3.20 0.65 +2.39 0.60 +1.19 0.55 +1.14 0.50 t9) In the third set of observations, made on the thirtieth day, the measure- ments were not taken from the amputated level, for at this stage of develop- ment it was difficult to tell with exactness where that level was. The last measurements were made from the end of the knob-like mass at the distal Fic. 2.—Thin line : represents regen- erating after 14 days; thick line after 24 days; broken line after 30 days. 150 140 130 120 110 100 90 80 70 60 50 40 Atypical Reg. Typical Reg. Atypic Reg. errr a SS end of the amputated arms, and the measurements in table 1, therefore, appear smaller than the corresponding measurements for the preceding periods. Bek The character of the curve is essentially the same as the 24-day curve, differing from it in the same way that it differs from the 14-day curve. It Changes in Salinity and Effects on Regeneration, etc. 89 is seen that the curves for 105 to 75 per cent solutions are almost parallel. Beyond these limits the curves diverge. In the 75 to 55 per cent solutions there has been either an absolute or relatively greater increment than in the solutions approaching the norm. In the hypertonic solution there is no increase, the curves for the three periods are essentially parallel and overlap. TABLE 3. Per cent of sea-water | Average regeneration in solution. for 30 days. 1.50 1.40 0.13 1.30 te) 1.20 0.50 I.10 I.91 1.00 4.81 0.90 5-34 0.80 4.25 0.70 3.68 0.60 2.77 0.50 1.43 0.40 fo) All these facts are made extremely clear after reducing some of the irregularities of the curves, by grouping the results on a basis of 10 per cent changes in salinity instead of 5 per cent, as shown in figure 2. Table 3 shows the results of table I grouped on a basis of 10 per cent changes in salinity, instead of 5 per cent differences. In the preceding tables and curves the fact was ignored that in certain solutions an arm was replaced by two arms. The supernumerary arms appeared in the median solutions, viz, 111 to 85 per cent, with a few desul- tory ones in more dilute solutions. The greatest number occurred in the 95 per cent solution, fewer in the 90 and 85 per cent, a sharp drop in 105 per cent and III per cent solutions, as shown in table 4. Table 4 gives the number of supernumerary arms in each solution and shows that they tend to appear in optimum dilute solutions. TABLE 4. No. of | No. of Per cent of |No. of arms\doublearms| Percent || Percentof |No, of arms\doublearms| Per cent sea-water in| removed. regener- of double | sea-water in | removed. regener- of double solution. ated. arms. | solution. ated. arms. 1.53 15 fe) fe) 85 15 4 26 1.42 15 te) o | 80 13 fe) ° 1.33 15 fo) to) 75 15 fe) ° 1.25 14 te) to) | 70 15 fe) | fe) 1.17 I5 (o) °o 65 15 2 13 III I5 I if 60 14 fe) °o 105 14 I 7 55 15 I 7 100 43 5 II | 50 15 ry oO 95 15 5 33 45 15 Oo o 90 15 4 26 40 15 (0) fa) When the supernumerary arms are included in table 1, it is found that the highest part of the curve, 85 to III per cent, is augmented still further. The maximum augmentation occurs in 85 to 95 per cent solutions, and only a moderate increment occurs in sea-water. go Papers from the Marine Biological Laboratory at Tortugas. Whether we consider only the regeneration of the supernumerary arms or of single arms, or normal as opposed to abnormal regeneration, or the variation in growth at different intervals, or all of these combined, it is clear that the maximum regeneration in Cassiopea occurs in sea-water di- luted to 95, 90, and 85 per cent. There are three sets of records complete enough to permit comparison, namely, Loeb’s work on Tubularia,! Goldfarb on Eudendrium,? and the present contribution on Cassiopea. Loeb’s data are based upon the regeneration of 65 pieces of Tubularia stems over a period of 8 days. The work was done at Serino where the sea- water is estimated on the basis of Forchhammer’s calculation to contain 3.8 per cent salts or more correctly 38.364 per kilogram. Loel» used 10 grades of solutions and plotted his results as shown in figure 3. 5 Fic. 3.—Thick line , is Loeb’s curve for Tubularia; broken line is curve for Cassio- . pea after 30 day 2 interval. 1 150 =140 130 = 120 110 100 = 80 70 40 60 40 The 323 Cassiopea arms observed at 14, 24, and 30 day intervals in sea- water whose salinity, though not definitely known, is calculated* to be 3.549 per cent. The medusze were placed in 20 graded solutions. The results differ from Loeb’s in at least four respects: 1. The salinity limits are different, 7. e., Tubularia regenerated in more dilute solutions than Cassiopea, though the extreme concentration is prob- ably the same for both. 2. Maximum regeneration occurs at different salinities, 65 per cent sea- water for Tubularia, 95 per cent for Cassiopea. 3. There is a greater range of supernormal regeneration in Tubularia 1. €., 100 to 50 per cent salinity, while the range for Cassiopea is limited to 15 per cent, 2. e., from norm to 85. 4. The character of the curves is reversed. For Tubularia there is a graded increased regeneration with increasing dilution from 133 to 65 per cent; beyond this limit there is a sudden drop to complete inhibition in 40 per cent. The Cassiopea curve on the other hand rises rapidly to 95 per cent, beyond which there is a very gradual decrease (fig. 3). The third set of observations were those on the compound hydroid Eudendrium ramosum. The results of one series, based on 953 stems, are plotted in figure 4, and when compared with Tubularia and Cassiopea show the following: 1Loeb, J. Organization and growth. Wurtzburg. 1801. 2 Goldfarb, A.J. Factors in the regeneration of a compound hydroid, Eudendrium ramosum. Jour. Exper. Zool., vol. 4, p. 43. 1907. 3 Clarke, F. W. The data of geochemistry. Bull. 491, U. S. Geol. Survey. tI911. Changes in Salinity and Effects on Regeneration, etc. gI FIG. 4. Se er et ee ae Eudendrium, 6 days. _......Hudendrium, 3 days. ......Cassioped. scene al UOUarEC. 150 140 130 120 110 100 90 80 70 60 50 40 1. Regeneration takes place in a more limited variation from the norm than either Tubularia or Cassiopea, 1. e., the maximum salinity in which regeneration takes place is 120, the minimum is 60. 2. Maximum regeneration is intermediate, 65 per cent solution for Tubularia, 85 for Eudendrium, and 95 for Cassiopea. 3. Normal or supernormal regeneration in Tubularia occurs over a wide range of solutions, from norm to 50 per cent; in Eudendrium it is inter- mediate, 100 to 80 per cent; in Cassiopea most limited, 100 to 85 per cent. 4. The nature of the curves varies. Tubularia and Cassiopea have asymmetrical curves, Eudendrium almost symmetrical, 7. e., beyond the optimum solution, increased dilution or concentration is equally injurious. Two sets of observations after 3 and 6 day intervals afforded the oppor- tunity of comparing these with corresponding sets of observations on Cassiopea. The regeneration in Eudendrium is given in table 5. TABLE 5. Pesieeat oes water i Number of polyps regenerated in solution. First 3 days. Second 3 days. Total, 6 days. 130 fc) a _ 117 5 27 32 III I4 or SI 100 (norm) 46 36 = 95 56 23 79 8s 62 22 84 75 28 45 73 65 0 34 34 It is apparent that the same phenomenon observed in Cassiopea occurred in Eudendrium, namely, that with increased intervals there is a greater relative regeneration, not in the optimum nor hypertonic solutions, but in the more dilute solutions, such as 75 and 65 solutions. It is very clear, both from Cassiopea and Eudendrium, that such changes in the sub-optimum solutions do not and can not alter the essential character of the curves sufficiently to make it analogous with Loeb’s. The differences in the behavior of the three organisms is made easier for comparison by table 6, which gives the results of changes in salinity upon the three organisms studied by Loeb and Goldfarb. 92 Papers from the Marine Biological Laboratory at Tortugas. TABLE 6. —-— Maxi- | Maxi- | Opti- Regenera- Beyond opti- | Beyond ne mum | mum | mum |, mum on con-| mum on di- =e dilu- |concen-| solu- eae _centrated | luted side of Salinity. Curve. tion. |tration.} tion. side of curve. curve. p.ct. | p.ct. | p.ct. p. ct. ae ee Mer ust a Tubularia...} 40 133 65 too to 50 | Less injuri- | Most injuri- 3.83 |Asymmetrical. ous. ous. Cassiopea'... 50 133 95 100 =>: 85, || Most injuri- | Least injuri- 3.54 |Less asymmetri- ous. ‘ ous. cally Eudendrium. 60 125 85 too =. 80._|| Equally in- | Equally in- 3.29 |Symmetrical. jurious. jurious. In these three organisms the increasing dilution of sea-water, with the consequent intake of water into the organism, makes for a more rapid and greater growth. Beyond an optimum dilution, any further dilution retards and ultimately inhibits regeneration. Concentration of sea-water, in all cases, makes for an expulsion of water from the organism, with a concomitant retardation, decreased regeneration, and finally no regeneration whatsoever. Beyond these fundamental similarities, there is little or no agreement. Though there appears to be little or no similarity in the behavior of these three organisms, on further examination it becomes quite clear that Eudendrium and Cassiopea are in essential agreement, both differing in a striking manner from Loeb’s results on Tubularia. Eudendrium and Cassiopea agree in the following respects: (1) maximum regeneration occurs in solutions that are hypotonic in a relatively small degree, 7. e., 95 and 85 per cent respectively; (2) regeneration in sea-water is equaled or exceeded in a narrow range of hypotonic solutions, 7. e., 100 to 85 and 80 per cent, respectively; (3) at any interval after amputation, the graphs show a sharp decline in the amount regenerated in increasingly hypertonic solutions; a maximum regeneration of not more than 60 per cent beyond the norm, 7. e., 60 per cent in Cassiopea, 35 per cent in Eudendrium; a gradual followed by a more rapid drop in the curve, in solutions more hypotonic than these optimum solutions. It is in these very points that both these curves differ from Loeb’s work on Tubularia, as follows: The maximum regeneration occurred in a markedly greater hypotonic solution, 7. e., 65 per cent; regeneration in sea- water is equaled or exceeded in a considerably wider range of diluted so- lutions, 7. e., 100 to 50 per cent solutions inclusive; the curve is completely reversed, the gradual increment and the sharp decline appearing on sides opposite to that in either Eudendrium or Cassiopea; the mode is nearer the extreme dilution; and the maximum regeneration is far greater (120 per cent) than in either of the other two organisms. It is estimated that the maximum salinity of the sea-water at Woods Hole, Massachusetts, is 3.29,1 for the Dry Tortugas 3.54 per cent.2 It might be urged that a definite relation obtains between the density of the sea and the regeneration of the contained organism, subjected to a graded 1 Bulletin of Bureau of Fisheries, vol. 31. IogItI. 2Clarke, F.W. The Data of Geocheaiarnge **Bull. 491, U. S. Geol. Survey. Io1I. Changes in Salinity and Effects on Regeneration, etc. 93 series of densities, for the increased density of the sea-water at the Dry Tortugas is paralleled by each of the changes already cited in table 6; but when Tubularia of Serino Bay, whose density is 3.83, is taken into consider- ation, no correlation between density and the character of the curve obtains, except in two particulars: (1) the greater the density of the sea, the greater the dilution in which regeneration will take place; (2) the greater the density of the sea the greater the regeneration in the hypotonic solutions, 1. e., Eudendrium A 3.29, maximum regeneration is 35 per cent above norm Cassiopea, A 3.54, 60 per cent, Tubularia, A 3.83, 120 per cent. Subject to a more accurate analysis of the density of the water in the moat at Fort Jefferson and at Serino Bay, these conclusions must be con- sidered tentative only. The fact remains, however, that Loeb’s limited data are not in accord with the results obtained in two different organisms from two different localities whose sea-water was of different densities. It is also clear that the more detailed and more numerous data upon Euden- drium and Cassiopea are in close agreement. And we are driven to the conclusion that either Tubularia, for some inexplicable reason, differs in its behavior from the other two ccelenterates, a view which is very im- probable, or that the insufficient data upon Tubularia has misled the author. The close similarity of Eudendrium and Cassiopea and the corroborative results of A. G. Mayer upon the effects of changes in density upon nerve- conduction leads me to believe that the results here given may be taken as indicative of the gross effects, without any effort to segregate the osmotic from the ionic influences, upon the regeneration of these organisms. SUMMARY. The object of this investigation was to ascertain to what extent changes in salinity affected an organism, Cassiopea xamachana, normally subject to relatively great variation in the concentration of the sea-water, and to compare the results with those of the hydroid Eudendrium and the hydroid Tubularia. Considerable attention was paid to render the following variable factors uniform for the series: Size of medusz; volume, surface, and depth of the solutions; extent of injury; level of amputation; temperature; crowding. Injurious or other variable factors were guarded against: (1) by aerating and changing all solutions daily; (2) by diluting with water containing known amount of sea-salts; (3) by concentrating the sea-water very slowly; (4) by examining the 323 regenerating arms at intervals of 14, 24, and 30 days. Cassiopea lived in solutions ranging from 40 to 153 per cent sea-water solutions. Regeneration occurred in solutions containing 50 to 133 per cent sea- water. Normal regeneration of the arms occurred within much narrower range, namely, 75 to 105 per cent. Beyond these limits regeneration was atypic. 94 Papers from the Marine Biological Laboratory at Tortugas. While the amount regenerated in sea-water was equaled or exceeded in solutions diluted to as much as 85 per cent, the optimum solution was 95 per cent sea-water. With increasing concentration, regeneration fell off rapidly; with increasing dilution the amount regenerated was diminished ' very slowly. During the succeeding 10 and 16 days there was a relatively greater regeneration in solutions below the optimum, 7. e., in the 75 and 70 per cent, and the character of the curve was correspondingly altered. The development of supernumerary arms likewise occurred most frequently in the optimum solutions, namely, 95 to 85 per cent; less fre- quently with increasing dilution or concentration. When these results are compared with those of Loeb on Tubularia and with the writer’s on Eudendrium, it is clear that in all the maximum regener- ation does not take place in sea-water, but in sea-water diluted, and that with increasing concentration or dilution regeneration is retarded and diminished. The character of the curves representing the influence of changes in salinity upon the regeneration of these organisms is strikingly different. 1. The most dilute solutions in which regeneration will take place is 4o per cent for Tubularia, 50 per cent for Cassiopea, and 60 per cent for Eudendrium. 2. The maximum regeneration occurs in 95 per cent solutions for Cassiopea, while it is 85 per cent for Eudendrium and 65 per cent for Tubularia. 3. The regeneration in sea-water is equaled or exceeded in solutions diluted as much as 13 per cent for Cassiopea, 20 per cent for Eudendrium, and 50 per cent for Tubularia. 4. The curve is most asymmetrical for Tubularia, less so for Cassiopea, and almost symmetrical for Eudendrium. 5. The Eudendrium behaves more like Cassiopea; both are strikingly different from Tubularia. It is very evident that Loeb’s curve is not representative of the behavior of marine organisms to changes in concentration, and it is altogether prob- able that such a curve can not be expressed in a simple curve based on two variables. In the absence of information concerning the exact concentration of the sea-water, where each of these experiments was made, it is unprofitable to make extended generalizations. oe V. REGENERATION IN THE ANNELID WORM AMPHINOMA PACIFICA, AFTER REMOVAL OF THE CENTRAL NERVOUS SYSTEM. BY A. J,, GOLDFARB, Of the College of the City of New York. 95 REGENERATION IN THE ANNELID WORM AMPHINOMA PACIFICA, AFTER REMOVAL OF THE CENTRAL NERVOUS SYSTEM. By A. J. GOLDFARB. In a previous publication,! the writer found that the head of the earth- worm Lumbricus was regenerated in the entire and permanent absence of the nerve-cord from the amputated region. Inasmuch as this was the first demonstrative experiment of a complete and perfect regeneration in the absence of the nerve-cord, it seemed extremely important to ascertain whether the phenomenon was peculiar to Lumbricus, and therefore of limited significance, or whether other annelid worms could likewise regen- erate without the aid or presence of the nerve-cord. With this object in view, the marine annelid Amphinoma pacifica was chosen, because it lives so well in the laboratory, because large numbers could be readily procured, but especially, because it was sufficiently large and hardy to permit of the necessary operations. There were two types of operations: 1. After narcotizing in a 5 per cent solution of alcohol in sea-water, the worms were amputated and the nerve-cord extracted. This was done by passing a fine forceps into the ccelomic cavity in such a manner that the nerve-cord lay between the tongs of the forceps, and at any desired position the forceps were firmly brought together and the contained nerve-cord was extracted with its ganglia and basesof the nerves. These could then be readily counted. I never succeeded in getting pieces so long as those in the earthworm, illustrated in my report.! I did succeed in extracting as many as nine consecutive ganglia. This method had the distinct advantage of causing very little injury to adjoining tissues. The importance of this fact will be later emphasized. 2. The second type of operation, used by Morgan? and modified by Nussbaum,’ consisted in cutting a ‘‘window”’ from the ventral side near the amputated level. The parts so removed or injured included cuticle, muscle layers, nerve-cord, often the ventral blood-vessel, and the nephridia. A modification of the operation consisted in making a ventral incision, then 1 Goldfarb, A. J. The influence of the nervous system in regeneration. Jour. Exp. Zool., 7, 4, pp. 643- 722. 1900. 2 Morgan, T. H. Experimental studies of the internal factors of regeneration in the earthworm. Arch fur Entw. Mech., 14. 1902. 7 , ’Nussbaum, J. Beitrage zur Frage iiber die Abhiangigkeit der Regeneration yom nervensystem bei Nereis diversicolor. Arch. f. Entw., vol. 25. 1908. 7 97 98 Papers from the Marine Biological Laboratory at Tortugas. deflecting the muscle layers sufficiently to permit the removal of the ex- posed nerve-cord. The muscle walls were then replaced and ligatured. There were 90 operations of the first type, 12 of the second, and 151 worms served as controls. The worms were amputated at different levels and the nerve-cord removed from the anterior as well as the posterior parts. It is well known that the regeneration from the two exposed surfaces at certain levels may be quite different. The flatworm Dendrocelum lacteum,' for example, regen- erated a tail from the anterior piece and a head from the posterior piece, provided the amputation was not made more than one-third from the anterior end. Posterior to this level the head piece regenerated a tail, but the tail piece did not regenerate a head. In the earthworm Lumbricus the same phenomenon occurs. The differentiating level occurs at about the fifteenth segment. Amphinoma differs from this type of regeneration in two respects: (1) the level at which the two cut surfaces can regenerate is much more: posterior, 7. e., about the middle of the body; (2) instead of the regenerative power of the tail pieces being limited to certain narrowly prescribed levels, the reverse obtains, 7. e., head pieces can regenerate tails only in the posterior half or third of the worm, while tail pieces regenerate heads at practically all levels. Table 1 gives some details of the regeneration in the head pieces of con- trol worms, 7. e., level amputated, the number that regenerated a tail, etc. It is evident that no head piece regenerated when amputated less than one-eighth from the anterior end of the worm. Between the third and the middle of the body some of the pieces regenerated, others did not. All, however, regenerated posterior to the forty-fifth segment or middle of the body. TABLE I. : Level amputated. No. of worms. | No. survived. | No. regen- | No.with no erated. regeneration. One-eighth from anterior end.... 9 8 °o 8 Twenty-fifth segment.......... I I I °o Thirty-fifth segment........... I I I °o One-third from anteriorend..... 3 3 2 I Middle lof iworittc. o. acwisiecatecica'e 5 5 2 3 Forty-fifth segment and farther posterior by actual count..... 9 9 9 o Three-fourths from anterior end. 4 4 4 oO In table 2 similar data are given for tail pieces of control worms. The data are arranged with respect to the level amputated and the number that did or did not regenerate. All of the tail pieces regenerated a head from the first quarter, the middle, and even from the distal quarter of the worm. Beyond this level a number of pieces did not regenerate. There were 9 such pieces that formed no head. Of these, 5 were small distal pieces 12 to 17 segments from the distal end, and probably too small to regenerate; in the other 4, I have no records of the exact level amputated. Regeneration in Amphinoma Pacifica, etc. 99 TABLE: 2: Level amputated. No. of No. sur- |No. regener-| No. that did not pieces. ne Vived | ated. regenerate. One-fourth to middle.......... | 10 3 “| 3 Oo One-eighth to middle.......... 22 13 | 13 (6) One-eighth to seven-eighths..... 10 8 8 fo) One-fourth to one-half......... (?) 19 | ae) o One-eighth to seven-eighths..... 18 07 10 7 One-eighth to seven-eighths..... 13 II | 9 2 Tables 1 and 2 establish the fact that at anterior levels the head piece can not regenerate a tail, while the tail piece can regenerate its head. At posterior levels the reverse occurs: the tail piece can not regenerate a head, though the head piece can regenerate a tail. In the median levels both pieces may regenerate their respective missing parts. These preliminary facts concerning the regeneration of Amphinoma, after ordinary amputation, will aid in understanding the behavior of the head and tail pieces after the removal of the nerve-cord as already described. Of the 56 operated head pieces, 29 survived 25 to 30 days, as shown in table 3, which also gives the level amputated and the number that failed to regenerate. TABLE 3. No. regen- No. | erating a tail. Operated head pieces. 6.5.50 ccc cccssewccsce 56 morava SUHEVIVEC es lace crate eisiciencleetieisle ele al eeinioieeioe 20 ae Amputated at one-eighth level............. 8 oO one-fourth level............. 3 fo) one-third levels oo .ccicc. eee 10 {e) one-halfilevel’s.. 5 nce «0%. o0,0:0 31 7 ° three-fourths level........... I °o These operations were made at anterior levels before it was known that head pieces did not normally regenerate from these parts of the worm. Though regeneration took place readily at posterior levels of control worms, not one operated head piece developed a tail. The experiment appears to demonstrate that, in these instances at least, regeneration had been pre- vented by destroying the nerve-cord near the amputated level. I have elsewhere suggested! that regeneration could be prevented and frequently was prevented by various means other than injury to the central nervous system. In the newt Diemictylus, the regeneration of the tail was prevented by plugging the cavity of the spinal column with paraffin. In the salamander Tornier? prevented the formation of the tail by operating the skin and muscle tissues in a manner to prevent their coordinate de- velopment. In Amphinoma regeneration was prevented not only by injury to the nerve-cord, but by removing the alimentary tract from five or more segments nearest the amputated level. In table 4 the anterior pieces with the alimentary tract so removed are arranged according to the level amputated. 1 Goldfarb, A.J. The influence of the nervous system in regeneration. Jour. Exp. Zool., vol. 7, 1909. The central nervous system in relation to the phenomenon of regeneration. Arch. f. Entw., vol. 32, A. TOTT. Ficbom zd = Kamp der Gewebe im Regenerat bei Begunstigung der Hautregeneration. Arch. f. Entw,, 22, 1900, D. 34 100 Papers from the Marine Biological Laboratory at Tortugas. TABLE 4. Number | Number eal d Number Number not of worms. | survived. evel amputated. | regenerated. | regenerated. =} 2 One-eighth ...... ° 2 3 2 One-third ....... oO 2 3 2 One-half ........ fa) 2 3 2 Three-fourths.... I I That regeneration was not prevented by injury to, or absence of, the nerve-cord can be demonstrated only by more direct proof. It must be shown (1) that regeneration did not take place, even though the amputated region was re-innervated by the regenerated nerve-cord; (2) that regenera- tion could take place without any trace of the nerve-cord in the region adjoining the amputated level. This proof was obtained by a detailed examination of serial longitudinal sections of posterior pieces. Table 5 in- cludes the 46 posterior pieces from which the nerve-cord had been removed and which were subsequently preserved 25 to 30 days later.! TABLE 5. Amputated level. | Regenerated. | No regeneration. QORE=CISRED io wo ic10 ietoveietere 5 II Oniasreiidilils -dpobevodos | m) 3 One-third rc. nycaeeisree oO | 2) Onesnale se aniese rei acae | 2 2 fo) | I Out of 28 worms, 7 regenerated a distinct head; the other 21 did not regenerate in the same interval. Worm number 33 is typical of the 7 operated worms which regenerated their heads. ‘The serial sections showed that the nerve-cord had been completely removed from the seven segments nearest the amputated level, agreeing with the record made at the time of operation. It was found that typical ganglia of the nerve-cord extended as far as the middle of the sixth segment from the amputated level, at which point the nerve-cord suddenly became narrow and extended in this very narrow, partially ganglionated condition into the fifth, fourth, third, second, and first segments, beyond the amputated level into the regener- ated head. It continued dorsally and ended finally in a clearly differ- entiated supra-cesophageal ganglion. The details are identical with those described elsewhere? for Lumbricus. These worms afford no clue to the dependence or independence of regeneration upon the presence of the nerve-cord at the amputated level, nor of the manner in which the central nerve system was formed. The second group of operated worms did not regenerate the head. I have arbitrarily designated a regenerated head one in which at least the stomadeum and “‘brain”’ were clearly differentiated. Worm number 32 may serve as an example of this group. In this worm the characteristic 1 The high mortality occurred chiefly in worms operated by the ‘‘ window’”’ method of removing the nerve-cord. 2 Goldfarb, A. J., vzde. Regeneration in Amphinoma Pacifica, etc. 101 nerve-cord was present as far as the eighth segment from the amputated level. Beginning at the seventh segment the nerve-cord was very narrow and extended in this attenuated condition through the sixth, fifth, fourth, third, second, and first segments. The regenerated nerve-cord had not only reached the amputated level, but extended slightly beyond, into the proliferated embryonic tissues laid down at the cut end. A similar regen- eration of the nerve-cord occurred in worms 13, 9, 45, etc., in which the nerve-cord regenerated two to seven or more segments, and in which the nerve-cord did not always reach or pass beyond the amputated level. In the worms whose nerve-cords had regenerated to the amputated level, supplying this region with nervous and trophic stimuli, the lack of regeneration could not have been due to any lack of innervation. In a third very small group the head regenerated without the presence of the nerve-cord at the amputated level. In worm No. 7, for example, the characteristic nerve-cord extended as far as the sixth segment from the amputated level. The regenerated nerve-cord was present in the fifth and part of the fourth segment. Beyond this region, 7. e., the fourth, the third, the second, and posterior half of the first segment, there was no trace of a nerve-cord. Close to the amputated level and slightly beyond it into the embryonic tissues, there was an unmistakable ganglionic mass, from which there extended posteriorly a short, narrow nerve-commissure. This supra- cesophageal ganglion had been formed independently of the direct contact or stimulation from the regenerating nerve-cord. Worm No. 34 affords a much better idea of the development of the newly formed head. The nerve-cord was intact as far as the sixth segment, from which point the regenerated nerve-cord extended through the fifth, fourth, and third segments. The next two segments were entirely free of traces of the nerve-cord. Beyond the amputated level there was a fairly large regenerated head with a clearly differentiated ‘‘brain,’’ from which the nerve-commissures extended ventrally and posteriorly almost to the ampu- tated level. These two worms clearly show that it is possible for a head to regenerate with the nerve-cord 2, 3, or more segments from the amputated level. They show that the development of the brain and commissures may secondarily unite with the regenerating nerve-cord, as was found to be the case in Lumbricus. In view of the regeneration of the head in Lumbricus, in the absence of innervation from the nerve-cord, and in view of the corroborative results obtained in Amphinoma, it seems evident that regeneration can take place under these circumstances. It is very probable that the presence of the nerve-cord accelerated regeneration. It may also be that regeneration does not take place ordinarily until the amputated level has been innervated from the cells of the nerve-cord as urged by Nussbaum.! It is equally clear, however, that a head can be regenerated in at least two annelid worms without the aid or stimulus derived from the nerve-cord. 1 Nussbaum, Josef. Beitrag zur Frage uber die Abhangigkeit der Regeneration vom Nervensystem be Nereis diversicolor, Arch. f. Entw., vol. 25. 1908. 102 Papers from the Marine Biological Laboratory at Tortugas. SUMMARY. The marine annelid worm Amphinoma pacifica readily regenerated a head at all levels except the distal eighth of the worm. The formation of a tail occurs only in the posterior half or third of the body. Only in the median levels of the body can the two cut surfaces regenerate a head and a tail respectively. Regeneration may be prevented by a severe injury, either to the digestive tract or to the central nerve system; the greater the injury the more likely will regeneration be inhibited. Many pieces did not regenerate after removing the alimentary tract from five or more segments nearest the amputated level. Many pieces, about one-third, failed to regenerate after removing the nerve-cord by the forceps, 7. e., with little injury to adjoining tissues. All failed to regenerate after removing the nerve-cord by the ‘“‘window”’ method, 7. e., with serious injuries to adjoining parts. The operated worms were examined in serial sections. In one group a regenerated nerve-cord connected the regenerated “‘brain’’ and commissures with the old intact nerve-cord. In a second group the regenerated nerve-cord approached and in many instances reached the amputated level, yet no head was formed. In a third group, the nerve-cord had regenerated, but several segments nearest the amputated end were yet without any nerve-cord or ganglia. These worms nevertheless had regenerated a head with its typical brain and nerve-commissures. The last group completes the demonstration that Amphinoma as well as Lumbricus can regenerate without the contact of, or stimulation from, the nerve-cord or central nervous system. VI. EXPERIMENTALLY FUSED LARVAE OF ECHINODERMS WITH SPECIAL REFERENCE TO THEIR SKELETONS. By A. J. GOLDFARB, Of the College of the City of New York. Fifteen text-figures. EXPERIMENTALLY FUSED LARVAE OF ECHINODERMS WITH SPECIAL REFERENCE TO THEIR SKELETONS. By A. J. GOLDFARB. The early work of Loeb, Morgan, and Herbst on the production of multiple embryos from a single egg suggested the reverse experiment of grafting or reuniting several fertilized eggs into one embryo. This im- portant experiment was first performed by Driesch, who succeeded in grafting together two eggs in about 0.04 per cent of the culture. With the same technique he subsequently grafted the eggs in other species of echino- derms and succeeded in keeping such grafted groups alive until they had developed into the larval or pluteus stage. He found that these larve, fell into several distinct types, as follows: true twins, twins with a common blastoccele, twins with a reciprocal influence on growth, fusion with a partially double archenteron, perfect fusion with a single set of organs, and single body with a second parasitic archenteron. In 1912, the writer repeated these experiments with the American form Arbacia punctulata and succeeded only after slightly modifying Driesch’s method. In the best cultures about 40 per cent were fused in groups of 2, 3, 4 or more eggs. These fused groups were preserved at different stages of development for further detailed study, but sufficient precautions had not been taken to guard against the solution of the very delicate skeletal struc- tures of the larve. Consequently the early stages only were available for detailed and complete study. Such studies corroborated and extended the results of Driesch. Subsequently, in the performance of other experiments, it was discovered that eggs could be agglutinated and fused quite as readily by a very different method, which was not only simpler but free of certain objections that might be urged against previously known methods. The new method consisted in using an isotonic or slightly hypotonic NaCl solution diluted with varying quantities of sea-water. NEW METHOD OF FUSING EMBRYOS. The best results were obtained by the following procedure: The eggs of Toxopneustes were artificially fertilized, and 15 to 20 minutes later approxi- mately equal quantities were placed in solutions made up of 70 to 80 parts of 5/8 m. NaCl (isotonic with sea-water at the Tortugas) and 30 to 20 parts of sea-water. After 5 to 8 hours in these solutions the eggs were trans- I05 106 Papers from the Marine Biological Laboratory at Tortugas. ferred to fresh sea-water. In the best cultures the number of agglutinations and subsequent fusions were as numerous as in the best cultures of Arbacia with the modified Driesch’s method. EARLY DEVELOPMENT. Inasmuch as I have elsewhere described in detail the early develop- mental changes in agglutinated and fused embryos of Arbacia and since the changes in Toxopneustes are essentially identical with those in Arbacia, I will only supplement the previous account by stating briefly at what stage and under what circumstances agglutination occurred after treatment with the NaCl solutions. The addition of an isotonic NaCl solution to sea-water involves a dilution of the other sea salts and a concentration of the NaCl. The excess NaCl stimulates the eggs to develop more rapidly, provided that the solution does not contain more than 50 per cent of the NaCl solution. In solutions con- taining 50 to 60 per cent NaCl solution, the excess NaCl seems to be neutral, but beyond this quantity with 90 or more per cent development was dis- tinctly retarded and atypic. That the eggs in the NaCl solutions absorbed water was evidenced by the distinct increase in size over the corresponding controls, and by the diminution in the space between the egg and the fer- tilization membrane. In many instances the fertilization membrane burst and a part of the protoplasm of one or more of the blastomeres protruded from the fertilization membrane, while in other eggs this membrane was completely thrown off. The differential rate of development in the NaCl solutions closely paral- leled the differential rate of regeneration in varying concentrations of sea- water. This suggested that the agglutinations might be due to a graded intake of water. But no agglutinations occurred in ordinary diluted or concentrated sea-water, or in sea-water to which MgSou, K, Ca, or Li salts had been added, or in sea-water to which various anesthetics had been added, or with eggs whose fertilization membranes had been removed, or in sea- water rendered alkaline. The NaCl in the above experiments must have affected the eggs in such a manner as to gelatinize their outer surfaces, and under these circumstances, when two or more eggs whose fertilization mem- branes were ruptured came in contact, they were agglutinated. Some of these agglutinated eggs subsequently separated, others remained aggluti- nated, developing into more or less perfect twins, while others fused together. I have never observed fusion during the egg stage, as reported by Haan. The fusion in both Arbacia and Toxopneustes occurred and was limited to the blastula and gastrula stages, as described in a previous publication. In these studies and in those of other workers in this field attention has been chiefly directed to the fusion of the ectodermal walls of the two embryos and to the fusion of their archentera. The skeleton seems to have been overlooked. Curiously enough the changes in the skeletons of two fusing larve indicate most clearly the nature of the regulatory phenomena during Experimentally Fused Larve of Echinoderms, etc. 107 this process. This paper will deal with the consideration of the changes in the skeletal structures in fusing larve. SINGLE PLUTEUS LARVA. Except where specifically mentioned to the contrary the drawings were made with the camera lucida and drawn to the same scale. Figure I is a lateral view of a single larva fully devloped. The figure represents the BiG 2: larva somewhat foreshortened, but with all of the parts, particularly the skeletal structures, outlined. Within the body of the larva are two ven- tral body rods, one on each side of the body. From these rods two long bars extend into the ventral arms. At right angles to these, and arising from the union of the ventral body rods and the ventral arm rods, are two ventro- dorsal connectives, which in turn bifurcate, one branch extending into the dorsal arm and the other aborally into the body to form the dorsal body rod. From the aboral end of the ventral body rods are two usually bifurcated aboral branches. Finally, there are two ventral branches between the ventral arms, but these are not shown in the figure. These bars are easily recognized from one another in both the single and fused larve, and they afford a definite and measurable index of the changes during the fusion of the larve. Figure 2 is another single larva drawn under greater magnification, It shows, from a foreshortened ventral view, the same characteristic parts of the skeleton, and also illustrates the small knobs and processes on the ventral rods, which are sometimes not reproduced in the drawings. Figures 3 and 4 are enlarged to the same scale as figure 2 and represent the ventral and lateral views of the skeleton. Figure 4 affords a particularly good example of the spinous ventral body rods. 108 Papers from the Marine Biological Laboratory at Tortugas. FUSED “PEUTET. Agglutinated or “true twin” larve are of no special significance for my present purpose, for each is perfectly developed and no reciprocal influence occurred. Figures 1 to 3 correctly represent any such agglutinated larva. It is only in fused larve, 1. e., larvae with a common body cavity and a com- mon and continuous ectodermal wall, that reciprocal changes are observed. Fic. 3: In the following drawings, for the sake of simplicity, I have not always indicated the whole skeleton when such omitted portions are normal and when the symmetrical right or left side is fully represented. Figure 5 is a partial fusion of two larve with a common body cavity and a continuous body wall. Each larva is completely differentiated and normal. The archentera are complete and independent, though they appear to be super- imposed in this particular view of the fusion. The skeletons are typical and the parts are complete and characteristic of the normal single larva. On closer examination it is found that certain parts have not developed or have not reached their full size, such as the left dorsal rods of the right or A pluteus, and the left dorsal arm of the left or B pluteus. Associated with this suppression or retardation of the dorsal rods there is a compen- satory growth of the left dorsal arm of the B pluteus. This supernumerary arm bifurcates, one branch extending into the right pluteus, the other branch into the left pluteus. And though the skeletal rods of the two larve are in close proximity and overlap in several places, there is no trace of a fusion of the rods. This suppression of the dorsal rods and the associated compen- Experimentally Fused Larve of Echinoderms, etc. 109 satory growth of other parts of the skeleton were observed in varying degrees of complexity in all or nearly all the fused larve. The two plutei of figure 6 are fused more completely together than the preceding pair of plutei. The two larve are perfect and the parts are Here A: 7 FIG. 5. typical and independent, except the body wall, which is common to both. In the right or A pluteus each of the skeletal parts is differentiated and normal. In the left or B pluteus the ventral body rods are incomplete in Fic. 6. so far as they do not possess their aboral bifurcated branches, and one of the ventral arm rods is abbreviated. The B pluteus in this fusion is incomplete, not with respect to the dorsal rods, as in the previous example, but in the aboral branches of the ventral body rods. 110 Papers from the Marine Biological Laboratory at Tortugas. It is well known that in the development of the skeleton the mesenchyme cells differentiate two triradiate spicules, one on each side of the body. The bars of the triradiate spicules develop unequally. The ventral body rod, the ventral arm rod, and the dorso-ventral connective are formed from the outgrowth of the original three bars. Subsequently the dorsal body rod and the dorsal arm rod are differentiated, and finally the aboral branches of the ventral body rods. In the two examples of fused larve just described, it was found that, when skeletal parts are suppressed, such suppression is limited to the parts last to be differentiated, namely one or both dorsal rods and the aboral branches of the ventral body rods. That this suppression was permanent can readily be demonstrated. It will be recalled that all parts of a pluteus are fully formed within 72 hours, yet these fused plutei, after a period of 4 to 7 days, had not differentiated their missing rods. B FIG. 7A. FIG. 7B. The last fusion does not offer such clear evidence of compensatory growth as the preceeding fusion; it is probably limited to the extremely elongated dorso-ventral connectives of the B pluteus. In the last two instances the fused plutei were equal or nearly equal, both in size and in the stage of development. Ina large number of instances, however, the two plutei are unequal in both these respects and in varying degrees of inequality. For example, figures 7A and 7B are two views of the same pair of fused plutei, of which figure 7A is somewhat enlarged. In these figures the A pluteus is clearly larger and more completely differenti- ated than its mate; the four arms, the shape of the body, and the archen- teron are completely differentiated and normal. The B pluteus appears to have but one abbreviated arm and a diminutive archenteron. The skeletons show very clearly the unequal development of the two plutei. The A pluteus contains all the skeletal parts of the normal pluteus, Experimentally Fused Larve of Echinoderms, etc. III and their shape and size and relative positions are essentially normal. The skeleton of the B pluteus is far from complete. It contains the two ventral body rods and two dorso-ventral connectives which are quite normal, but the two dorsal body rods, one of the ventral arm rods, and one dorsal arm rod are very much reduced in size, and there is no trace at all of the aboral branches. The compensatory growths of the skeletons in this fusion are very striking indeed. They are altogether limited to the A pluteus and confined to processes on the right ventral body rod at H' and on the left ventral body rod at H?, where the outgrowth is much branched and extends into the B pluteus. The formation of these hypertrophied and supernumerary bars is explicable on the assumption that when the mes- Fic. 8. FIG. 9. enchyme cells were prevented from developing within zs own larva they migrated into the adjoining or A larva and there formed the additional skeleton at or near the center of active skeleton building, namely, the region of the union of ventral rods and dorso-ventral connectives. Figure 7B is a dorsal view of the same pair of fused plutei, which shows more clearly some of the points referred to. Figure 8 represents two plutei much more completely fused into one body. The dominant or A pluteus is typical and complete; the suppressed or B pluteus is smaller and incomplete, and this inequality is shown par- ticularly well in the skeletons. It will be observed that the B pluteus is deficient in these skeletal parts that were missing or abbreviated in the previous examples, namely, the dorsal arm rods and the aboral ventral 112 Papers from the Marine Biological Laboratory at Tortugas. branches. The other parts are perfectly differentiated. If the hypothesis of the migration of the mesenchyme cells from the suppressed to the other larva is correct, one would expect in this fusion correspondingly smaller compensatory growth to have taken place; this is found to be the case, namely additional processes on the ventral body rod shown at / and an elongated dorso-ventral connective. A rather interesting instance of form equilibrium is seen in the A pluteus, where the left dorsal arm is absent and the right has formed an accessory arm rod and dorso-ventral connective. This fusion affords particularly strong evidence that the skeletons of the two plutei do not fuse in the ordinary acceptance of that term, for the overlapping of the parts is here very considerable. In figure 9, the inequality of the two fused plutei is very great. The A pluteus is perfect in every detail, the B is very incomplete. In the first place, there is but one in place of four arms, and the gut is fused to the foregut of the A pluteus. The skeleton consists of two ventral body rods, only one of which possesses an aboral branch; two dorso-ventral connectives, and one large and one very small ventral arm rod. There is no trace of the dorsal arm or body rods. As in previous instances, such suppression of the skeleton is associated with compensatory growths, which, in this instance, is confined to the supernumerary ventral body rod and to a marked thickening of the other ventral body rod of the B pluteus. The fusion shown in figures 10A (somewhat enlarged) and in I0B is particularly interesting. It resembles figure 9 in so far as the A pluteus is complete and true to type, while the B pluteus is very incomplete. The dominant pluteus has all the skeletal parts, the other (B) has but one ventral arm and an atypic archenteron. The skeleton of this pluteus consists of only three out of at least twelve parts, namely one ventral body rod, one ventral arm rod, and one dorso-ventral connective. Over three-quarters of the skeleton has not been differentiated. The compensatory growth of the skeletal structures is to some degree commensurate with the degree of incompleteness or suppression. Close examination of figure 10 will show that the right dorso-ventral connective is not the normal single rod, but is much branched, and that from it a very long bar extends across the body of A into the body of B; it reaches almost as far as the ventral body rod of B, and might easily be mistaken for its symmetrical mate. Figure 10B affords another view of this fused larva and shows in a more striking manner some of the facts just referred to. In passing, it might be noticed that the three short rods derived from the dorso-ventral connectives of B are probably other supernumerary or com- pensatory outgrowths of the skeleton. The extension of the large supernumerary (S) rod across one pluteus and into the other might be accounted for in one of two ways. Either it was formed originally as a ventral body rod in the B pluteus and secondarily united with the skeleton of the A larva at its dorso-ventral connective, or Experimentally Fused Larve of Echinoderms, etc. 113 it was an hypertrophied dorso-ventral connective of A which grew beyond the limits of its own body into B. The matter can be settled conclusively only by continuous observation of the living fusion, which was not made in FIG. IOA. FIG. IOB. the case of this particular fusion. The preceding and certain of the follow- ing fusions make it altogether probable that a fusion of the two skeletons did not take place, but that the mesenchyme cells migrated from one part of the body to another, as in figures 5, 6, 8, and 9, or from one body into the other, as in figures 7 and 10. On comparing figures 7 and 10, ic will be observed that the supernumerary rod in both originated in the same region, is much branched, extends from one larva into the other, differing only in the matter of length, and finally that they replace the suppressed parts of the skeleton of the weaker larva. The close similarity in these respects is significant and supports the view of a redistribution of the mesenchyme cells and the subsequent outgrowth of the hypertrophied rods into the second larva. Figure 11, which is somewhat enlarged, might be mistaken for a single larva with a supernumerary arm. On closer examination it will be found that there are clearly two larve fused very completely together. One of these, the dominant or A pluteus, has the characteristic four arms, perfect 8 114 Papers from the Marine Biological Laboratory at Tortugas. archenteron, etc. The other or B pluteus is represented externally by a single arm. This arm, however, contains a ventral and a dorsal arm rod united by a dorso-ventral connective, from which a characteristic ventral and a dorsal body rod extend into the common body cavity. The arm of B is really the right or left half of the B larva and if there be any doubt as to the dual character of figure 11, the diminutive archenteron in the B larva should at once set it aside. There are several instances of compensatory growths in the dominant larva. There is the hypertrophied dorso-ventral con- nective, and better still the dorsal body rod which is not merely considerably longer but extraordi- narily thicker than in the con- trols, and finally an elongated left dorsal body rod. It should again be noted that although there is_ considerable overlapping of adjoining skeletons there is no fusion at any point. In the fusions of the larve so far described, the archentera were either independent or partially fused. In the following examples the two larve are more completely fused together, and there is but one archenteron, the one belong- ing to B having been either sup- FIG. 11. pressed or fused completely in that of A. Figure 12 is an example of a very complete fusion of two larve with but one archenteron. Detailed inspection shows that one larva is complete and perfect, while the other has but a short blunt arm as the external evidence of the second larva. Within this arm there is a normal and char- acteristic ventral body rod with a number of hypertrophied processes, a true dorso-ventral connective and an unusually thin ventral arm rod, and an accessory dorsal arm rod. In this larva the right or the left half of the larva has been differentiated, but less completely than the half larva of figure IT. Figure 13 resembles the fusion shown in figure 12 in several essential points. Each fusion has but one archenteron, a dominant and perfect Experimentally Fused Larve of Echinoderms, etc. 115 larva, and a very incomplete larva represented by a single broad arm, in which there is a ventral arm rod, a ventral body rod, and a dorso-ventral connective. In figure 13 there are two thin, straight bars—one long, the other very short, neither of which connects with either the B or the A skeletons. If it be assumed that retrogressive changes have taken place and that these have affected one or both ends of certain rods, we should expect the connections between rods to disappear, the distal ends to be shortened, and the bars themselves to be thinner than corresponding rods of control larve. All these results are found in figure 13 and the inde- pendent bars probably represent the former dorsal and ventral arm rods. vy} Fics 12. In figure 14 form equilibrium is more closely established. Larva B is reduced to a small, very blunt protuberance on the left side of the figure. It consists internally of a ventral body rod and its aboral branch, a very much dwarfed ventral arm rod, a dorso-ventral connective, a dorsal body rod, a supernumerary rod at s and a mere rudiment of the dorsal arm rod. In other words only one half of one side of the skeleton has been formed in this larva. The gradual decrease in size and contents of the B larva, seen in figures II to 14 inclusive, reaches an extreme state in figure 15; and if it were not for the intermediate stages already described, figure 15 might easily be overlooked or mistaken for a single larva. In figure 15 the A larva contains all the parts in their proper relative positions typical of the control larva. The gut is somewhat enlarged, ~ indicating the incomplete fusion of the two archentera. The skeleton is single and perfect and to describe it is to describe the control larva. The B 116 Papers from the Marine Biological Laboratory at Tortugas. larva is represented only by a short, blunt arm containing a single straight rod, probably a ventral arm rod, which passes the ventral body rod of the A larva and ends blindly in the common body cavity. This single rod is less than one-twelfth of the complete skeleton whose parts have either been suppressed or disintegrated after their formation. If either or both of these FIG. 14. Fic. 15. processes were to continue beyond the stage shown in this figure the blunt arm would disappear with its contained single bar and all trace of the B larva would be gone. It would then be absolutely impossible to distinguish such a larva from the non-fused larve in the rest of the culture. DISCUSSION. In a previous publication I have shown that two or more blastulz or gastrulz derived from separately fertilized eggs, when united, tended to be remolded into a single gastrula somewhat larger than the controls. In this paper I have shown that this process of form regulation continued through the larval period, during which the tissues were completely differentiated. Such fusing larva tended to be remolded into a single one of normal size, but this tendency is conditioned by certain factors, of which the following are the most important: (1) Area of the agglutinated surfaces of the two blastule: When the de- veloping blastulz are merely agglutinated by an inconsiderable area of their surfaces, each develops into a gastrula and a pluteus—complete, perfect, Experimentally Fused Larve of Echinoderms, etc. EL? and independent of the other except for the common surface of attachment. There is no reciprocal influence of one upon the other. When, however, the agglutinated surfaces extend over a large area the blastule tend to fuse, to have a common blastoccele, and the gastrule tend to approximate the form of a single embryo. It can not be stated definitely which is cause and which effect, whether greater common surface makes for more complete fusion or whether the internal forces making for fusion also affect the surfaces of the agglutinated embryos. (2) Mass of agglutinated or fused embryos: When more than four or five eggs are agglutinated the disintegration of one or more of them at an early stage in their development brings about the disintegration of the entire cluster before they can fuse. When three blastule are agglutinated they may develop into three perfect and practically independent larve, or they may fuse together, under which circumstance the reciprocal influences are exceedingly complex and the resulting larva is quite atypic. For this reason I have omitted all consideration of such complex fusions and limited myself to fusions of two eggs, so that the regulative changes may be observed with the least number of disturbing factors. (3) Differential rate of development: When two gastrula are fused and each has developed at an equal rate, twin larve like that shown in figures 5 and 6 result, in which case each larva is practically complete and equal in size, and has little or no influence upon the other. When two gastrula have not developed equally the conditions are then most favorable for form regu- lation. At least two factors are involved in form regulation, namely, the sup- pression of structures and the disintegration of those already differentiated. I have elsewhere shown, from the study of living fusions, that the archenteron of one of the fused gastrule may be entirely inhibited or may never develop beyond an early stage of its differentiation. Driesch and de Haan have also observed aborted or dwarfed archentera. I have shown in this paper that certain parts of the skeleton also may never be developed, that such suppressed parts are nearly always the last to be differentiated and confined to but one of the fused larve. It has been urged by de Haan that the suppressed larva is a sick larva, that there is no regulation but that a union of a healthy and an unhealthy larva occurs. Under such circumstances we should expect the sick larva to be atypic and some relation between diminutive size and degree of ir- regularity in development. But the fused larve are quite perfect and often of normal size. It might be urged that only very sick larve would be atypic and would not complete their development. But how can one account for the lack of development of the right or the left sides of the skeleton on such an assumption? It would also be exceedingly difficult to account for the absence of the oral half of one side (fig. 14), particularly in the face of the fact that the skeletal parts that are differential are characteristic and normal and show no trace of irregularity or sickness. 118 Papers from the Marine Biological Laboratory at Tortugas. Since the suppression of skeleton, arms, and archentera, etc., do not occur in control larve, it seems reasonable to suppose that the more rapidly developing larva influences the other in such a manner that the mesenchyme cells of the latter are prevented from completing the skeleton, and the degree of suppression seems to be related to the differential rate of de- velopment, the greater the difference the greater the suppression of the parts of the slower developing larva. The second factor making for form equilibrium is the loss of parts that have already been differentiated. I have shown from the study of living gastrule that the archenteron in the slower developing gastrula in a fusion may be reduced in size and completely disappear. The skeletons may also actively degenerate in one of the fused larve. This degeneration is evi- denced in the reduction in thickness of various rods, in the disintegration of one or both ends of rods, or both of these processes. As a result of the incomplete development of the skeleton and the subsequent disintegration of parts that have been formed the skeleton may ultimately consist of but a single or a few independent rods. And since the size of the body is so dependent upon the size of the skeleton the suppressed larva may decrease in size until it almost disappears, as shown in figure 15, and I believe may completely disappear. What happens to the tissues of the degenerating larva? Are they ab- sorbed as food by the dominant larva as nurse cells supply egg cells with nutriment; or are they disintegrated after the manner of certain grafts without any influence upon the host; or are they transferred and recon- structed within the body of the dominant larva? There is unmistakable evidence that two of these processes ordinarily take place, but in unequal degree and affecting different structures of the organism. The transfer and rebuilding of the materials of one larva into the other are limited to the last differentiated tissues, namely, the skeleton. It was shown that when two fused larve developed unequally the dominant pluteus may prevent the completion or the development of parts of the skeleton; that under these circumstances the unused mesenchyme cells may migrate to other regions of the body or into the body of the dominant larva and there give rise to additional skeletal material, either by the thickening or elonga- tion of the rods or by the growth of supernumerary processes and rods. The extent of such hypertrophied or supernumerary growths was commen- surate with the suppression of skeletal parts in the other regions. De Haan seems to have shown that two eggs may be completely fused into one somewhat larger than the control. Driesch, Goldfarb, and de Haan have demonstrated that blastule may be completely fused together into a single larger blastula. The evidence is not so conclusive but seems to point to a complete fusion of gastrule and the enlargement of the gut as well as the body. Beyond this stage in the differentiation of the tissues a complete fusion of the organs of the two larve does not take place, at least in Toxopneustes . Experimentally Fused Larve of Echinoderms, etc. 119 variegatus. The skeleton under no circumstance fuses in the two individuals and there is no evidence of an enlargement of the larva to form the “‘ein- heits pluteus’’ or completely fused larva of Driesch. There is, on the contrary, definite evidence that disintegration of skeletal materials takes place in the suppressed larva without any corresponding enlargement of the skeleton or the body of the dominant larva. The dominant larva in figures 12, 13, 14, and 15 are no larger than in figures 9, I10B, 5, 6, 7B, and 8—all drawn to the same scale—or figures 11, 10B, and 7A, drawn to greater magnification. In the second place the dominant larva in these fusions are no larger than the corresponding control larva, and the variation in size is approximately the same in both groups. There are no giant larva in the sense employed by Driesch and Morgan and other investigators. There are completely fused larva as there are completely fused blastulze and gastrule. The facts clearly show that in the blastula, gastrula, and larval stages the less or earlier differentiated tissues (such as ectoderm and entoderm) after their union to form the giant body and giant gut diminish in size until a normal single body and gut are approximated. This diminution must occur either as a consequence of the degeneration of certain cells or the diminution in size of the cells. The latter can readily be shown not to be the case. It can also be shown that the degenerating substances in one larva do not serve as food or stimulant of the other, for the dominant larva are no larger nor do they develop any faster than the controls. In fact they are considerably slower than the controls. The increased growth occurs not in any increase of the larval body or of the contained organs, but in the hypertrophied and supernumerary processes and bars whose origin and development have been described. SUMMARY. A new method was discovered by which large numbers of eggs of Toxop- neustes variegatus could be fused together, during the blastula and gastrula stages. The method consists essentially in placing the fertilized eggs in a solution composed of one part of sea-water to three parts of either an isotonic or a slightly hypotonic NaCl solution. This solution has two effects upon the eggs, namely, it increases the volume of the egg and thereby bursts the fertilization membrane, and it gelatinizes the surface of the egg. When any such eggs are in contact they become agglutinated and may remain agglutinated and produce twin larvee or fuse together during the blastula or gastrula stages. The conditions that favored the fusion process were: (1) The number of agglutinated eggs: More than two eggs rarely gave rise to a fused and characteristic larve. When two eggs were agglu- tinated they either developed into twin larve or fused more or less completely together. 120 Papers from the Marine Biological Laboratory at Tortugas. (2) Time of fusion: The earlier in development that fusion occurred the more completely did the parts tend to fuse together and vice versa. (3) Area of common surface: The greater the surface common to both eggs or blastule, the more completely and the earlier did they tend to fuse, and vice versa. (4) Differential rate of development: When the two embryos developed equally fast, each was essentially perfect, though fused to the other. When one developed less rapidly than the other, reciprocal changes took place, as a result of which the larve tended toward complete fusion and form equilibrium. Form equilibrium is determined by at least these three factors, namely: (1) Differential rate in the development of the two eggs. (2) An inhibition in the development of certain structures. The skele- ton in the more slowly developing larva was rarely complete, lacking those portions of the skeleton that were last differentiated. Such portions may include one-half, three-quarters, or more of the skeleton. The parts that do appear are perfectly normal. (3) A disintegration of certain structures. Skeletal parts of the larva after their complete differentiation may break down and completely dis- appear. Such decrease in the skeleton was accompanied by a decrease in the corresponding parts of the body of the larva. Such regulative changes are nearly always limited to the more slowly developing larva. They result in the complete atrophy of this larva without any reciprocal or compensatory effect upon the shape, size, or parts of the more rapidly developing or dominant larva, except for the skeleton. The ectoderm and the entoderm of one larva may fuse with that of the other, and may subsequently be absorbed or be disintegrated without any visible effect upon the other larva. The skeleton never fused. When it was suppressed in one larva, some or all of the unused mesenchyme cells migrated to other regions of the same larva, or into the adjoining larva, and there gave rise to hypertrophied or supernumerary parts of the skeleton, and in certain instances replaced in part the missing skeleton of the sup- pressed larva. BIBLIOGRAPHY. DE Haav, J. A. B. 1913. Ueber homogene und heterogene Keimverschmelzungen bei Echiniden. Archiv f. Ent., 36, p. 473. Driescu, H. 1900. Studien iiber das Regulationsvermogen der Organismen. Part 4. Die Ver- schmelzung der Individualitat bei Echinodermkeimen. Archiv f. Entw., 10, p. 411. 1910. Neue Versuche iiber die Entwickelung verschmelzener Echinodenkeime. Archiv f. Entw., 30, p. 8. GOLDFARB, A. J. 1913. Studies in the production of grafted embryos. Biol. Bull., 24, p. 73. HERssT, C. 1900. Ueber das Auseinandergehen von Furchungs und Gewebezellen in kalkfreiem Medium. Archiv f. Entw., 9, p. 425. Logs, J. 1894. Ueber eine Einfache Methode, zwei oder mehr zusammengewachsene Embryonen aus einem Ei herforzubringen. Pflugers Archiv, 55, p. 525. 1895. Beitrage zur Entwickelungsmechanic der aus einem Ei enstehenden Doppel- bildungen. Archiv f. Entw., I, p. 453. 1908. Ueber den Mechanismus der Agglutination. Zeitsch. f. Chemie, 3. 1909. Ueber die chemischen Bedingungen fur die Entstehung zwillinge beim Seeigel. Archiv f. Entw., 27, p. 119. Moreau, T. H. 1895. The formation of one embryo from two blastule. Archiv f. Entw., 2, p. 65. I2I VIL. BADERIMENTS ON THE PERMEABILITY OF CELES: BY d. F.sMCGLENDON, Instructor in Physiology, University of Minnesota Medical School. Three text-figures. 123 4.) Ova EXPERIMENTS ON THE PERMEABILITY OF CELLS. By J. F. McCLenpon. One of the most important steps in the analysis of life was the discovery of oxygen. Ever since that time it has been known that animals absorb free oxygen and give it out in a combined form. The discovery of enzymes has given a clue to the rapidity of oxidation within cells; but enzymes are always present in living substance, and yet oxidation may vary from almost zero in certain eggs and seeds during rest to a high rate during activity. We know that oxidation within muscle is greatly increased during contraction, yet hardly anyone would suppose this to be due to the sudden increase of oxidases. Loeb, however, did use this explanation for the increased oxidation of the egg on fertilization, by sug- gesting that the sperm carries an oxidase into the egg. Certainly the sperm contains an oxidase, as does every other cell, but the sperm is minute in com- parison with the egg, and no one has found any indication of an unusual amount of oxidase in the sperm. Furthermore, the egg may be made to divide by artificial means without the introduction of any enzyme. An organism may be caused to absorb more oxygen by any one of a number of different stimuli. But what is stimulation? What is the first change induced in the organism or excised organ by all stimuli? Pfeffer showed that stimulation of sensitive plants causes a change in permeability of certain cells (of the pulvinus). This change in permeability causes a change in turgor (internal osmotic pressure) and movement results. The question arises: what relation does this bear to oxidation? According to Ralph Lillie oxidation is suppressed in the cell by the accumulation of waste products to which the resting cell is impermeable. Stimulation increases the permeability of the cell to them and consequently oxidation is increased. Lillie maintained that on fertilization the egg became more permeable, and linked all subsequent phenomena with this change. Not finding any other explanation of the stimulating action of the sperm, I attempted an exact study of the permeability of the egg. EXPERIMENTAL. Three methods of procedure were followed: (1) the use of cell masses as partitions (on eggs of Lytechinus); (2) the use of quantities of eggs sus- pended in a liquid medium (on eggs of Fundulus); (3) experiments on individual eggs (of Arbacia). 125 126 Papers from the Marine Biological Laboratory at Tortugas. The first of these methods was tried at the Tortugas laboratory on the eggs of the sea-urchin Lytechinus (Toxopneustes) variegatus. Advantage was taken of the fact that the electrolytes of the eggs and the medium dissociate into ions which bear electric charges, and therefore their move- ment through the eggs could be detected, with a high degree of accuracy, by the electric conductivity method of Kohlrausch.! Several bushels of sea-urchins were collected each day and the eggs of the ripe females placed in large dishes of sea-water. The mucus egg membranes (jelly, zona pellucida) were washed off and the eggs were intro- duced into a conductivity vessel made especially for the purpose (fig. 1). The conductivity vessel was placed in the centrifuge and the eggs were precipitated until they were closely packed together. By microscopic examination it was found that only a trace of sea-water was left in the spaces between the eggs (fig. 2). By raising the electrodes above the level of the eggs they came to lie in the supernatant sea-water that had been pressed out from between the eggs. In this way the conductivity of sea- water and egg-mass could be measured separately within two minutes. The conductivity of the egg-mass, when moderately centrifuged, was about one-twentieth of that of the sea-water, in- dicating that the conductivity of the egg itself must be almost nil. The conductivity of the egg-mass was greatly affected by the degree of packing. However, the vessel was so long and narrow that the packing of the eggs could be recorded accurately by mark- ing their upper limit on the side of the vessel (fig.1,a). Theeggs could then be removed from the conductivity vessel, replaced, and centrifuged to the same level, when the con- ductivity was found to bethesame. In this Picx: Fic. 2. way I was able to measure the conductivity before and after fertilization and before and after artificial stimulation causing cleavage. It was found that the conduc- tivity increased about one-fourth when the developmental processes began. The sources of error and methods of controlling them are discussed in another paper.’ One point, however, the elimination of the fertilization membrane, needs further consideration. Soon after the egg is fertilized, a fine line appears close around it and gradually expands until it forms a circle, widely separated from the egg. Almost every one (except Kite) supposes this to represent a spherical membrane, the ‘fertilization membrane.’’ Before fertilization the egg is covered by a thick transparent jelly or mucus layer, called the chorion or zona pellucida by some authors. I stained this jelly with basic dyes (neutral red and methylene blue). These dyes coagulate the jelly, causing it to become membranous. 1Ostwald-Luther. Messungen. 2McClendon. Amer. Jour. Physiol., vol. 27, p. 240. Experiments on Permeability of Cells. 127 Fic. 3. in these experiments all toxic substances were eliminated from the water by redistilling it in the apparatus shown in figure 3. The stock bottle (to the left) is filled with distilled water, which siphons over into the beaker and keeps the latter filled up to the mouth of the wide tube. A small flame under the beaker boils the water and drives out all dissolved gases (since the water contains no salts to hold back COse, and no non-volatile substances to liberate ammonia). The boiling water from the beaker siphons over into the Kjeldahl flask over a large flame. If the siphon be drawn out into a fairly small opening the Kjeldahl flask may be set lower than the beaker, as in the figure; but this has the disadvantage that when the still is cooled the Kjeldahl flask fills up to the neck and makes it inconvenient to start again. The siphon has a bulb blown on it at a, to stop the mouth of the flask. It is not necessary that this stopper be ground in, provided it fits fairly snug, as water condenses on the stopper and seals the crack. A hole is blown in the side of the neck of the flask to admit the condensing tube. This hole may be reamed out with the end of the tube itself, while the glass is hot, to insure a fairly close fit. The condensing tube is of fused quartz and the only water that collects in the receiving bottle is condensed on the quartz. The cold water dissolves glass very slowly, so that it is permissible to use a glass receiving bottle if the water is distilled fresh each day. Hard glass (Jena or Bohemian) dissolves more slowly but contains heavy metals which are toxic even in the minute quanti- ties that dissolve away in the cold distilled water, even though the glass be previously steamed out. Soft glass liberates traces of sodium, calcium, and potassium, which render the water less toxic, but which, if present in sufficient quantity, may be a source of error in the experiment. In a previous paper! I concluded that the fertilization membrane is the result of a mutual coagulation and precipitation between the jelly and another transparent colloid filling the space between it and the egg. I found that if the jelly be removed from the egg no fertilization membrane can be formed. Elder? subsequently made the same observation and came to the same conclusion. No matter what is the nature of the membrane, ?McClendon. Science, vol. 33, p. 387, Mar.10, 1911. *Arch.f. Entwicklungmech., vol. 33, p. 143, I912. 128 Papers from the Marine Biological Laboratory at Tortugas. the fact that it can be suppressed is of importance in my technique, since the presence of the membrane might be considered as a source of error. The increased conductivity of the eggs indicates increased permeability to electrically charged particles, since electrons can not pass in the free state through solutions without very soon attaching themselves to atoms. The particles concerned must be ions, since colloidal particles move slowly and carry little electricity per unit mass, and could hardly cause the great in- crease in conductivity observed. This increase in permeability was confirmed by plasmolytic experiments on the eggs of Arbacia at Woods Hole. What ions are concerned in the increase in permeability has not been determined. Experiments which I made on the Fundulus egg are sugges- tive. Fundulus eggs (which normally develop in sea-water), when develop- ing in distilled water, do not give up chlorides to the water. If placed in sulphates or nitrates of sea-water metals, they still do not give up chlorides. If placed in distilled water they do not give up magnesium (or calcium). But if placed in sodium salts they lose magnesium (and probably calcium and potassium), and chlorine, because the pure sodium increases the permeability. A critical review of the literature on permeability is given in a previous paper? but I shall here refer to some of it briefly. In the summer of I910 R. Lillie, Lyon and Shackell, E. N. Harvey, and the author, all working on the sea-urchin’s egg, had come to the conclusion that its permeability increased on fertilization. Very shortly there appeared a paper by Loeb?’ in which he made this statement: In former papers, especially in a book published a year ago, I pointed out that the process of membrane formation, or a certain alteration of the surface of the egg, is the essential cause of the development of the egg; and I pointed out, also, that this alteration of the surface might increase the permeability of the egg, especially for hydroxy] ions. I have found one such paper,‘ but it was directed against the view that the cell is impermeable to electrolytes, and gave evidence to show that the egg is more permeable to salts than to sugar. In the book referred to® Chapter x11 deals largely with permeability and ends with the following conclusion (on page 118): Wir kommen also zu dem Schluss, dass die Hydroxyloinen des Seewassers auch in das unbefruchtete Ei diffundieren, und dass die Membranbildung des Oxydationen im Ei nicht dadurch steigert, dass sie die Durchgangigkeit desselben fiir Hydroxylionen erhéht, sondern dass sie die fiir die Entwicklung nétigen Oxidationen auf anderem Wege entfesselt oder mdoglich macht. If Loeb has found evidence that the permeability of the egg increases on fertilization, or fatty acid treatment, I would be very glad to have my 1 Amer. Jour. Physiol., vol. 29, p. 295, Appendix II. 2McClendon. Biol. Bull., vol. 22, p. 113. 1912. 3 Loeb, J. Science, n.s., vol. 32, p. 412. Sept. 23, 1910. Loeb. Univ. of California Publications, Physiol., vol. 3, No. 11, p. 81. 1908. & Loeb. Die Chemische Entwicklungserregung des tierischen Eies. Berlin. 1909. Experiments on Permeability of Cells. 129 attention called to it, but it is difficult to find a paper without a reference to the journal in which it is published. The degree of permeability is of importance in development. R. Lillie maintains that too great an increase in permeability causes cytolysis of the egg. Gray! attempts to explain the abnormality of Echinoderm hybrids by assuming that the foreign sperm produces the wrong degree of perme- - ability and this leads to some of the chromatin becoming pathological and being thrown out of the spindle. The changes in permeability are probably due to changes in the aggre- gation state of colloids: either proteids, lipoids, or proteid-lipoid combina- tions or associations. Many investigators have looked for such changes with the microscope. That some change takes place at the surface of the egg on fertilization has long been known, but opinions differ as to its exact nature. Spherical concentric surfaces, when viewed tangentially with the microscope, cause optical illusions, and the real structure can not be exactly determined without knowledge of the refractive indices of the substances composing it. Kite, however, claims that a thin layer may be observed on the surface of the unfertilized egg and, on fertilization, this layer absorbs water and swells. He claims to have stained it differentially.2. I do not understand, however, his observations on unstained eggs. He says that this layer which he calls the vitelline membrane becomes more distinct (of different refractive index) from the sea-water on swelling, whereas it is well known that all hydrophile colloids, on swelling, become less distinct and more nearly of the same refractive index as the medium. The swelling of this layer might cause an increase in permeability, since the swelling of all hydrophile colloids increases their permeability to water- soluble substances. Lillie, however, believes that a reversible increase in permeability takes place in the superficial layer of the cytoplasm (plasma membrane) and J. Gray who has confirmed my determination of the in- crease in permeability of the egg finds that impermeability returns in 15 minutes, whereas Kite’s vitelline membrane remains swollen. I consider Kite’s vitelline membrane a new name for a colloid postulated by Loeb. Mitchell and the author® studied experimentally the role of hydroxyl ions in the increase in oxidation of sea-urchin’s eggs, which follows fertili- zation or artificial stimulation. We concluded that the increased perme- ability allows the hydroxyl ions of the sea-water to penetrate the egg. This increase in hydroxyl ions within the egg might increase oxidation. CONCLUSIONS. The permeability of the egg to ions and perhaps some other substances increases on fertilization. The unfertilized egg is perhaps in a dormant condition and the increase in permeability probably allows a rapid inter- change with the surrounding medium necessary for activity (development). 1Gray, J. Proceedings of the Cambridge Philosophical Society, vol. 17, p. I, 1913. 2 Kite. Science, n. s., vol. 36, p. 562. I912. § Jour. Biol. Chem., vol. 10, p. 459. 9 130 Papers from the Marine Biological Laboratory at Tortugas. Whereas this supposed significance of permeability has not been proven, the sea-urchin’s egg is not an exception. Spores, eggs, seeds, and pupz are usually inclosed in relatively impermeable membranes. A striking example is the cocoa nut. Whereas an active organism may be partly inclosed in a relatively impermeable skin, great activity is always associated with a rapid interchange with the medium, and this is possible through more permeable portions (lungs, gills, kidneys, gut). The relation of permeability to oxidation can hardly be determined until more is known about the mechanism of animal oxidations. These seem to depend on structure since complete oxidations cease when structure is completely destroyed. Reference is made only to oxidations resulting in the formation of CO., Oxidizing enzymes such as tyrosinase, which are inde- pendent of structure, do not completely oxidize the substances acted on. VIII. HE RELATION BETWEEN THE RATE OF PENETRATION OF MARINE TISSUES BY ALKALI AND THE CHANGE IN FUNCTIONAL ACTIVITY INDUCED BY THE ALKALI BY E. NEWTON HARVEY, Instructor in Physiology, Princeton University. One figure. 131 THE RELATION BETWEEN THE RATE OF PENETRATION OF MARINE TISSUES BY ALKALI AND THE CHANGE IN FUNCTIONAL ACTIVITY INDUCED BY THE ALKALI. By E. NEwton HARVEY. OBJECT AND METHOD. The present study, made at Tortugas in the summer of I9QII, is a con- tinuation of permeability investigations undertaken at Columbia Univer- sity in I910 to 1911.!. My aim has been twofold. First, to compare the permeability of the cells and tissues of salt-water organisms with those of fresh-water forms. Second, to determine the relation between the rate of penetration of the alkali and the appearance of structural or functional changes in the cell. Rate of penetration of the alkali may be best deter- mined by the color change of some indicator within the cell. As in previous work, neutral red was used for the purpose. It is non-toxic, readily staining the living cell, and the color change (of red in acid or neutral to yellow in alkaline solution) is sharp and well marked. Neutral red was first made use of as an indicator in studying perme- ability by Bethe.2 Meduse stained in the dye become a deep orange-red, which changes to bright red in HCI and to yellow in NaOH. Warburg? found that sea-urchin eggs stained in neutral red were entered readily by NH,OH but not by NaOH. My own experiments‘ gave similar results for another species of sea-urchin (Toxopneustes variegatus) and for Paramecium, Elodea, and Spirogyra cells. The penetration of the following alkalies was studied: 1. Weakly Dissociated: 2. Strongly Dissociated: Ammonia—NH,0OH Tetraethylammonium hydroxide— Methylamine—NH;CH;0H N(C2Hs) 4OH Dimethylamine—N H2(CH;)20OH Sodium hydroxide—NaOH Trimethylamine—NH(CH;);0H Potassium hydroxide—KOH Ethylamine—NH;C2H;OH Strontium hydroxide—Sr(OH)2 Normal propylamine—NH;C:H;CH20H Isopropylamine—N H;(CH;)2CHOH The addition of the stronger alkalies to sea-water precipitates the Mg so that a Mg-free sea-water containing 0.625 m. (100 NaCl + 2.2 KCl + 2.5 CaCl.) was used to dissolve the alkali. Such a solution is a fairly well- balanced medium for marine tissues, being only slightly inferior to van’t Hoff’s solution. The salt present in solution with the alkali affects also the rate of entrance of the latter, so that it is important to use always the 1 Harvey, E.N. Jour. Exp. Zool., vol. 10, p. 507. I9II. 2 Bethe. Pfluger’s Archiv, vol. 127, p. 219. 1909. 3 Warburg, O. Zeit. f. Physiol. Chem., vol. 66,p.305. I9I0. 4Harvey, E.N. Loc. cit. 133 134 Papers from the Marine Biological Laboratory at Tortugas. same type of salt solution for comparative results. As stated above, Mg- free sea-water of 0.6 m. concentration was employed. The magnitude of the difference between pure NaCl and a balanced medium is brought out in table 1. TABLE I. Color Color Egg. Concentration | change in | change in of NaOH. Mg-free 0.6 m. sea-water. NaCl. min. min. Toxopneustes variegatus........... N/100 Io to 12 5 Palolo-worm (Eunice fucata)....... N/80 12 5 Holothuria floridana (immature).... N/80 6 I PART I. THE PERMEABILITY OF MARINE TISSUES. PENETRATION RATE AND DEGREE OF DISSOCIATION. Just as in the case of fresh-water organisms (Paramecium, Spirogyra, Elodea), the weakly-dissociated alkalies (group 1) penetrate marine tissues almost instantly, the strongly dissociated only after an interval of 30 to 60 minutes. This statement applies to all the cells or tissues of salt-water organisms investigated, including the following: Holothuria floridana, immature eggs. Eunice fucata (palolo-worm), segmenting eggs. Holothuria floridana, respiratory tree. Eunice fucata (palolo-worm), trochophores. Hipponoé esculenta, mature eggs. Pomatostegus stellatus, mature eggs. Toxopneustes variegatus, mature eggs. Cassiopea xamachana, gonads. Toxopneustes variegatus, plutei. Cassiopea xamachana, subumbrella epithelium. Sabella sp., tentacles. Two unidentified species of marine alge. Thus, the anomalous result again appears—that it is the least dissoci- ated, the weakest group of alkalies, which penetrate living cells most rapidly. Two of the members of this group are considerably less dissoci- ated than the rest—ammonia and trimethylamine. We might expect them to penetrate most rapidly of allif rate of penetration bears an inverse relation to degree of dissociation. There is, however, no marked difference in pene- tration rate between ammonia or trimethylamine and the other amines,® but a well-defined difference in toxicity. TOXICITY AND DEGREE OF DISSOCIATION. The lower toxicity of ammonia and trimethylamine is indicated in table 2. Efficiency in preventing development of the sea-urchin’s egg and cytolytic action were taken as criteria of toxicity. In addition to ammonia and the amines, tetraethylammonium hydroxide was also studied. Al- though a substituted ammonia, it is strongly dissociated, ranking with the inorganic alkalies, and in table 2 it may thus serve as a representative of the inorganic alkalies. The unfertilized eggs were placed in the solution, and a certain number removed to sea-water after intervals of 2 and 5 minutes and fertilized. 5 Ammonia and the amines enter cells so rapidly that it is impossible to detect constant differences in pene- tration rate. TABLE 2.—Effect of ammonia, amines, and tetraethylammonium hydroxide on toxopneustes Penetration of Marine Tissues by Alkali. eggs, stained in neutral red. 135 Condition of eggs re- Condition of eggs re- Solution of alkaliin | Color change | Effect of solution moved to sea-water moved to sea-water Mg-free sea-water of red to on egg after 3 after 2 minutes and after 5 minutes and N/250-concentration. yellow. hours. fertilized. Examined | fertilized. Examined after 3 hours. after 3 hours. BELO Ene svercicisiets as Instantly..... Irregular fragmen- | Segmentation; some ir- | Many irregular seg- tation and glo- regular. mentations. bule formation. Methylamine....... Instantly..... Clear cytolysis....| Irregular segmentation. | Unsegmented eggs; a few fragments. Dimethylamine..... Instantly..... Clear cytolysis....| Irregular segmentation. | Unsegmented eggs; a few fragments. Trimethylamine.....| Instantly..... Irregular fragmen- | Segmentation; some ir- | Many irregular seg- tation and glo- regular. mentations. bule formation. BER VlAMINEs,. .:. 6:60:65 | Instantly..... Clear cytolysis....| Irregular segmentation. be prscetiae eggs and ragments. Propylamine........ Instantly..... Clear cytolysis....| Irregular segmentation. | Unsegmented eggs and fragments. Isopropylamine..... Instantly..... Clear cytolysis....| Irregular segmentation. | Unsegmented eggs; some irregular seg- mentations. Tetraethylammo- ‘Not yellow Many red eggs, and | Irregular segmentation; | Irregular segmenta- nium hydroxide. after 3hours.| some fragments. red in color. tions; red in color. | | I. Control—fertilized in Mg-free sea-water—red, segmentations normal. II. Control—unfertilized in Mg-free sea-water—red, unsegmented, normal. It will be noticed that eggs are not cytolyzed after 3 hours immersion in N/250 NH,OH, NH(CHs)30H, and N(C2Hs)sOH (column 3) and after 5 minutes immersion are still capable of fertilization (column 5); whereas the opposite is true of the remaining alkalies of table 2. N(C2Hs)sOH, most highly dissociated of all, is no more toxic than NH,OH, least dissociated of all. The explanation of the difference must lie in the penetrating powers of the two alkalies. N(C:Hs)s,OH can not penetrate the egg-surface and is consequently less toxic than NH,OH, which readily penetrates. Pene- tration is of first importance in determining toxicity. If the alkali pene- trates readily, then degree of dissociation is a further factor in determining toxicity. Those weak (ammonia and trimethylamine) alkalies which are least dissociated are least toxic. The same fact appears when other cells are tested; for if segmenting palolo eggs (Eunice fucata) are placed in N/250 solutions of NH,OH and the amines for two minutes, and then returned to sea-water, only those eggs develop into swimming larve which have been immersed in trimethylamine and ammonia. Eggs exposed to the remaining amines are for the most part killed; only a few larve develop. GENERAL CONSIDERATIONS. On being returned to sea-water the palolo eggs which have been turned yellow by immersion in the weak alkalies again become red in color. The return of color is much more rapid in ammonia and trimethylamine than in the remaining amines. The explanation of this phenomenon lies in the lesser degree of dissociation of the former substances. Neutral red occurs in the egg in small granules, a combination between the neutral red base (BOH) and some organic substance having acid properties (HA), forming a salt (BA). BOH + HA = BA + H:20 136 Papers from the Marine Biological Laboratory at Tortugas. This red salt BA is broken up by NH,OH (or any other alkali), neutral red (BOH) is freed, and a compound NH,A is formed. The weaker the alkali the greater will be the hydrolytic splitting which its salt, NH,A, will undergo. NH.A + H.O0 = NH,OH + HA The NH;OH will diffuse away more rapidly than a stronger base, such as methylamine, and recombination of HA and BOH will consequently occur more rapidly with the reappearance of the red color. The results with marine organisms agree in general with those obtained on fresh-water forms. One point of difference is worthy of mention. Ammonia and trimethylamine were found to be less toxic than the remaining amines for Elodea cells, just as in the case of marine forms; but while ammonia was less toxic, trimethylamine was ranked with the remaining amines, not with ammonia, for Paramecium. Iam unable to explain this difference. It is possible that my result with trimethylamine and Para- mecium may be due to some unforeseen experimental error. In other respects the results agree entirely. The effect of alkalies on living tissues in general may be represented as follows: Less strongly dissociated weak al- kalies. Ammonia and trimethyl- amine. More strongly dissociated of the weak alkalies. Amines (except tri- methylamine). Strongly dissociated inorganic al- kalies, including tetraethylammo- nium hydrate. Penetration very rapid. c Penetration very rapid. Most toxic. i Penetration very slow. Less toxic. Least toxic. Penetration of Marine Tissues by Alkali. 137 PART If PERMEABILITY AND FUNCTIONAL CHANGE. THEORETICAL. Two quite opposite points of view have been taken by physiologists in regard to the relative importance of the permeability of cells. Some regard the cell-surface as of no consequence in regulating cell processes, being freely permeable at all times to crystalloidal substances in the medium or in the cell. Others see in the surface an important regulator of chemical reactions in the cell and a barrier of a high degree of impermeability, retaining cell constituents as a test-tube retains the substances reacting within its walls. Theoretically, no more simple method of regulating the velocity of chemical reactions could be found than by means of a membrane whose permeability to the reaction product might vary. Actually, there is a large mass of evidence indicating that the cell-surface is a highly impermeable membrane for substances normally occurring in or used by cells. The evidence has been advanced in numerous papers by Lillie’ and Hoeber’? and need not be discussed here. Crucial evidence has only recently been put forward that the cell-surface (particularly its permeability) plays an important part in cell activities— that it may vary at different periods of activity and in the presence of definite substances. The evidence may be classified under three heads: 1. Measurements of permeability changes induced by narcotics, anes- thetics, and salts or other substances which affect cell activities. 2. Measurements of permeability during different states of functional activity. 3. Determinations of the rate of penetration of a substance and rate of change in functional activity induced by the substance. PERMEABILITY CHANGE THROUGH CHEMICALS. Of particular interest here is the experiment of Osterhaut,’ in which it is shown most clearly by the plasmolytic method that the plasma membrane possesses greatest impermeability to salts in a balanced medium or in the medium which most nearly approaches that of the normal. Pure NaCl or pure CaCl, both enter Spirogyra cells or marine algz slowly, the latter more slowly than the former; but if the two are mixed in the proportions of a balanced solution, they mutually retard each other’s entrance. The NaCl, alone in solution, must increase the permeability of the plasma membrane to such an extent that its entrance is greatly facilitated. Osterhaut® has recently confirmed his results, mentioned above, by the electrical conductivity method of measuring permeability. Disks cut from the kelp, Laminaria, were placed face to face so as to form a cylinder several centimeters long, and its conductivity was measured. In NaCl the con- ductivity of the cylinder was found to increase regularly (as compared with 6 Lillie, R.S. Amer. Jour. Physiol., vol. 24, p. 14. 1909; vol. 28, p. 197. 1911. Also Biol. Bull., 27, 192. 7 Hoeber, R. a ae Chemie der Zelle und Gewebe., 3d Aufl. torr. 8 Osterhaut, W. J. V. Science, n. s., vol. 34, p. 187. IQII. ® Osterhaut, W. J. V. Science, n.s., vol. 35, p. 112. I912. 138 Papers from the Marine Biological Laboratory at Tortugas. its value in sea-water), while in CaCl, or lanthanum chloride the con- ductivity decreased. Under the conditions of the experiment, increase or decrease in conductivity denotes a corresponding increase or decrease in permeability. The changes in permeability induced by NaCl, CaCls, or LaCl; are readily reversible within certain limits on returning the tissue to sea-water. My own experiments, cited above, indicate that NaOH may enter many kinds of eggs much more readily when dissolved in 0.6 m. NaCl than in 0.625 m. Mg-free sea-water. The pure NaCl evidently decreases the resistance” of the egg-surface for NaOH. The surface of Elodea and Spirogyra cells is likewise affected by minute traces of substances in solution. For instance, N/40 NaOH dissolved in tap-water entered Spirogyra in about 40 minutes; dissolved in double dis- tilled water, condensed in soft glass,!4in about 10 minutes; and dissolved in N/1o NaCl, instantly. Again the resistance of the surface is greatest in the medium most nearly normal. Addition of CaCl, to the NaCl prevents so rapid penetration of NaOH.” Anesthetics give similar results with Elodea cells. Chloroform added to N/4o NaOH to 1/6 saturation increases the penetration rate from 90 to 13 minutes. The protoplasmic rotation characteristic of the Elodea cells ceases in 1/6 saturated chloroform, but the effects of the chloroform are completely reversible, for, on removing the leaves again to pure water, rotation promptly begins. Alcohol and ether act in a similar manner. That chloroform increases permeability in 1/6 saturated solution may be verified by the plasmolytic method. Such “reversibly’’ chloroformed cells are plasmolyzed by urea much less readily than normal cells, a result indicating that the urea enters more rapidly.¥ On the other hand, Lillie! cites evidence to show that very small con- centrations of anesthetics decrease permeability. PERMEABILITY AND FUNCTIONAL ACTIVITY. Lillie has advanced evidence showing that stimulation of muscle-tissue is accompanied by an increase in permeability of the tissue. Those sodium salts which are most effective in stimulation are most effective in causing the loss of pigment from the tissues of Avenicola larve. Loss of pigment indi- cates increased permeability to pigment. The same is true as regards artificial parthenogenesis. These salts most effective in stimulating devel- opment cause most rapid loss of pigment from the eggs (Arbacia) ;* they are also most toxic; in them death of the egg ensues most rapidly. It has long been known that the condition of death is one characterized by marked increase in permeability. The change on stimulation is in the direction of that accompanying death. 19 The word resistance is used instead of permeability toindicate that the cellis normally impermeable to NaOH which only enters after affecting the cell-surface. s 11 Such water is perfectly harmless for both Spirogyra and Paramecium, a good test of the purity of a water. 12 Harvey, E. N. Loc. cit., p. 540. 13 Harvey, E. N. Loc. cit., p. 539. 4 Lillie, R.S. Amer. Jour. Physiol., vol. 29, p. 372, 1912; vol. 30, p. I, I9I2. 15 Lillie, R.S. Amer. Jour. Physiol., vol. 24, p. 14. 1909. 1 Lillie, R.S. Amer. Jour. Physiol., vol. 26, p. 106. I910. Penetration of Marine Tissues by Alkali. 139 Further evidence that the egg increases in permeability immediately after fertilization is furnished by McClendon.” He found that the elec- trical conductivity of the sea-urchin egg is increased after fertilization, thus showing that its permeability is increased. A similar result may be obtained by the plasmolytic method.” Fertilized eggs are plasmolyzed more readily than unfertilized by a molecular solution of cane sugar. This is due to the fact that the salts diffuse out of fertilized eggs more rapidly, the osmotic pressure within falls more rapidly, and plas- molysis occurs more readily. Yet another fact indicates that the surface of fertilized eggs has under- gone amarked change. They are entered more readily by NaOH. In this particular case we may say that the resistance to the entrance of alkali has decreased, using the word resistance, as before, to indicate that the NaOH enters only after modifying the normal impermeability of the surface. Trondle,!’ and also Lepeschkin,” using the plasmolytic method, have found that the permeability of plant cells for salts and sugar increases in the light. The increased permeability has an adaptive significance in that the sugar formed by photosynthesis may be more rapidly removed and photosynthesis itself thereby favored. PENETRATION RATE AND CHANGES IN FUNCTIONAL ACTIVITY. Under this head are classed the cases in which the rate of penetration of the substance is reduced to zero, 2. e., cases in which marked functional or structural changes appear, while we have no evidence that the substance enters the cell at all. The obvious conclusion is that it produces the effects observed by acting on or combining with the cell-surface itself, and affords excellent proof of the important rdle played by the surface in all activities. It has long been known that muscles lose their irritability in isotonic solutions of potassium salts. Overton®! showed that the muscle might remain in the potassium salt solution for days without change of volume, so that the potassium salt could not have entered and its effect in changing irritability must have been a surface action. Warburg” showed that the oxygen consumption of the egg was greatly increased in the presence of dilute NaOH, although the NaOH did not enter; yet NH,OH could enter readily and in concentrations containing too few OH ions to affect to any marked extent the oxygen consumption. Neutral red is therefore a more delicate test for OH ions than increased rate of oxidation. Here is unequivocal evidence that a chemical process in a cell can be affected by a substance which does not enter the cell at all. In my earlier paper I have shown that initiation of cell division in sea- urchin eggs, stoppage of ciliary movement and cytolysis of Paramecium, and stoppage of protoplasmic rotation in Elodea might be brought about before any appreciable amount of NaOH had entered the cells. ~ McClendon, J. F. Amer. Jour. Physiol., vol. 27, p. 240. I910 18 Harvey, E.N. Science, n.s., vol. 32, p. 565. I910. 1 Tréndle. Ber. d. deutsch. bot. Ges., vol. 27, p. 71, 1909; and Jahrb. f. wiss. Bot., vol. 48, p.171, 1910. ®Lepeschkin. Ber. d. deutsch. bot. Ges., vol. 26a, pp. 198,231, and 724, 1908; and Beihefte z. bot. Zentralb., vol. 24I, p. 308, 1910. 21 Overton, E. Pfluger’s Arch., vol. 92, p. 115. 1902. 22 Warburg. Loc. cit. 140 Papers from the Marine Biological Laboratory at Tortugas. It is the purpose of the following paragraphs to indicate how far other functions may be affected by NaOH before it penetrates the cell. It may be stated in advance that every cell activity thus far studied is affected, indeed abolished, before the NaOH enters, while a great many cell activities remain entirely unaffected even after NH,OH has penetrated and turned the neutral red to yellow. Evidence showing that the indicator method is an adequate one for detecting the penetration of both strong and weak alkalies has been presented in a previous paper (Amer. Jour. Physiology, vol. 31, 1913). EXPERIMENTAL. MUSCLE AND NERVE TISSUE. CASSIOPEA XAMACHANA. The only experiments on the permeability of the cells of medusz of which I am aware are those of Bethe, cited above. The method of experimentation with Cassiopea, a scyphomedusa well adapted for physiological investigation, was the same as that previously used in studying the effect of temperature on the muscular contraction and nerve conduction. The method was originally described by Mayer. A long strip of tissue cut from the disk of the jelly-fish is laid across three dishes (A, B, and ¢, fig. 1). Dishes A and € are filled with sea-water. Dish ay Loa TRC h or wes BiGay- B is filled with the alkali solution to be tested. Stimuli are constantly arising in the sense-organs in dish A and pass along the nerve-network of the strip of tissue toward C, stimulating the muscles as they go. When the nerves are so affected by the alkali in B that they no longer conduct a stimulus, the muscle in c will no longer contract. In the case of some al- kalies, acids, and most other substances, the muscles in B are affected before the nerves. The time required to stop contraction can be determined by mere inspection, since the muscles are being constantly stimulated by nervous impulses carried from the sense-organs in A. In the case of certain of the alkalies, the nerves in B are affected before the muscles. We may then determine by direct stimulation of the muscles in B with the electric current when they have been so far affected as to cease contracting. Entrance of alkali is of course detected by the color change of neutral red. % Harvey, E. N. Carnegie Institution of Washington, Pub. No. 132,p. 27. I9II. % Mayer, A. G. Carnegie Institution of Washington, Pub. No. 102, p. 128. 1908. — Penetration of Marine Tissues by Alkali. « 141 The subumbrella surface of Cassiopea is covered by a thin layer of epi- thelium and muscle and nerve tissues. The epithelium is outermost, nerve- network next, then come the muscle-fibers, and finally, embedded in the jelly, are clumps of symbiotic algal cells scattered in lines running parallel with the muscle-fibers. Microscopic examination of the Cassiopea strips used in the following experiments reveals a layer of red-stained granules near the surface and on focusing downward lines of red-stained granules are observed following the muscle-fibers and appearing to be within them. Transverse sections of the same strips showed, however, that very little more than the outer third of the epithelio-muscle layer contained cells stained in neutral red. The dye is chiefly taken up by globules and granules in epithelial mucus-cells. These granules are practically all cast off in the mucus formed by exposing Casstopea to chloroform-saturated sea-water or other injurious solutions. Neither nematocysts nor algal cells stain. No definitely stained nerve-cells can be made out, although red-stained granules are present where nerve- cells should be. Only the region of the muscle-fibers away (i. e., toward epithelium) from the contractile fibrils stain. It should be borne in mind, then, in interpreting the following results, that red-stained granules are not present in the immediate region occupied by the contractile fibrils, but are present in a region first reached by alkali diffusing into the tissues and in a region containing the nerve-network, although it is impossible to say whether granules within the nerve-cells are stained or not. It is probable, however, that both nerve-cells and the outer ends of the muscle-fibers con- tain the red indicator. The results are given in the following table: TABLE 3.—Relation between color-change, cessation of contraction, and cessation of conduction in Cassiopea. (Tortugas, July 9-10, 1911.) Alkaline solution | Time of : 4 Recovery of | Recovery of N/250 concentra- | €XPOSUre | Color change in Cessation of Cessation of contraction in | conduction tion in Mg-free to solu- minutes. contraction in conduction in |minutes on re-| in minutes sea-water. tion in minutes. minutes, moval to sea- | on removal minutes. water. to sea-wate Ammonia...... 20 I.5 Feeble, 5; Missing,* 10; 5 2 ceases, I0..... ceases, 16. Methylamine... 5 I torso oe Bovanaccaes Slight contrac- 6 tion in 10. Dimethylamine . 5 I Slight? contrac=|)25 2 osc os ccc AG elere cise e/ele ais I tion, 4; ceases, 4.5. ee 2 Trimethylamine. 25 2.5 Feeble, 10; Missing, 15; AAs Oconee ceases, 18...] ceases, 20... Ethylamine.... 5 I Feeble, 3; Missing, 2; ee SOer CO 6 ceases; 4.4...|| Ceased, 3. c « Normal propyl- 5 I Feeble, 4; 7 Se pete ee Sars ouate eh biare. eyeks 5 amine. CREASES TA Se are Widnes cece Isopropylamine . 5 I Feeble, 4; Missing, 3; Eine dieateteiatew eis I ceases, 5... ceases, 3.5... : Tetraethylammo- 5 Nocolor change} Feeble, 5; Missing, 2; Dio feyavoresevexcnopets Missing, 2; nium hydroxide.) ceases, 8 ceases, 4.... regular, 5 NAOH.» d52s.0. 10 Nocolorchange} Muscles con-| Occasional, 3; | I .........+: I tract when! ceases, 4..... impulse pass- es. Very fee- ble, ro. USLOYS eget eli ae No color change} Feeble, 8...... Ceases,6.... sr(OH)e).c:'. ose Nocolor change} Feeble, 7...... Ceases, 4.5... *“ Missing”’ indicates that sometimes the impulse can pass from A to c, sometimes not. 142 Papers from the Marine Biological Laboratory at Tortugas. The actual times vary somewhat in different experiments, but the above is typical. The results of interest in the above table may be sum- marized as follows: The weak alkalies (first seven) enter almost instantly, yet contractions and conduction do not cease till some time after the color change occurs. The strong alkalies (last four) do not enter until long after contraction and conduction have ceased. In ammonia and trimethylamine the muscles regularly lose their power of contraction before the nerves. In the remaining alkalies the nerves cease to function before the muscles as direct stimulation shows. The phenomenon is more marked with the inorganic hydroxides. Thus far these alkalies are the only substances known which affect the nerves first and the muscles afterwards. Recovery of both contraction and conduction will take place if the tissue is removed to sea-water soon after stoppage of contraction or con- duction occurs. Occasional exceptions to this rule have been recorded in which the nerve-tissue has been so injured in one spot that the impulse was unable to pass. The red color returns only in the tissue exposed to the weak alkalies. Recovery from the effects of the strong alkalies never occurs if the tissue is allowed to remain in the solution long enough for the color change to take place, nor does the red color return on transfer to sea-water. Ammonia and trimethylamine are less toxic than the remaining amines. The strong alkalies take an intermediate position between the two groups of weak alkalies as regards toxicity for Cassiopea muscle. That the resistance to the entrance of neutral red is a property of the living cell may be shown by treating the Cassiopea strips with chloroform- saturated sea-water. The cells are killed and the red-stained granules are cast out in a slime from which the neutral red dye slowly diffuses. If such a strip is placed in N/250 NaOH, the slime is turned yellow in less than two minutes, 7. €., just as rapidly as in N/250 NH,OH. Toa certain extent the slime prevents free access of NaOH to the tissue. When stripped off with a camel’s-hair brush, the underlying tissue, still pinkish in color, is rapidly turned yellow by the alkali. The slime forms most rapidly in the strong alkalies, much less rapidly in the stronger amines, and very slowly in NH,OH and trimethylamine. In the following experiments on tissues other than those of Cassiopea and on eggs, the effect of NaOH has always been compared with that of NH:.OH, for two reasons: (1) My general experience indicates that NH,OH may be taken as a representative of the weak alkalies and NaOH as a repre- sentative of the strong. (2) Since NaOH always affects the cell before it enters, we might consider the visible alteration in function a more delicate test for the presence of the OH ion within the cell than the neutral red indicator. The control experiment with NH,OH shows that this is not the case. The color change always begins and in many cases is complete before Penetration of Marine Tissues by Alkali. 143 any alteration in function occurs. Even though NH,OH enters readily it is too weak an alkali to affect the cell immediately. PENNARIA TIARELLA. The tentacles were stained in neutral red and the time for color change was compared with the time required for spontaneous movement to cease. Globules much like those occurring in Cassiopea take up the dye. They are present in all parts of the tentacles. It is probable (but can not be stated positively) that stained granules are present within muscle-fibers. In NaOH of any concentration movement always ceases long before any red granules are turned yellow. In N/10o0o0 NH,OH movement ceases in about 15 minutes. The red begins to change, but is not completely yellow at this concentration for a long time. In N/500 NH,OH movement ceases before the color change is complete. The above results were also obtained using the gill muscles and tentacles of an unidentified Amphitrite-like annelid from Tortugas, Florida. SALPA DEMOCRATICA. Red-stained granules undoubtedly occur within the muscle-fibers of the circular bands of this animal. In NaOH of any concentration movement ceases before the color change. In N/1to0oo NH,OH contractions cease in about 15 minutes. The color change begins in 5 minutes, but is not complete even after an hour. In N/500 NH,OH contractions cease before the color change is complete. CILIA AND MODIFIED CILIA (CTENOPHORE SWIMMING-PLATE). BERGE OVATA. Granules stained in neutral red occur in the swimming-plate cells, not in the plates themselves. In NaOH the movement always ceases long before the color change occurs. In N/500 NH,OH, the plates cease beating at first, begin again in less than a minute, but have mostly stopped in 5, and all by Io minutes. The red granules are only partially yellow even after 45 minutes, although the color change begins in the course of 3 to 4 minutes. TOXOPNEUSTES LARVZ. Red granules occur in the ciliated cells, but not in the cilia themselves. In NaOH the cilia always stop before the color change occurs. In N/500 NH,OH color change occurs in I to 2 minutes. Movement is stopped at first, begins again in about 4 minutes, continues slowly after half an hour, and has not entirely ceased even after an hour. A similar result is obtained with palolo trochophores (Eunice fucata). These experiments with marine cilia agree in every way with those ob- tained on Paramecium. 144 Papers from the Marine Biological Laboratory at Tortugas. The above results taken as a whole admit of only one interpretation. If NaOH prevents the normal functioning of many different types of cells without appreciably entering those cells, we must conclude that its effect is on the surface and that the alteration in function is connected with a change in the surface layer. Muscle contraction, nerve conduction, and the move- ment of cilia and modified cilia (swimming-plates of the ctenophores) must all be dependent, primarily, on changes in the surface layer of the muscle, nerve, or ciliated cells. CYTOLYSIS OF MARINE EGGS. Cytolysis of marine eggs is accompanied by (1) an increase in volume of the egg as a whole; (2) in many cases a solution of the yolk and pigment granules. Two quite distinct views have been held as regards the cause of the swelling: (1) Swelling may be due to an increase in permeability of the egg surface-membrane such that the salts of sea-water can readily pass in; they therefore no longer balance the osmotic pressure of substances within the egg and the egg swells; (2) swelling may be due to a breakdown of protein, lipoid, or other substances and the development of a higher osmotic pressure or a swelling pressure within the egg. According to the first view, the action of a cytolytic substance must be chiefly on the surface; according to the second, action is on the cell contents. The relation of cytolysis to penetration of alkali is therefore of special interest as furnishing evidence in favor of one or the other theories of cytolysis. Details of typical experi- ments are given below. TOXOPNEUSTES VARIEGATUS (MATURE EGG). N/80 NaOH—Egg remains in solution absolutely unchanged for 6 min- utes; diameter 5.5 units;2® then swells suddenly to 7.0 units. Color change is coincident with swelling. N/80 NH,OH—Color change instantaneous. Egg remains in solution for 15 minutes unchanged; diameter 5.5 units. In the next 5 minutes the egg begins to appear coarsely granular and at the end of 2 more minutes has swollen to 7.5 units and is filled with a mass of globules. Swelling is relatively rapid. EUNICE FUCATA (PALOLO WORM) MATURE EGG. N/80 NaOH—Egg remains in solution unchanged for 8 minutes; diame- ter 19 units. Then swelling begins and continues slowly for 5 minutes; diameter 21.5 units. The color change occurs gradually during the swelling. N/80 NH,OH—Color change instantaneous. Egg remains unchanged in appearance and size and only becomes broken up into oil-like globules after an hour. 2 A unit is 19.2 microns. Penetration of Marine Tissues by Alkali. 145 HOLOTHURIA FLORIDANA (IMMATURE EGGS). N/80 NaOH—Eggs remain unchanged in size and appearance for 6 minutes; diameter 14.2 units. The surface layer then becomes clear, the clear region extends inward, and in the course of 1.5 minutes more the whole egg is clear and swollen to 16 units. The color change occurs as the egg contents become clear. In N/160 and N/320 NaOH the color change occurs some time before the egg clears and swells. Thus in N/160 NaOH the color change occurs in 18 minutes and cytolysis in 40 minutes. N/80 NHsOH—Color change is instantaneous. Eggs remain the same in appearance and size for an hour; diameter 14.4 units. Then swell to 18 units in the course of 2 to 5 minutes. The immature eggs of Fissurella sp., Eurythoe sp., and an Amphitrite- like worm behave as do those of the palolo. It was often noticed that if segmenting eggs were placed in NaOH differ- ent blastomeres were entered at very different times by the alkali. Thus the resistance of each individual blastomere is specific and on division the daughter cells may become perfectly distinct from each other so far as permeability relations go. Whether such differences are related to the hereditary potencies of the individual blastomeres was not determined. It will be seen from the foregoing observations that several types of eggs may be distinguished as regards penetration of NaOH and cytolysis, viz: (1) Eggs in which cytolysis and color change are both simultaneous and rapid (Toxopneustes and sea-urchin eggs in general); (2) eggs in which cytolysis and color change are simultaneous but gradual (Eunice and other annelids); (3) eggs in which color change may occur before cytolysis (Holothuria floridana, immature eggs). In addition to the above: (4) eggs in which the color change occurs after cytolysis (Cumingia, observed at Woods Hole, Massachusetts). In all types ammonia and the amines enter instantly, long before cytolysis occurs. Only one egg thus far examined (that of Cumingia tellinoides) belongs to type 4, and one egg only (that of Holothuria floridana) to type 3. In other cases penetration of NaOH and swelling are simultaneous. Leaving aside for the moment the egg of Holothuria, which forms an exception to the rule that cells are visibly affected before or at the moment of entrance of NaOH, let us analyze the results obtained with eggs of types I and 2. The rapid color change in ammonia shows how small the concentration of OH ions need be to affect the red-stained granules. We may assume, then, that if even a trace of NaOH entered the color change would be rapid. Since practically all marine eggs remain quite unchanged in size and in color (of the red-stained granules) in NaOH solution for some time and only after a definite interval begin to swell and to change color relatively rapidly, we may conclude that the NaOH can only force an entrance after funda- Io 146 Papers from the Marine Biological Laboratory at Tortugas. mentally changing the normal nature of the cell-surface. The manner in which cytolysis is connected with such a change in the cell-surface may be conceived as follows: The NaOH combines with substances, probably with proteins at the surface of the egg, forming a compound through which it may pass; 7. e., the surface is made permeable to NaOH. This compound must also allow the salts of sea-water to pass readily; the egg is likewise made permeable to the salts of sea-water.28 Osmotic equilibrium is upset and the egg absorbs water and salts. Breaking up of the granules is due largely to the increase of water in the egg. All marine eggs swell when placed in fresh water and when the increase in volume reaches a certain value the contained granules break down; true cytolysis results. The breaking-up of the granules by distilled water is obviously not due to any chemical substance, such as alkali, entering the egg. An increase in permeability might result in swelling without the entrance of the cytolytic substance. Such is the case when Cumingia eggs are placed in NaOH. However, I am inclined to think that the presence of a sufficient number of OH ions within the egg may aid in the breaking down of the granules and that this breaking down increases also the degree of swelling of the egg. Cytolysis in Holothuria appears to be largely of this type, since here NaOH enters before the increase in volume begins. From this point of view both theories of cytolysis contain an element of truth. Swelling of marine eggs is due both to an increase in permeability of the surface and also to the breakdown of lipoid or protein granules within. The latter tends to increase the swelling pressure or the osmotic pressure of the egg, but is secondary to the increase in permeability of the surface. 26 McClendon (Am. Jour. Physiol., vol. 27, p. 240, 1910) finds the electrical conductivity in sea-urchin’s eggs to be greatly increased after cytolysis, a result indicating ready passage of the salts of sea-water. IX. PHYSIOLOGICAL STUDIES ON CERTAIN PROTOZOAN PARASITES OF DIADEMA SETOSUM. By MERKEL. H. JACOBS; Instructor in Zoology, University of Pennsylvania. 147 a rn ee . PHYSIOLOGICAL STUDIES ON CERTAIN PROTOZOAN PARASITES OF DIADEMA SETOSUM. By MERKEL H. JAcoss. INTRODUCTION. It has been shown by the author and others that different species of protozoa have certain physiological characteristics, often almost as striking as their morphological ones, and which are probably of considerable signifi- cance in the interpretation of their habits of life and their relation to their environment. The study of such characters in the past has been greatly neglected on account of the general belief of biologists that they are exceed- ingly inconstant and subject to modification by external factors and there- fore of little value in an understanding of the fundamental nature of the organisms possessing them. Previous attempts to detect characteristic physiological differences between different forms have for the most part been confined to organisms whose general habits of life are so dissimilar as to leave room for the objection that the differences observed may have been due merely to the effect on the animals of the different environments to which they have been accustomed and not to any innate peculiarities of the organisms themselves. It occurred to the author as desirable to test, if possible, a series of forms which naturally live under essentially the same environmental conditions, and which may be assumed to have done so for many past generations, in order to see whether they show greater likenesses than a number of forms selected at random, or whether each has preserved its individuality in spite of the similarity of its environment. In the case of a number of parasitic protozoa living together in the same organ of the same host, we have a favorable opportunity for such a study. Even here we can not assume that each form has exactly the same environment, but, there can be no question that this environment is far more similar than that of a number of selected free-living forms. Furthermore, in order to become parasites of the same host the nature of the forms in question must have been more or less changed. In the case of intestinal parasites, for example, it would be necessary for them to acquire a resistance to the digestive juices of the host and to become accustomed to a more or less anaerobic habit of life, etc. Such changes would necessarily be in the same direction for all of them. If, therefore, a number of parasitic forms, which have had to meet and adapt 149 150 Papers from the Marine Biological Laboratory at Tortugas. themselves to the same new conditions and live under them together for thousands of generations, still show as characteristic physiological differ- ences as free-living forms, it is good evidence of the fundamental nature of these differences. Is such the case? The present paper is a preliminary consideration of this question. The work on which it is based was per- formed at the Tortugas Laboratory of the Carnegie Institution of Wash- ington during the summer of IgIo. MATERIAL. The material used in these experiments consisted of four species of ciliate protozoa obtained from the alimentary tract of the large black sea-urchin, Diadema setosum, which can be procured in practically unlimited quantities from the coral reefs about the Tortugas. This form is particularly favorable as a source of material, not only on account of its abundance but on account of the high percentage of infection shown by it with respect to the protozoa in question. Out of over 100 adult Diademas examined during the progress of the work, not a single one failed to show one or more kinds of protozoa in the intestine, which often fairly swarmed with them. Doubtless the large size of the intestine, its pouch-like recesses in which stagnation of the food can occur, and the character of the food itself are factors favoring the pres- ence of protozoan parasites or, perhaps more correctly, protozoan commen- sals, since none of the forms in question seem to be injurious to the host or to live at its expense, except in so far as they use up a small portion of its practically unlimited supply of food. The forms studied were of four kinds. Since apparently none of them have been previously described or named, they are for convenience desig- nated in the account that follows by the letters, A, B, C, and D. Of the four, B is the most regular in its occurrence, having been found in 100 per cent of the adult Diademas examined. It is a large, heterotrichous form allied to Metopus, with a hook-like anterior end and showing considerable individual variation in appearance. The form next in abundance is C, which is found in about 75 per cent of the cases. Itis a medium-sized, holotrichous form, elongated in shape, with the anterior end of the body transparent and longitudinally striated and the remainder filled with refractive spherules. A nucleus is visible about the middle of the body and no mouth opening can be detected. The next form in importance is the one designated as D, which is found in 40 per cent of adult Diademas. It is flat and leaf-like in form, with the body slightly concave on the ventral and convex on the dorsal surface. The one margin of the body is almost straight and the other curved, giving the animal an asymmetrical shape. At the posterior end is a spur-like pro- jection. Evidently this form is allied to the hclotrichous genus Dysteria. The form which occurs in the smallest number of cases (33 per cent) is A. When present at all, however, the number of individuals is generally very large. In size it is rather small and in shape elliptical and flattened. oe Physiological Studies on Certain Protozoan Parasites. I51 Very little could be made out of its structure, owing to its small size, rapid movements, and its peculiar tendency to disintegrate with almost explosive suddenness when kept under observation on a microscopic slide. In addition to these four characteristic forms, several other ciliates were observed more rarely, and on one occasion an ameba. They were not found frequently enough, however, to be considered in the experimental part of the work. So far as the present observations go, the important group of flagellates appears not to be represented by parasitic forms in Diadema. The parasites of Diadema seem to occur exclusively in the alimentary tract, none being found in the body-fluid so long as the intestine is uninjured. In the alimentary tract itself the number is not the same in the different regions, some parts being more favorable than others. The greatest numbers are found in the distended pouches of the second, or upper, coil of the intestine. In the lower coil they seem to be almost or entirely absent, even when present in great abundance elsewhere. It is interesting to note in this connection that free-living forms (e. g., Foraminifera, etc.), which are taken in with the food, are invariably dead before the second coil is reached. The rectal portion of the intestine also shows very few protozoa, form D being the one most frequently encountered there, though B and C may occur rarely and in small numbers. On several occasions cysts, pos- sibly of D, were found in this region. It seems quite probable that cysts thus formed may pass out and eventually infect other Diademas. It does not always happen that the different protozoa present show the same distribution, thus, A, for example, may be quite absent in some of the pouches and present in others, while B, C, and D show similar irregularities. It is usual to find several of the forms together, but there are no constant combinations; any form may occur associated with any of the others. The above remarks concerning the abundance and distribution of the parasites apply only to adult Diademas. As might be expected from our knowledge of other animals, young individuals contain either no parasites at all or at most but a few. The youngest ones examined (0.5 inch in diameter) contained nothing. Somewhat older ones (1.5 inches in diameter) contained sometimes nothing at all, sometimes large numbers of a form not found in any of the adults examined, and in a few cases small numbers of Cand D. It is rather striking that form B, which appears to be present in all adults, should be so constantly lacking in younger individuals. Whether this is because conditions for its existence are unsuitable or simply because no opportunity for infection has occurred could not be determined. A search for the parasites in question was also made in sea-water and in other animals, related and unrelated to Diadema. The only situations in which any of them were found was in the alimentary tract of other sea- urchins, where they may occur in varying numbers. B, for example, was found rarely in Toxopneustes and Echinometra, and C on a number of occasions in Toxopneustes, while D seems to be present in every adult Toxopneustes, and rarely in Echinometra. A was never found outside of 152 Papers from the Marine Biological Laboratory at Tortugas. Diadema. A number of specimens of Hipponoé were examined for pro- tozoan parasites without results. METHODS. In comparing the physiological nature of the forms in question certain characters were chosen which could be expressed quantitatively and the animals examined with respect to them. Those selected all had to do with the length of time the organisms could maintain their normal activities under various conditions differing from those to which they were accustomed. The method adopted was to place the animals under the new conditions and note the time that elapsed until all visible movements ceased. In the case of ciliates this time corresponds very closely with that at which irreparable injury, resulting in death, has been done to the cell, and therefore furnishes a more or less accurate criterion of the general vitality of the organisms. In the present experiments the number of individuals studied was always large and the range of variation within the species small, and there- fore it was found possible to choose, as the point for comparison, the time at which all the individuals of the given species had been rendered motion- less. This point was easy to determine and the general agreement in the results obtained on successive days by the use of the above method showed it to be sufficiently accurate in a case, such as the present one, where it was desired to express the relative and not the absolute resistance of the four forms in question. Most of the experiments were repeated a number of times with material obtained from different sources, so as to eliminate acci- dental errors. The characters, in respect to which the different forms were compared, were as follows: (1) Ability to live outside the body of the host; (2) ability to live in the body-fluid of other related animals; (3) length of life of the parasites after the death of the host; (4) resistance to COs; (5) resistance to H.S; (6) resistance to the decomposing proteid substances; (7) resistance to H;SQOu4; (8) resistance to KOH. EXPERIMENTS. ABILITY OF THE PARASITES TO LIVE OUTSIDE THE BODY OF THEIR HOST. It is well known that protozoan parasites show a varying ability to live apart from their host. In general, the less specialized the parasite and the looser the relation between it and its host, the longer it may be expected to live under conditions other than the natural ones. The forms in question are not highly specialized and, as might be supposed, are for the most part able to live for a considerable time outside the intestine of Diadema, though so far as these experiments go, a continued existence of this sort is in no case possible. The point of particular importance, however, from the standpoint of the present investigation, is that the four forms studied show marked differences in behavior when removed from the body of their host. If a number of food pellets from the intestine of a Diadema be shaken up with a little fresh sea-water, none of the parasites obtained in this way is at first Physiological Studies on Certain Protozoan Parasites. 153 visibly injured. If, however, a few cubic centimeters of the mixture be placed in a covered glass dish which prevents evaporation but which, being only partially filled, does not prevent contact of the liquid with the air, it will be seen that all of the animals present will die after a length of time which is fairly constant for each species. The first one to disappear appar- ently always is A, which is no longer present after 2 to 3 hours. It may be mentioned that no way of keeping A longer than this in arti- ficial cultures was found except to place a relatively large quantity of the intestinal contents in a test-tube with the addition of only sufficient water to keep it moist, in which case the animals retained their activity somewhat longer. D is next to A in respect to its behavior in sea-water culture, and usually is dead at the end of 24 hours, though a few individuals may survive several hours longer. C does not differ very markedly from D, but is some- what more resistant, living on an average 30 hours or a little over. B is by far the most resistant form, often surviving and continuing active for 2or3 days. The relative ability of the four forms to live outside the body of their host may therefore roughly be expressed by the ratio I : 24% 12 : 10. To determine to what extent the injurious effects of removing the para- site from their host were due to the action on them of the oxygen of the air, a parallel series of experiments was tried in which oxygen, so far as possible, was excluded. This was done by shaking up the food pellets from the intestine with sea-water from which the air had been removed by boiling (the concentration of salts being kept constant by the addition of distilled water) and immediately placing the liquid in dishes which were completely filled, covered, and sealed with vaseline to exclude the air. Such treatment was found not materially to alter the length of life of any of the forms except B, which in one such experiment lived for 7 days, the longest time for which it was found possible to keep any of the forms alive by any of the methods employed. A, C, and D died in about the usual times, the life of A, if anything, being slightly shortened and that of C slightly lengthened. Inno case except that of C, however, were the differences significant. ABILITY OF THE PARASITES TO LIVE IN THE BODY-FLUID OF TOXOPNEUSTES. Early in the course of these experiments it was observed that the forms in question live in cultures made with the body-fluid of Diadema about the same length of time that they do in sea-water. This is not surprising, since the osmotic properties of the body-fluids of sea-urchins and sea-water are about the same. A point of some interest seemed to be whether the body- fluid of a nearly related genus of sea-urchins would be as favorable as that of the host or as sea-water, or whether there might be present in it in addi- tion to the salts of sea-water certain other substances of more specific nature which would have a toxic effect. To determine this point, therefore, experiments were made, but unfortunately they were not very extensive. One such experiment may be mentioned. A, B, and C obtained from Diadema were placed in a small quantity of the body-fluid of Toxopneustes. 154 Papers from the Marine Biological Laboratory at Tortugas. A was found to be dead at the end of 1.5 hours, B in 3 hours, and C was living after 15 hours, but dead in 24. In a control culture in which the body-fluid of the original host was used in place of that of Toxopneustes, A lived about 2 hours, B was still active at the end of 24, and C died somewhere between 15 and 24. It will be seen, therefore, that A and C are less affected by the change than is B, which normally lives 10 to 20 times as long in the body-fluid of its own host as in that of Toxopneustes. Unfortunately D was not available at the time when this experiment was performed. The converse of the above experiment, namely, the subjecting of parasites ob- tained from Toxopneustes to the body-fluid of Diadema, was also tried with similar results. In this case only D was available, it being the only one of the four forms found in quantity in Toxopneustes. In the control culture in which the body-fluid of their own host was used the animals were normal after 15 hours; in the body-fluid of Diadema they were dead in 2 hours. The injurious effects of transferring the parasites to the body-fluid of an animal other than their host are almost certainly not due to physical (e. g., osmotic) differences between the body-fluids of the two animals, since they may be transferred with impunity from either host to sea-water, the body-fluid of sea-urchins having approximately the same osmotic properties as the latter. Furthermore, in a number of experiments in which the con- centration of the liquid surrounding the parasites was purposely suddenly altered, the forms in question were shown to be quite resistant to changes of this sort. It may beassumed, therefore, that chemical rather than physical differences in the two body-fluids are responsible for the death of the animals. It is rather interesting that different races of the same species are found adapted to different hosts and apparently can not be suddenly transferred from one to the other. Where adaptive characters such as these are concerned, we can look for a considerable amount of variation within the species. But in the case of non-adaptive characters, such as some of those mentioned in the succeeding experiments (resistance to H,S, H.SOg, etc.), it was found that the properties of form D, for example, were practically the same whether it was obtained from Diadema or Toxopneustes, and consequently that such characters are fairly constant for the species. LENGTH OF LIFE OF THE PARASITES AFTER THE DEATH OF THE HOST. If a dead Diadema, either opened or unopened, be covered with sea- water and allowed to stand in a warm room, the number of parasites present soon begins to diminish and after a variable time all disappear. Sometimes in as short a time as I2 hours none are to be found, though in other cases they may be active for 24 hours or more. In no instance, however, do they seem to persist more than 30 hours at the room temperature of the Tortugas laboratory (85° to 95° F.). They usually disappear in the order, D, A, C, B. In one series of observations D was dead in 12 to 15 hours, A in a little less than 20, while C and B survived somewhat over 24. While the absolute Physiological Studies on Certain Protozoan Parasites. 155 time of disappearance differs in different experiments the relative order for the four forms seems to be quite constant. It will be seen, therefore, that in respect to their ability to live in the body of the host after the latter’s death, just as in many other respects, they show certain characteristic physiological differences. The relative length of life of the four forms is roughly 1.5:2:2:1. Just what factors cause the death of the parasites under these conditions is not certain, but probably the products of decomposition present play an important part. As will be shown later, some of these substances are markedly toxic, though each of the four forms is affected differently by them and the immediate cause of death may perhaps be different in the different cases. EFFECTS OF COg ON THE PARASITES. Carbon dioxide certainly occurs normally in the intestine of Diadema and therefore small quantities of it can not be markedly injurious to any of the four forms in question. With large quantities the case is different, all being killed in a fairly short time, though the four forms are by no means alike in their resistance. If a drop of fluid containing them be exposed to a contin- uous stream of the gas in a gas-chamber and kept under constant obser- vation, it will be seen that D, the least resistant form, is rendered motionless in 3 or 4 minutes; A comes next, the time required for it being 5 to 6 minutes; C remains active for about I5 minutes, and B for from 45 to 60 minutes. The ratio of resistance is therefore roughly 1.5:15:4:1; B being the most resistant form and D the least. The same order of resistance was found to hold if instead of subjecting the animals to a stream of gas they were placed in a relatively large quantity of sea-water charged with CO, by means of a ‘‘sparklet”’ bottle. EFFECT OF HeS ON THE PARASITES. Hydrogen sulphide begins to appear in the intestine of Diadema in con- siderable quantities soon after death. The vapors given off on gently warming a little of the intestinal contents taken 12 hours after death darken lead acetate paper very perceptibly, and after 24 hours the reaction may often be obtained without the application of heat at all. As this substance might be suspected of playing a part in causing the disappearance of the parasites from the intestine of the host after death, it was of interest to examine its effects on the four forms in question. As might be expected, it proved to be markedly toxic to all, but the different forms resisted it in pronouncedly different manner, and, furthermore, the order of resistance is different from that shown to CO:. C, for example, which is next to the most resistant form to COs, is killed first when a current of the gas is passed through the gas-chamber containing the animals, the time required being only 1 to 2 minutes; D comes next, with a period of about 3 minutes; A lives 3 to 4 minutes, and Bioto15. The ratio of the resistance of the four forms is therefore approximately 2.5 :8:1: 2. 156 Papers from the Marine Biological Laboratory at Tortugas. EFFECT OF DECOMPOSITION PRODUCTS ON THE PARASITES. In an attempt to throw light on the cause of the disappearance of the parasites after the death of the host, the effect on them of the direct products of the decomposition of the tissues of Diadema was tried. The method used was to allow fragments of Diadema tissues to decay in the water for 3 days at room temperature; 5 drops of the resulting infusion were then added to 10 drops of the fluid containing the animals. The results were most surprising. B, which had hitherto proved the most resistant form in most of the experiments, was killed almost instantly—certainly in less than 5 seconds; A lived from Io to 60 seconds; D 20 to 25 minutes; and C as long as 2 hours. C, therefore, which is less resistant than B to most conditions, is in this case many more times resistant. The ratio for the four forms is roughly 7 :1:1,500:250. These figures seemed so remarkable that the experiment was repeated several times, but always with essentially the same results. THE EFFECT OF H2S04 ON THE PARASITES, A limited number of experiments were tried with thissubstance. It was necessary first to determine the proper concentration, since if the solution is too strong death occurs so quickly that its time can not be accurately observed, and if too weak it occurs so late that other factors have time to complicate the results. A series of preliminary experiments showed that a favorable proportion is 1 drop of N/10 HeSOx, to 5 drops of the fluid con- taining the organisms. Of course, care had to be taken that the acid was immediately mixed thoroughly with the culture medium, otherwise the organisms were not all subjected to exactly the same conditions. Treated in this way, the four forms showed striking differences: A had a very low resistance, being killed in about 5 seconds; D was killed in less than 30 seconds, but the exact time (somewhere between this and 5 seconds) was not noted; C died in about 30 seconds; while B survived 1.25 hours or 900 times aslong as A. This is seen to be a surprisingly high resistance to acid when it is remembered that Paramecium and many other fresh-water forms are killed almost instantly by exposure to N/500 H,SO,. Summarizing the results of the observations on the resistance of the four forms to H2SO,4 we have the ratio, I :900:6: <6. EFFECT OF KOH ON THE PARASITES. This substance used in the same concentration as H,SOu., as might be expected from results obtained with other animals, is on the whole much less injurious. A (the least resistant form) was found to be killed in about 2 minutes, D in somewhat less than 45 minutes, B in about 1 hour, and C in 3 to 4 hours. It will be noted that while A, C, and D are much more resistant to KOH than to H.SO,, B is actually less resistant. The ratio for the four forms is roughly I : 30 : 100 : 20. Physiological Studies on Certain Protozoan Parasites. 157 CONCLUSIONS. The general results of the experiments performed is to show surprising differences in the resistance of the parasites of Diadema to various unfavor- able conditions. In some cases the most resistant form may live several hundred times as long as the least resistant one. B is 900 times as resistant to H,SO,as A, and the difference between B and C with respect to products of decomposition is still more striking. Furthermore, a form which is strongly resistant to one condition may be only feebly so to another, and vice versa. For example, B is 8 to 10 times as resistant to H2S as C, but 1,500 times less resistant to certain of the products of decomposition of Diadema tissues. Cis more resistant to CO, than A or D, but less resistant to HS. A, C, and D are 24 to 450 times more resistant to KOH than to H.SOu,, while B is somewhat less resistant. Other similar facts could be mentioned. Comparing all of the results obtained, it is therefore seen that the similar habit of life of the four forms in question has not brought about physio- logical similarity except in certain adaptive characters which are a sine qua non for continued existence in the same host (e. g., ability to resist the digestive juices of the latter, etc.). In other respects they are just as different as almost any four free-living forms that might be selected and so far as the evidence of these experiments goes it seems to show that the physiological characters of an organism are not merely the result of its environment, but may be as fundamental and characteristic as its morpho- logical ones. X. SGIN OF THE ELECTRIC TISSUES OF GYMNARCHUS NILOTICUS. By ULRIC DAHLGREN Professor of Biology, Princeton University. Nine plates and nine text-figures. 159 foe cee ets te oh he shire lena Died en atoms aos A auioainlete 40 SS Neely: Sereda UOT OAM? AYSOIPACT OURS ot cperanl opt Ahh keyalg is weet marienetest aati Ene estalg oni ' tt er rt ee “=o el akametien acacia Pakpes eebancaeests ees gabe ‘ ee ee ee eae nS mr sree SERIE PIERS 2a eT SRS FS nN ie Ne IE ee eee ORIGIN OF THE ELECTRIC TISSUES OF GYMNARCHUS NILOTICUS. By ULric DAHLGREN. In the seven types of electricity-producing fishes the exact development by which the electric organs and tissues are produced during the creature’s life is known in only two, leaving the other five unknown. It also happens that the two forms which have been studied as to the histogenesis of their electric organs are the only two elasmobranch fishes among the seven, so that we have not as yet seen how the remarkable electric tissues in Malopterurus, Gymnotus, Astroscopus, the mormyrids, and Gymnarchus are developed. Also, of the five teleost types we know the structure of the full-grown electric organs in all of them pretty well, exceptin Gymnarchus. This fish is found in Africa and has been rather rare, so that but two workers have published observa- tions on it, both of them a long time ago and from poorly preserved material. It was, therefore, with great pleasure that the writer came into possession of some embryos of this rare form through the kindness of Dr. J. Graham Kerr, Dr. Arthur Shipley, and Dr. Richard Assheton, to whom he wishes to express his most sincere thanks. This material was collected in Africa by Dr. Samuel Budgett some years ago and was in most excellent condition, owing to the great care and skill with which Dr. Budgett put it up and cared for it. The collecting was done unflinchingly and faithfully, under con- ditions of hardship and sickness that few white men could stand, and Dr. Budgett lost his life from exposure and illness incurred in part by this work. A full account of his trip and of the scientific results should be read in the Budgett memorial volume issued by Dr. Shipley, Dr. Kerr, Dr. Assheton, and others in 1907, through the Cambridge University Press (24).! It is somewhat unfortunate that the structure of the electric organs in the adult fish could not be worked up at the same time that this paper was written, but the writer has material on the way from Khartoum and hopes to publish a second paper shortly. But three papers have been published on the electric organ of this inter- esting fish, one by Erdl (15) in 1847 and another by G. Fritsch (19) in 1885. Riippel’s publication on the subject could not be found, but Fritsch states that Riippel mentioned the peculiar structures which we are considering, and so he stands at present in the writer’s knowledge as the first one to have seen and reported to science the electric organs of this fish, although he was in doubt as to their significance. Erdl used a specimen which was so 1 The figures in parenthesis refer to the literature cited, p. 193. II I6I 162 Papers from the Marine Biological Laboratory at Tortugas. poorly preserved and so soft that when he cut the animal across its body, the electroplaxes and connective tissue ran out of the muscle in which they are embedded. Nevertheless Erdl stated that it was probably an electric organ, coming to this conclusion by a comparison, of such features as he could make out, with the structures found in the tails of the other electric fishes, as Raja, Mormyrus, and Gymnotus. Fic. 1.—General view of Gymnarchus niloticus. (Drawn from a lantern slide made from a figure in Jordan's ‘‘Guide to the Study of Fishes,’"” New York, 1905.) Fritsch (19) in 1885 worked on better material and gave a more com- plete account of the anatomy of the organ, especially of the histology of the electroplax; but he came to the rather strange conclusion that it was not an electric organ at all, assuming the erroneous position that, on account of the large blood-supply, the organ acted in some way as a storage for oxygen during the period of hibernation made necessary by the drying of waters at certain seasons. Fritsch was also mistaken in calling the fibrous Fic. 2.—Scene drawn from descriptions to illustrate the habitat of Gymnarchus and its manner of swimming. (Drawn from lantern slide made from a drawing, by Bruce Horsfall.) contents of the electroplaxes ‘‘connective tissue.’ He went a great deal farther than Erdl, however, in describing the gross anatomy of the electric organs and surrounding tissues. Not having full-grown material, the writer must rely on Fritsch’s figures and descriptions for the adult gross anatomy, although the oldest embryos ee eee Origin of Electric Tissues of Gymnarchus Niloticus. 163 (about 42 days) used for this paper represent practically adult material so far as the histology of the electroplaxes is concerned. A short résumé of the adult anatomy will make a good basis for the embryological descrip- tions to follow. ' The fish (see text-fig. 1) is a mormyrid of elongate form, so much so as to make it almost eel-shaped, although not quite so much so as the Gymnotus of South America. It possesses an extensive development of the dorsal fin, which extends from forward on the neck to within a short distance of the tip of the tail, where it suddenly stops, leaving the tip of the tail naked of fin; whence the name of the fish, Gymnarchus niloticus. This heavy fin, well provided with a series of lateral ray-muscles, is used extensively by the fish as a means of propulsion, by holding the body straight and stiff and causing a series of lateral undu- Fic. 3.—Section through mid-tail region lations to pass from behind for- ward, thus driving the body back- ward (see text-fig. 2) or, from front to rear, which causes the fish to of body of an adult Gymnarchus, showing position of the eight electric spindles and their relations to sur- rounding muscular and bony struc- tures. D, dorsal spindles; U.M., upper median spindles; L.M., lower median spindles; V, ventral spindles. move with its head forward. I Giiver Fetsche}: 26 RabHos a: have not heard, but I presume that in moments of unusual effort /s the animal can swim by means of | the common sinuous body-move- ments used by other elongate fishes, as is well illustrated in the eel. The posterior tip of the creature’s body is interesting. As has been mentioned, this end is free from the fin for some distance (see text-fig. 2). Also it is round in section and ends bluntly. When swimming backwards the animal uses it like a finger to feel its way. The peculiar round and blunt end may be explained by the fact that this tip contains the largest and best-developed portion of the electric organ, which fills the lateral parts of the body at this point almost to the exclusion of the ordinary muscle. As Erdl and Fritsch have described, the electric organ consists of eight long ‘‘tube-like’’ or cylindrical structures, four on each side, embedded in the muscle tissue as close to the median bony parts as a little connective tissue in between will permit. Four of these are present on each side (see text-figs. 3 and 4), and they may be called in order from above down- ward, the dorsal, the upper middle, the lower middle, and the ventral cylinders, or spindles, of the electric organ. In a section cut through the body at a point midway between tail-point and anus (see text-fig. 3) the dorsal spindles are to be found, just above the union of the neural spines of the vertebre, and set closely together with only the dorsal spine and some connective tissue between them. The upper median spindles are more widely separated by 164 Papers from the Marine Biological Laboratory at Tortugas. the neural canal and its contained spinal cord. This second pair of spindles are at the level of the upper part of the spinal cord. The lower median spindles are found much below, on each side of the ventral processes of the vertebra. They lie at about the level of the caudal artery and almost as widely separated as the upper median pair. Lastly, the ventral pair of spindles are found just below the latter pair and slightly below the level of the caudal vein, which lies between them. They are separated a little less than the lower median pair by the narrower bony structures at this point. Upper median z Dorsal . x > ‘Ventral Lower median Fic. 4.—Diagram showing longitudinal position and extent of electric spindles in Gymnarchus. Those of one side only can be shown in figure. The four spindles are largest in diameter in the tail, especially out in the thick, finger-like, naked extremity. From this part they taper to a smaller size as they go forward in the body and they finally become thin and end at points in the neighborhood of the anal opening (see text-fig. 4). All are not of equal length. As Fritsch has shown, the dorsal organ is the shortest and extends for about 20 cm. ina fish of 89cm. length. The ventral pair reach for about 5 cm. further, or 25 cm. in length in a fish of the same size. The two median spindles reach for about 40 cm. from the tip of the tail. Each spindle is marked off clearly from the neighboring muscle, and other tissues which are found next to it, by a distinct connective-tissue covering. In my largest specimen, which is a young fish of 40 days, this is well shown and is exactly like other dividing connective-tissue sheaths that surround the various muscle divisions. Like them, it often contains pigment cells which show golden-brown pigment granules. The important contents of these spindles are alternate, cylindrical segments of a denser, deeper-staining, muscle-like substance, the electro- plaxes; and a connective tissue of jelly-like, grayish transparence which in all ways appears to be similar to the ‘‘electric connective-tissue”’ found between the electroplaxes in the other electric fishes. In this tissue are found the blood-supply, which is largely in contact with the ends of the electroplaxes, principally the anterior end; and the nerve-supply of thick, medullated fibers which run towards, and are attached to, the posterior ends of these organs. In this case, according to both Pacini’s law and the fish’s relationship to Mormyrus, the direction of the current at time of discharge should be from tail towards the head. I have been unable to learn, from the literature of travelers and scientific collectors and observers, if the electric discharge is strong enough to be felt by the hand. Budgett does not mention it, and no other does, so I conclude that it is not a strong shock and that the organ must be classed with the weak electric organs, Origin of Electric Tissues of Gymnarchus Niloticus. 165 as is the case with its relatives, the various mormyrids. The natives of Africa are much afraid of the creature, especially at nesting time, and one of its Arabian names, ‘‘Abu rhad”’ meaning ‘‘father of thunders,” might seem to indicate perceptible electric powers. The embryos and young fishes put at my disposal by Dr. Kerr and Dr. Assheton were five in number, and of these three were the suitable stages from which this paper was worked out. The significant development of the electric tissues in Gymnarchus takes place between the ninth day of embryonic life, at which time the embryo possesses a fully formed and complete musculature in the tail with no sign of an electric organ, and the fortieth day of development, at which time it can be seen that the embryo has developed its electric organs, out of a certain part of the previous muscu- lature in the tail, to a degree that shows the farthest advanced electro- plaxes in a practically adult condition. The most interesting and critical stages in this metamorphosis of muscle into electroplax appear to take place within much closer limits, and stages from the eleventh to the fifteenth day would include them. These signifi- cant changes have been studied and drawn principally from an embryo 12 days of age, fixed in sublimate-acetic, and showing the changes very much to my satisfaction. A point of interest and importance in this study is that, in earlier embryos, the myotomes and electric spindles are youngest, least developed, and growing fastest in the posterior part of the body or nearest the tip of the tail; while in older embryos and in the adult the greatest, most complete, and most characteristic development of the electroplaxes is to be found in the end of the tail or at the posterior end of the spindle. Thus the adult structures in the anterior part of the spindles represent a somewhat inferior and less complete change of the muscle tissue into electric tissue than the posterior parts of the same organs do. The same importance attaches to the fact that the rates of development of the several spindles seem to vary. The lower median spindle starts first to differentiate, extends farthest forward in the body, is larger than the others when developed, and during early development is always in advance of the corresponding parts of the other spindles. The upper median spindle closely follows the lower. The ventral spindle is much behind the two median ones, while the dorsal spindle represents the latest and weakest development and is shorter than any of the others. These facts have made it possible in the present study to get many stages of development from very few embryos. STUDIES OF AN EMBRYO NINE DAYS OF AGE. This little fish was 26 mm. in length and, while the egg-membrane had been ruptured and cast away, the animal was still forced to remain in its nest because of its huge, elongate yolk-sac, still unabsorbed, and because of its otherwise undeveloped organs of alimentation, locomotion, etc. The posterior part of the body was carefully cut into four portions (see 166 Papers from the Marine Biological Laboratory at Tortugas. text-fig. 5) to be sectioned as follows: First, from in front of the anus to a point about 26 vertebral segments posteriorly. The last three segments of this portion were sectioned transversely and serially (region A) while the remaining anterior part was sectioned vertically and longitudinally (region A). Another portion of 19 vertebral segments further caudad was removed and its posterior two segments sectioned transversely (region B), while the anterior part was cut as before in vertical, longitudinal sections ! ' ' ios Se 5 aeansed PE oe ——— i Saag THAT EZ Wi et cale aia i. 8, A, Zag Fic. 5.—Outline of body of an embryo of Gymnarchus nine days old. Transverse lines and letters indicate parts sectioned for study. For explanations see text. (Copied from Assheton in ‘‘The Work of John Samuel Budgett.’”’) xX 5. (region B). A third part, of 18 more vertebral segments, was treated in the same way, except that no transverse sections were taken and the entire piece was cut vertically and longitudinally (region C), while the remaining portion or tail-tip was cut serially in transverse sections and forms a series (region D). Figure 1, plate 1, shows a transverse section through the body a rather short distance from the extremity of the tail or at a point where we are sure that the electric organ will be well developed a little later in life. It may be thought that such a section could be taken for study to better advantage in a more anterior position on account of the earlier anterior development just discussed, but it must be taken into account that the tail segments are being added and are still growing rapidly at this age, that they are very short and very crowded, and therefore the location of this section is, in reality, fairly well forward in the future electric spindles. Conditions were much the same in region C. This section shows a good development of muscle fibers as indicated by the shaded area. As is usual in vertebrates, the most advanced stages of muscle-cell development are to be seen at the lateral periphery of the myotome. Here, at the point indicated in figure 1, plate 1, by the dotted line, a layer of the outer muscle cells, two or three deep, has acquired an average of about 26 myofibrils (an average of 14 counts). These are grouped in one (fig. 2, plate 1, B and C) or sometimes two bundles in the cell (fig. 2, plate 1, A) and their correct spacing and their thickness and staining power indicate muscle cells of normal development and good functional activity. The remaining and inner muscle cells of the myotome, forming its larger bulk, show, as one examines them successively farther inward (toward the median line), a series of earlier stages, until at many points on the inner edge of the Origin of Electric Tissues of Gymnarchus Niloticus. 167 myotome the smallest cells are seen with large nuclei and no myofibrils at all (fig. 2, plate 1, C). These youngest cells are particularly abundant at the dorsal and ventral edges of the myotome. From what we know of the position of the future electroplaxes and their relation to the muscle-masses, we can be sure that it is from the inner edge of the myotome that the electric tissue is to come, and a close scrutiny of the cells which form this edge shows that at two points only is there any indication of such a development. One of these points, marked with a circle (@ ) in figure 1, plate 1, is where the two myotome segments (dorsal and ventral) meet and close in against the notochord. Here we see, in some of the sections, several especially large and strongly developed fibers, somewhat detached from the rest of the myotome and resting against the vertebral disk. Several reasons exist, however, why these fibers do not represent the future electric organ. First, they are in the exact median position which remains constant during growth and in which no electric tissue is to appear. Second, they are very short and are not attached to each other longitudinally, as other fibers are, by means of connective tissue, but are attached to the bodies of the future vertebra. They may be called the vertebral fibers. A second point can be seen where some muscle-tissue shows unusual development. At this point (marked with a circle (O) in fig. 1, plate 1) are several muscle cells which show from Io to 18 myofibrils each, and a degree of development almost equal to the cells in the outer layer. These cells are represented as seen under high magnification in figure 3, plate 1. They are, I believe, destined to be the future electric cells, several of which will unite and form one of the electroplaxes of the lower median spindle of the electric organ. I base this assertion only on their position and their some- what advanced development as muscle fibers, for they show at this time no indication, other than their size, of developing into electric tissue. Their position is slightly too far dorsal for the lower spindle, in an adult Gymnarchus, but when we examine the 12-day-old and 42-day-old stages, we find that the normal growth of the muscle-mass will carry this point to exactly the proper position for that spindle to lie at, in the grown fish. An important point in studying these cells in figure 3, plate 1, is to note that they are separated from the rest of the myotome and from one another by other muscle cells of weak or of earlier development and containing only a few myofibrils. These weaker cells, and even some of the more advanced ones, are destined to degenerate during the development and growth of the electroplax. In figure 3, plate 1, 7 cells are present that will probably take part in the formation of the electric spindle at this point. Some of them are contiguous and others are widely separated. In figure 4, plate I, we have a fortunate longitudinal section from the adjacent region of this same embryo. The section shows in longitudinal view, and under low magnification, the same group of muscle cells marked with the circle (O) at both ends. Also, it shows a portion of the ventral 168 Papers from the Marine Biological Laboratory at Tortugas. part of the myotome with its strongly developed outer layer (0.f.). The vertebral fibers mentioned above are not visible, but their position in a few following sections is indicated by a line one can imagine to be drawn between the points marked with the figure ©. It has been considered unnecessary to figure the myofibrils longitudinally under large magni- fication. The cross-striation is very plainly visible and is the same in all parts. Each of several regions of this embryo was examined and in all its parts were the same conditions found. We may sum up by stating that the embryo of 9 days age shows no electric tissue and but a very weak indication of the development of such a tissue, all of the myotome cells being decidedly of the muscular type. Our only evidence is drawn from future stages, as to the position of such electric tissues and from the otherwise unexplained precocity of the fibers in our 9-day embryo, in that same position. STUDIES OF AN EMBRYO TWELVE DAYS OF AGE. This animal (text-fig. 6) showed a considerable increase in size, not so much in length as in thickness. The posterior part of the body of this specimen was divided into the following pieces: First, from a short dis- tance behind the anus to 14 vertebral segments further caudad. Two-and-one-half segments were cut off of this by serial, transverse sections from the posterior end (re- gion A —1) and the remaind- er was cut into vertical, lon- gitudinal sections from right to left (region A). Second, Bp SEE MENTS ORS: ETE Hea! I reais atl (eae, Pee wee ete eae ete moved and 3 segments of Sra rae aaa (Copied from the same source as fig. 5.) the posterior end were again cut as serial, transverse sections (region B,) and the remaining 12 segments were cut serially in horizontal, longitudinal sections (region B). Third, 14 segments were again cut off and the posterior 3 segments were cut trans- versely (region Cy), while the remainder was cut in vertical, longitudinal sec- tions (region C). The next 18 vertebral segments formed a piece that was cut in vertical, longitudinal sections (region D), without any transverse sections being made, and this left a small bit of somewhat curled tail-tip, which was so young in development that its segments could not be easily counted. This is cut transversely in series (region £). Study of this embryo may best begin by examining a transverse section through the posterior part of the body (fig. 5, plate 2) to see what has become of the inner parts of the myotomes. Looking first for the point L.M. as seen before in figure 1, plate 1 (marked with O), we can see at once that it is present as a compact mass of muscle-like tissue now clearly Origin of Electric Tissues of Gymnarchus Niloticus. 169 separated from the rest of the myotome. Further, there are to be seen on each side three other similar sections of the same muscle-like tissue, all more or less also separated from the main mass of the myotome. The most important part of our study now consists on the one hand in proving that this mass L.M., in figure 5, plate 2, is derived from the undoubted muscle- fibers (marked with O), as seen in figure I, plate 1, and on the other hand in showing that this same structure is to become the finished electric tissue as seen in such advanced development as in figure 21, plate 8, for instance. In tracing it back to the muscle, we are much assisted by the fact that the dorsal and ventral spindles are always in an earlier or in a less complete stage of development than the upper middle spindle, or, particularly, the lower middle spindle under consideration. And, since in this embryo of 12 days the development has gone to considerable length, and since a slightly younger stage, say a 10- or 11-day embryo, was not included among the embryos at my disposal, this fact is of much importance, because it will be fair to take the left ventral spindle as an intermediate step in the compari- son. Figure 9, plate 3, is a highly magnified section of the locality of the left ventral spindle from region C in the embryo of 12 days, and in it we can see a mass of muscle-like cells, closely associated and lying at the inner edge of the myotome. Certain changes clearly differentiate them from the rest of the muscle, however. One change is the fact that the myofibrils have shown a large diminution in size or thickness and have also suffered in power to take the stain; particularly on the periphery of some of the muscle- columns they almost refuse to take it. Their spacing is also irregular and they show a distinct tendency to clumping together and, in some cells, to get close to the nucleus or even to surround it. They can be readily com- pared in figure 9, plate 3, with the well-developed young muscle cells just outside and to the left of them. The dotted line marked XX indicates a separation of the two. All of those to the right of this line show the con- dition of the myofibrils mentioned above; those to the left show the usual condition of muscle cells of this age in fishes. A second characteristic of the changing muscle cells under discussion is in their cytoplasm. It appears more abundant, although this may be due to the smaller fibril bundles. But it also stains more heavily with such stains as eosin, erythrosin, and orange G. With the eosin, for example, it also shows a more yellowish tinge than the cytoplasm of the usual muscle cells in the same sections. And lastly, some of the muscle cells seem to have entirely disappeared or to have greatly shrunken. This latter fact causes a loose and separated condition to obtain among the metamorphosing cells which shows in sharp contrast (fig. 9, plate 3) to the compact con- dition seen in the typical muscle cells to the left of or outside of the line XX. Another important fact can be seen among the changing cells in figure 9. Those near the center of the group show a tendency to touch or coalesce with each other. Already at this early stage this marks a differ- ence in the group. Those cells within the dotted circle are destined to 170 Papers from the Marine Biological Laboratory at Tortugas. unite with one another in a compact bundle to form the future electroplax; while those of the group which lie outside of the dotted ring are to degenerate and atrophy altogether, in order to make room for the growing electroplax. Just mediad of, or to the right of, the dotted ring is seen a cell that has entirely lost its myofibrils and whose cytoplasm is vacuolated and about to be absorbed. At other points, to the left, still other cells can be seen that have dwindled to a smaller size than any of the normal muscle cells. The cells within the ring are judged by the writer to be those which will form the electroplax at this point, because they are starting to unite and, also, because they occupy the position at which the electroplax will lie. Another reason is that their myofibrils are weakening and clumping. Some of those cells lying inside of the ring may also atrophy. This can not be infallibly judged; but certainly all of those outside of it and to the right of the line XX are about to degenerate and are not found in later stages. The connective tissue creeps into the neighborhood of and among these metamorphosing muscle cells at this time, and good mitotic figures can be seen, showing that it is increasing the number of its cells. It does not, however, penetrate the groups of future electric cells inside the dotted line. Also blood, pigment, etc., are to be seen in characteristic positions. A transverse section of the body at region B; need not be illustrated at this point by alow magnification figure, because it is so like figure 5, plate 2, in general appearance. But it happens that in such a section several very interesting stages, forming a sequence of which figure 9, plate 3, can be taken as the first member, were noticed, and figure 10, plate 3, is the second in this series. This drawing represents the right ventral spindle in region Bi, and a number, five to be exact, of the transforming cells can be seen here in closer union than the corresponding cells were in figure 9, plate 3. One or possibly two of these may disintegrate a little later. A dotted ring is not necessary, because the connective-tissue cells have partly marked off the electric cells, and outside of this incomplete ring and above it can be seen two muscle cells which are atrophying. A third cell is shown in a final stage of disintegration. Its cytoplasm is almost clear, or all gone, and its myofibrils have united in a single lump, which will soon become a round droplet or cell-inclusion that will afterwards disappear. Moving to the right upper middle spindle (fig. 11, plate 3), we need very little explanation to see how the four or more muscle cells that first com- posed this structure have come into a still closer union. Below and to the right (outer) are still seen some of the degenerating muscle cells—five in this figure. Several important points must be discussed in connection with this figure; the component cells, as seen here, are not simple, single muscle fibers. The lower one can easily be seen to have two myofibril bundles as well as two nuclei. The presence of two widely separated fibril bundles, as well as the large size of the cell, makes it nearly certain that the structure was formed by the coalescence of two muscle fibers. In the upper region Origin of Electric Tissues of Gymnarchus Niloticus. 171 of the future electroplax, however, a large cell is seen which has only one fibril bundle and four nuclei. It is possible here that this was one cell and that it is growing in size and multiplying its nuclei by amitotic division. As this is the same process which goes on in ordinary muscle cells of this age, it is not surprising to find it going on here, and in older specimens we shall find it the rule. The cytoplasm of all these electric cells is abundant at this stage and is dense staining with the acid dyes. As compared with figure 11, plate 3, the next illustration, figure 7, plate 2, is most interesting and is a step of some magnitude in the development of the electroplax. The cytoplasm of such cells as compose this young electroplax is all united into a single mass and the relation of nuclei to fibril bundles is completed. The nuclei are always peripheral and the fibril bun- dles appear to form a single central mass. This is the permanent condition which will obtain throughout the life of the organ, and is also the condition common to some other electroplaxes, as, for instance, Raja and Mormyrus. Just how the several fibril bundles become massed as a single bundle is not to be positively stated at this time. The individual bundles can hardly be imagined as moving together through the cytoplasm. It is probable that some of the several bundles as seen in figure 11, plate 3, are lost and absorbed, but it is not probable that all but one aresoremoved. The central bundle in figure 7, plate 2, looks large enough to be composed of several, such as are seen in figure 11, plate 3. The best explanation is that several remaining bundles are moved toward each other by growth currents in the cytoplasm, or by the absorption of material which lies between them. At the same time, of course, the peripheral cytoplasm is growing in mass and all nuclei tend to remain in this external layer. We shall see later that a very few nuclei are left behind in the central fibrous mass in some electro- plaxes. In figure 7, plate 2, we see, plainly and indubitably, the first form of the electroplax as found in older fishes. The connective tissues which surround the electroplax are becoming more decided in figure 7, plate 2. Also blood-vessels and pigment-cells are oftener seen as in this drawing. One muscle cell, with its myofibrils clotted into several irregular masses and its nucleus in an advanced stage of dis- integration, is seen just between the lower end of the pigment-cell and the electroplax. It will be well at this point to examine some of the longitudinal sections of these early stages, in order to make clear several points which can not be so well studied in the transverse views. Figure 12, plate 4, is a low magnification picture (X 140) of 6 segments of the caudal part of the body at region D in this 12-day embryo. Only the dorsal part of the body is shown, where a fortunate slant of the section has permitted the knife to pass through both the dorsal and the upper median spindles at the same time. The drawing is an accurate projection from three different sections, so that all parts of each spindle might be 172 Papers from the Marine Biological Laboratory at Tortugas. shown, as well as their relations to the rest of the myotomes. Bracket No. I (zone 1) in this figure indicates the inner tips of the ventral halves of the myotomes. Zone 2 shows a mass of connective tissue and nerve which runs between these two halves (compare fig. 5, plate 2, Conn T.). Zone 3 embraces the inner tips of the dorsal myotomes where they lie between the upper median spindle and the horizontal connective-tissue septum. It will be noticed that this set of muscle fibers appears to be very short, much shorter than the connective-tissue regions between them (not filled out in this drawing). The reason can be seen at once when these same muscle cells are examined under the high power (fig. 15, plate 5, bracket 3; and fig. 6, plate 2), where it is apparent that they are degenerating fibers. Parts of four of these fibers are shown in figure 15, plate 5, and that one nearest the electric tissue which is represented in figure 15, plate 5 (brackets 4 and 4), is the farthest gone, having lost its myofibrils altogether. The nuclei are also distorted and, besides the single large nucleolus, are noticeably empty of any chromatic matter or even of linin (notice the nuclei of healthy muscle and electric tissue). The next fibers to the left in this figure serve well to show how the degeneration takes place from the ends, where the cytoplasm is gathered in heavy lumps which stain light and yellowish with eosin. In fact, there is a singular similarity between the process of degen- eration of these muscle cells and the transformation of the others into the electric tissue. Above, in zone 3, is a rounded cell which I take to be a muscle cell in a final stage of dissolution. It contains large granules of a chromatic substance, which appear to be the remains of myofibrils. Figure 6, plate 2, also represents three fibers from this same region in another myotome and shows three stages of the degeneration of the muscle. Two of them indicate that the degeneration begins in the middle of the fiber. Zones 4. and 5 in figure 15, plate 5, as well as in figure 12, plate 4, show the young electric tissue of the upper median spindle. Let us first consider this tissue in figure 12, plate 4, where we can get a larger, low-powered view of it. In the first place, the most noticeable feature of the electric tissue here is that it shows scarcely any signs of a division into the original myotomes. A trace of this is seen in zone 4, but in a careful and systematic search through a number of the spindles of this stage very few instances could be noted. Even the myofibril bundles of successive myotomes seem to have united, and these bundles were most carefully examined under a Zeiss 2 mm. 1.40 ap. lens. I have no doubt that each spindle does constitute a con- tinuous reticulum of muscle cells at this period. Laterally, the various muscle cells are not continuously united, but, as is shown in figure 15, plate 5, they are united at a number of points much in the way that some heart- muscle fibers are. The myofibrils are best seen, of course, in these longitudinal sections, and it can be seen that they are typical. The transverse striation is as perfect as in the functional muscle, but owing to the slighter fibrils they are not quite so dark. In the upper median spindle of figure 15, plate 5, some Origin of Electric Tissues of Gymnarchus Niloticus. 173 points were found where this striation was slightly weakened. Cases of relaxation and semi-contraction uniformly characteristic of the normal muscle prevailed. The distances between the bands, as well as the length of the anisotropic parts, was the same as in ordinary muscle and remains as long as striation can be seen. This same fact does not hold true in Raja or in Astroscopus, where it is much shortened. It does hold, however, in Mormyrus, which is closely related to Gymnarchus. The longitudinal sections, shown in figure 12, plate 4, and figure 15, plate 5, correspond to the spindles before it has become possible to see that they have segmentally divided into electroplaxes and before a central core of fibrous material has become definitely differentiated from a superficial layer of cytoplasm containing all of, or nearly all of, the nuclei. We will now advance toward the head of this same specimen, to find material in a more advanced stage of development, in order to study the segmentation of the embryonic spindle into its individual electroplaxes (this segmentation has just been described as absent in fig. 12, plate 4), and to study further the differentiation of the inner fibrous core from the outer layer, and lastly to see how the myofibrils lose their transverse stri- ation (muscle striation). The reader will remember that at this stage the anterior electroplaxes are in advance, developmentally, of the posterior ones. Figure 13, plate 4, shows four vertebral segments, taken from the region C of this same embryo, under low magnification. Bracket 1 embraces the zone of the epithelium. No connective tissue is shown. Between the muscle-zones 2 and 4 lies the zone of the long, narrow, embryonic electric spindle, marked by bracket 3. It is the lower median spindle which is shown, one which is always in advance of all the others in development. In this figure it becomes quite plain that, while the muscle tissue is arranged segmentally to correspond to the vertebra, the electric organ is not. In this case the electric spindle is separated transversely by connective tissue into three segments, instead of four. Nor is this proportion always main- tained among the several electric spindles themselves. In some cases, as will be seen later, a single electroplax corresponds to as many as five vertebree. Even the electroplaxes do not correspond with each other. A subsequent figure will also demonstrate this (fig. 23, plate 9). Just what factors do determine the length of the electroplax I am unable to say. A closer study of the nerve distribution and blood-supply may throw some light on the matter. The transverse sections of the stages represented in figure 13, plate 4, do not show anything of interest over and above what was discussed in con- nection with the conditions seen in figures 9, 10, and 11, plate 3, and figure 7, plate 2. We will, therefore, advance toward the head one step further in this embryo and examine a longitudinal section from the region marked B in text-figure 6. Here (fig. 17, plate 6) the electric tissue is seen in what may be considered as its maximum development in this 12-day embryo. The electroplaxes are distinct from each other and have grown considerably in 174 Papers from the Marine Biological Laboratoryai Tortugas. size. The one selected for illustration is from a lower median spindle and shows an actual length of 0.9 mm. It appears with smooth, well-defined boundaries and good separation of core and outer layer. The myofibrils are shown in the core as several closely associated bundles, only to be distinguished from one another by a slight curving and twisting in their course. This would not be visible in a transverse section at most levels in the electroplax, and it gives strength to the view, expressed before, that the fibril bundles of a number of young muscle cells combine to form the single fibrous core of each electroplax. At the levels examined it would appear that a group of from 4 to 7 cells from each myotome is concerned in the formation of any electroplax and that these same groups of from 2 or 3 myotomes are likewise united end to end, thus making it possible that from 12 to 21 muscle cells are united to form each electroplax. The beginning of the loss of transverse striation is quite visible in figure 17,plate6. This striation merely fades or loses its staining power, beginning at several points, but usually at the middle of the electroplax. The ends retain the staining power of the M stripes longest. The multiplication of nuclei must be spoken of at this point. It is well known that the nuclei of future muscle tissue divide by mitosis with a consequent and subsequent division of the cell-body as long as the cells are in the young myoblast stage. As soon as muscle differentiation begins to take place, the mitotic division of nuclei is changed to an amitotic type of division and the cell-body ceases to divide and begins to lay down myo- fibrils. Some few observations against this view are recorded, but it seems to hold in Gymnarchus. In figure 15, plate 5, one can readily see many cases of amitotic nucleus division. The few cases of mitosis visible in the figure are of connective- tissue cells. Such cells, it is known, always divide by mitosis during their whole existence. As the electroplax grows, more nuclei are needed, and they are supplied by the amitotic divisions above mentioned. Just how long this process keeps up is not known, but it is probably mostly done before the first 15 or 20 days of development are passed. It is evidently going on fast in the 12-day embryo, and it is possibly finished in the 42-day stage. The nuclei are typical muscle nuclei in appearance and are not to be dis- tinguished from the nuclei of real and active muscle in the same preparations, except, perhaps, they are slightly larger and have a somewhat heavier chro- matic content (larger nucleolus and chromatic granules). This does not hold, of course, for the degenerating nuclei, whose differences have already been described. Connective-tissue nuclei can at once be distinguished by their delicate outline and small chromatin content, which is distributed as very small granules. STUDIES OF A LARVA FORTY-TWO DAYS OLD. In order to continue tracing this history of the development and growth of the electric tissue, we will now be obliged to pass to an embryo of some one ee Origin of Electric Tissues of Gymnarchus Niloticus. 175 considerable size and one in which the oldest tissues are almost the same, for an understanding of the adult structure as those of a fully grown fish would be. Text-figure 7 gives an outline of this specimen, which was about 63 mm. long and whose tail part was cut off and divided and sectioned as follows: Beginning at about the level of the anus, a portion composed of 6 vertebral segments was sectioned in a vertical and longitudinal direction (region A); then passing caudad, the next 2 segments were cut transversely (region A,); then the next 12 were cut longitudinally and horizontally (region C); then the next 6 were cut transversely (region C,); then 13 others were cut longitudinally and vertically (region D); then 7 more were cut transversely (region D,); then the next 14 were cut longitudinally and approximately vertically (region £), while the tip of the tail, composed of some I2 or more segments, was sectioned transversely (region EF). In the embryo of 12 days the youngest developmental stages of the electric tissue were found in the most distal portion of the tail, which at that time had but recently been extended by growth from the body and was still in process of extension. The oldest and consequently the most-developed electroplaxes were to be found in the anterior or cephalic end of the spindles. In the present embryo, or larva, of 42 days, the conditions are reversed, and the electroplaxes in the extremity of the tail have passed those farther up in the body in their differentiation, and have reached a much greater and more complete development. Accordingly we will select for study one of the most anterior and least developed examples and compare it with that electroplax last studied, which is represented by figure 17, plate 6. Figure 18, plate 6, is drawn from one of the ventral electroplaxes of the region C, as shown in text-figure 7. This position makes it fairly well forward in the spindle, although portion B would have shown slightly younger stages. Fic. 7.—Diagram of body of an embryo, or larva, of Gymnarchus, about 42 days old. Lines and letters indicate regions studied. For explanations see text. (Copied from the same source as text-fig. 5.) XX about 1 and 1.5. The fibril core will first attract our attention. The first noticeable feature is that this core is growing in mass and volume all through the electroplax, but also far faster in its center than at either end. Throughout its course it has assumed a unified appearance which shows no trace of the several fibril bundles which have gone to compose it. In the narrower ends the mass is straightest and its component fibers appear most parallel, being but slightly wavy. As we follow them from either end towards the middle it can be seen that their course becomes more and more wave-like, until, in the middle, they have been thrown into decided folds. Their actual 176 Papers from the Marine Biological Laboratory at Tortugas. course can be somewhat better followed with thicker sections and deep eosin staining than with preparations of the usual kind. As to the fibrils themselves, they have lost the transverse striation altogether. By this I mean that it is no longer stainable. It seems that, in the straight ends, with a good immersion lens one can see traces of this striation, left as a thickening of the fibrils, at points that may represent the previous position of the anisotropic M spindles. Whether the increase in the mass of the fibrillar core, which goes on as the electroplax grows, is due to the thickening of the individual fibrils or to the laying down of more and new fibrils or to the deposit of an interfibrillar substance between them, was not decided. The first two conditions seemed the most probable, because of the apparent absence of much interfibrillar substance in the oldest and largest electroplaxes. The cytoplasm of the outer layer begins to be of interest at this stage. Its most particular point of interest lies in the fact that it is decidedly different in structure at its anterior end from its posterior end. This is shown in its staining capacity as well as in its actual structure. In the specimens stained in iron hematoxylin and eosin the cytoplasm at the posterior end stains deeper than that at the anterior end, with both dyes. It can also be seen to be granular in structure—a sort of general granulation with a few larger granules of a substance which stains somewhat like chro- matin. In particular, its cytoplasm is darker than the fibrillar core at this posterior end. As the cytoplasmic layer is examined in an anterior direction, it is seen to stain lighter and at the same time to contain an occasional vacuole. This condition increases until, at the anterior end, about one-fifth of the length of the entire structure is covered with a cytoplasm which is so much vacuo- lated that it appears as a delicate reticulum in the meshes of which the nuclei lie. The fibril core extends out to the end or almost to the end, and it can be seen that it is darker in color and denser than the cytoplasm covering it. Some of the finer, granular material is found out as far as the reticulated tip. A new element of interest begins to become apparent in this stage, and that is the point of attachment of the nerve. The strong medullated fibers come from the cord, pass through the spinal ganglion, and enter the connective-tissue ‘‘tube’’ of the spindle at the level of the posterior third of each electroplax. The fibers wind and turn considerably, in this vicinity, and are finally applied to the surface of the electroplax over an area which may be described in this specimen as its second sixth part from the posterior end. Thus the extreme posterior end does not receive any nerve-endings at this period. The cytoplasm shows a number of indentations where the nerve is ap- plied and the axis-cylinders of the nerve can be traced into these spaces, which they apparently fill with a club-shaped nerve-ending. This ending can not be satisfactorily described until some of the special nerve methods can be used to elucidate it. Origin of Electric Tissues of Gymnarchus Niloticus. Li Another and more simple step in development is indicated in figure 19, plate 6, which was taken from an electroplax in a dorsal spindle of this embryo at the region E (see text-fig. 7). Two points of interest will be spoken of in connection with this stage. The shape has changed as follows: The middle part has both actually and comparatively widened over the breadth shown by the middle part of the electroplax seen in figure 18, plate 6, and this increased width is due to a broadening of the fibrous core alone, the outer nucleated layer remaining the same. Accompanying this widening is also an actual shortening of the structure. Thus, part of the increased bulk of the middle is due to an absorption of the two ends. Apparently the anterior end suffers the greater amount of absorption, for it is usually shorter and thinner than the posterior. In this connection we must remember the previous vacuolization of this anterior end as seen in figure 18, plate 6. It would appear that the vacuoli- zation was part of an absorption, or rather of an atrophic process. Con- siderable traces of it remain in figure 19, plate 6, and it is also still noticeable that the anterior end does not stain deeply. The posterior end retains its length and vigor to a greater degree. It seems to become considerably narrowed, however. The form of the electroplax also begins to show a marked change through the development of spurs, branches, or papillae which begin to protrude from both its ends, particularly from its posterior end. Some indication of this was visible already in figure 18, plate 6, as small lumps or ‘‘shoulders”’ that marked the organ roughly into three parts—anterior, middle, and posterior thirds. These growing spurs begin to give the middle third a distinct truncate or cylindrical form. The nerve fibers also attach themselves to the sides of the growing papilla, particularly to their bases. The papillz show an inclination to grow out of the sides of the posterior third. We will now pass to the last and oldest stage of the electroplax which can be found in the embryos which Budgett collected in Africa. This is found in the £ region of the 42-day embryo or larva of Gymnarchus and is represented in longitudinal sections under a low magnification by figure 23, plate 9, which shows parts, or the whole, of about 11 electroplaxes lying in place in the ventral, lower middle, and upper middle electric spindles of this region. Figure 21, plate 8, also shows a transverse section just cephalad of this point, in the D region, where cross-sections of the 8 electric spindles reveal partial or complete transverse sections of 7 electroplaxes, but only the electric connective tissue which fills the tube at this point, in the eighth. Figure 20, plate 7, represents a longitudinal section of one of the electro- plaxes shown in figure 23, plate 9, and will serve as a basis for our last study of this series. The form has continued to follow the development which was indicated in figure 19, plate6. The total length of the structure has shortened somewhat more, while its width in the middle part has increased twofold. These I2 178 Papers from the Marine Biological Laboratory at Tortugas. statements are the result of averages of about 15 measurements. The papillze on the posterior surface have increased in length and some of them have begun to rival the posterior portion of the organ in length. We have no suitable figures of the adult electroplax, but from Fritsch’s (3) descrip- tions, and from the tendencies shown by these embryos, it would seem that, as Ewart describes in Raja, the original posterior portion of the electroplax shortens and the new papille lengthen until they all form approximately similar structures. This can only be completely studied when we have secured suitable sections of the grown fish. The development of papillz is noticeably weak on the anterior surface. The figure does not show as many as some of the electroplaxes in figure 23, plate 9, but it is a fair illustration. Neither does it show well the usual condition of the main anterior process of the cell at this time, which can be better seen, in some electroplaxes of figure 23, plate 9, to be still in evidence and of considerable length, but of very weak development. This anterior process shows no trace at this age of the general vacuolization of its cyto- plasm which we saw in an earlier stage, and the fibrils extend as a very thin and uncertain core through its length. The fibrillar mass which forms the core of the electroplax has now assumed what appears to be its permament condition. The fibrils are very fine and, after having passed straight down through the posterior process, are thrown into flat waves by the process of packing them into the shortening and widening middle part which now constitutes the principal bulk of the structure. This causes the larger part of the fibrils to lie nearly at right angles to the longitudinal axis of the organ, and this appearance was first taken by the writer, who examined the oldest embryo first, to be a trace of the transverse striation of muscle from which the organ is formed. We now know it to be the myofibrils, lying at right angles to their original course. No conclusion was arrived at, in this stage, concerning the growth of this mass, as to whether the number of fibrils was increased, or whether the larger size was due to a thickening of the original fibers, or to the deposit of interfibrillar substance. The fibrils seemed to be as fine if not finer than in the earlier stages; certainly they are much finer than functional myofibrils. Since muscle tissue increases the number of its myofibrils long past this age, I see but little reason why this modified muscle should not also do so. A closer study of the cytoplasm and nuclei of the peripheral layer was next undertaken in this oldest stage. Beginning on the posterior surface and on the papille, we find the layer thickest here and composed of at least three distinguishable materials. One was a dense material which was reticular in structure and stained with chromatic stains deeper than almost any other pure cytoplasm that I know of. This material was one con- stituent, while the other substance composing the general field at this point was a far lighter staining material which was homogeneous and clear. This latter seemed to bear the same relation to the denser material that the ‘“‘nuclear sap”’ does to a linin alveolum or reticulum in the nucleus. Like Origin of Electric Tissues of Gymnarchus Niloticus. 179 the linin reticulum, this denser material was more refractive and took more of the chromatic as well as more of the acid counterstains than the homo- geneous material did. I shall adopt Schneider’s (31) name of “‘Linom”’ for the denser substance and his name of ‘‘ Hyalom”’ for the clearer and more homogeneous material. It seems probable from the works of Biitscheli (9), Rhumbler (27), Hardy (35), Wilson (34), Andrews (1), Strasburger (32), and others that these structures of cytoplasm, as seen under the microscope in fixed material, do not represent the exact condition as it exists in life. After reading the pages of my recent work on the electric-motor cells of Torpedo, one can more easily see that the cytoplasmic linom is a structure which depends upon the fixative used as one factor and upon the chemical and physical peculiarities and the contents of this plasma at the time that the fixative is applied, as another set of factors. Its reticular or alveolar arrangement can most certainly be immensely varied, and all these artificial conditions must be very much different from that which obtains in life. Since I have only a few fixed specimens to discuss, over whose earlier prepa- ration I had no control, I shall not try to solve the question of what the structure was during life, but will merely describe the present specimen as it appears. The linom of the cytoplasmic layer on the posterior end is very fine and can only be seen with the best lenses and under the best conditions. This holds particularly for the outer portion of the layer, for as we examine the inner portion the reticulum grows larger meshed until, at its point of contact with the fibrillar core, the meshes are quite easily seen. The same is true as we examine this layer in a more anterior position. Here all the meshes are proportionately larger, until in the layer, as found covering the extreme anterior surface, the meshes of the linom are visible with ordinary high powers. I do not refer to the vacuoles which are found at various points, for there is a great difference between the largest meshes and the vacuoles, although the vacuole may be derived from, or originate in, an overgrown mesh. The meshes always contain the hyalom, while the vacuoles do not. My definition of a vacuole in this case will be a space in the linom into which the hyalom does not extend. Also, the rounded outline of a vacuole shows a surface tension between its content and the cytoplasm, which does not seem to exist between a mesh of the linom and its contained hyalom. These vacuoles appear to have contained a soluble substance during life, and the hyalom is not soluble in fluids used to prepare the specimen. Such vacuoles are found at various points in the cytoplasmic layer, most often at its outer edge, while the large meshes appear at the inner edge and around the nuclei. The vacuoles become so large and numerous at the anterior end of a stage such as figure 18, plate 6, that the cytoplasm looks like foam containing the nuclei and surrounding the unchanged fibrillar core. In addition to this structural basis of linom and hyalom, another very prominent content of the cytoplasm is a series of granules of some material 180 Papers from the Marine Biological Laboratory at Tortugas. that is denser and somewhat more stainable than either of the others. These granules are very numerous and fine at the periphery, where they cause the outer part of the layer to stain deepest with eosin, and they lie in the finer-meshed linom, and grow larger and fewer as one examines the inner parts of the layer. They are of a slightly more refractive quality than the linom and stain deeper with chromatic dyes the larger they get. The largest also have some color of their own, a slight golden-brownish color, somewhat like that of pigment granules. While the smallest granules seem to lie in or attached to the strands of linom, the larger seem to lie in its meshes in the hyalom. They remind the observer of the granules found in certain other cells, particularly in the electric-motor cells of Torpedo, as well as in other nerve-cells. In figure 16, plate 5, these finest granules are seen in the outer part of the layer stained with eosin, while the inner part does not show the larger ones that lie there. The larger granules are prone to gather about the nuclei and about the inner part of the electric layer, as seen in figure 22, plate 8, where they took the iron hematoxylin stain well. The larger, chromatic-staining granules are also found, in groups or singly, in many parts of the fibrillar core. While the larger ones found in the core, around the nuclei, and at the inner edge of the cytoplasmic layer seem to stand in sharp contrast to the more numerous and finer ones found in the outer part of the cytoplasm, one can trace steps between the two kinds. The writer believes these granules to be the secreted or prepared nitro- genous material used by katabolic processes to produce electricity, either directly or indirectly. They must be, physiologically, the same granules described by Ballowitz (5) in the cytoplasmic layer of the electroplax of Raja, and by Schlichter (30) in the cytoplasmic regions on both surfaces of the electroplax of Mormyrus. In this latter they were very large and not easily demonstrated. The writer has seen some indication of them in Astroscopus (12), but they would seem to be noticeably absent in other forms, as Torpedo (although they have been figured here by Fritsch (20)), and in Gymnotus (3) and Malopterurus (6), where such granules as Ballowitz has described or figured would seem to be inadequate in size and number for so heavy a duty. It will be noted in the above list that the weak- electric fish seem to have these cytoplasmic granules best developed, while the strong-electric fish show least of them. It is possible that these latter possess them, or their equivalent, in a more fluid and less visible form. The cytoplasmic layer is marked off from the fibrillar core by a very definite and sharp line. It can not be said that a definite membrane exists here, although one may exist. The boundary between the outer layer and the core is not an even or a straight one, but shows a wandering course, especially at the two ends. At some points strands of the linom seem to branch into the core (fig. 16, plate 5). At the anterior end in particular it shows many extensive and complicated invaginations of its granule-containing substance into the fibrillar core. Certain small portions Origin of Electric Tissues of Gymnarchus Niloticus. 181 of it also have been permanently left in the main body of the core, usually near its anterior end. These inclusions (fig. 20, plate 7) may or may not include the nuclei. They always contain some of the largest of the granules. The nuclei do not show any internal peculiarities which would dis- tinguish them as electric nuclei from some of the other tissue nuclei, par- ticularly from muscle nuclei, which they much resemble. They have a large chromatic content and a particularly large plasmosome which stains deeply. They can be sharply distinguished from some other nuclei, as connective-tissue nuclei for instance, where the more delicate outline and different chromatic pattern is discernible at a glance. Each nucleus shows some sign of a surrounding differentiated layer of cytoplasm. This consists of larger granules and a zone in which the hyloplasm seems to be in greater proportion than elsewhere. At different places in a preparation one may see more or less of a contraction zone around the nuclei. While this may be a physiological condition, it is more probably an artifact due to the fixing or hardening. One interesting condition is to be seen in most of the few nuclei which become detached from the outer layer and included with some small portion of the outer cytoplasm in the fibrillar core. These nuclei probably become so placed during a very early stage, and the further they are separated from the layer to which they rightly belong, the larger they grow and the more diffuse their chromatin becomes. The plasmosome diminishes in size and the whole structure looks more like a connective-tissue nucleus, except that it is very much larger. I have seen this same condition in the electro- plax of Raja. It was not possible to find a real electrolemma or cell-membrane covering this electroplax. A connective-tissue covering, more or less closely applied to the surface, was always present, but the fact that this covering possessed its own nuclei seems proof that it was a real connective-tissue covering and not any product of the activity of the electroplax tissue. At such points, as this connective tissue did not actually touch the electroplax, a careful examination was made to see if some actual cell-membrane did exist. Beyond the fact that the outer edge of the electric layer was sharply defined and that its surface was rounded and even as if some membrane was present, no real membrane could be demonstrated, either by its refractive properties or by its color. INNERVATION. A general survey of the innervation is desirable, as too little exact topo- graphical work has been done on those fishes in which the electric-motor centers are thinly distributed in character over large areas of the cord, as, for instance, in Gymnotus, Raja, and the mormyrids. Regions D and E were selected in the 42-day-old embryo as the most favorable parts to study. The work was not as exact as it could be if the investigator had had plenty of live material, especially adult material, on which to use some of the well- known neurological methods. But even in this embryo, which was well 182 Papers from the Marine Biological Laboratory at Tortugas. fixed in sublimate-acetic, much could be accurately made out, and it is hoped, too, that the description will prove suggestive. The motor cells were first looked for in the spinal cord, especially that region which was posterior and in the neighborhood of the electric organs. Assuming for the present, as a law, that electric-motor cells are larger than muscle-motor cells which innervate an equal weight of muscle, it was very easy to find groups of large, heavy nerve-cells scattered through the sub- stance of the spinal cord from near the anterior beginning of the spindles to the very last point in the tail to which they extend. Fic. 8.—Side view of reconstruction (semi- diagrammatic) of spinal cord and motor electric nerves of a larval Gymnarchus 42 days old. Arrow indicates anterior and posterior directions. Ni, No, Ns, and N4 are the four electric nerves formed by branches from motor roots of spinal nerves. E indicates small branches from electric nerves that innervate posterior surfaces of electroplaxes. g marks spinal ganglion, whose afferent nerves have been cut off. These cells were situated just above the central canal and from I to 4, or even 5, could be seen in every transverse section in most of this length (fig. 14, plate 4, A, B, and C). They did not appear to be divided into two symmetrical groups, but rather to lie in one median group. On the one hand, their dorsal position might appear to be evidence that they were not motor-cells or that they were not to be considered as modified muscle-motor cells. On the other hand, the presence of real ventral muscle-motor cells (fig. 14, plate 4, D and £) in the whole length of the cord, the correspondence of the cells under discussion with the position of the electric spindles, and the fact that Fritsch describes similarly placed cells in the same position in mormyrids as the motor-nerve cells of the electric organ, all seemed to constitute very strong indirect evidence that these were the nerve cells which innervated the posterior ends of the electroplaxes. In addition, the writer was able to trace, in a number of cases, the course of the neuraxons through the cord, out into the ventral roots of the nerves, and finally into special bundles of nerve fibers that undoubtedly innervated the electric tissue. No one nerve process was actually traced for the whole distance unbroken, but the various parts and regions were so pieced together that the course was well established and corresponds in many ways with Fritsch’s observations on Mormyrus. Text-figure 9 shows a semidiagram- matic cross-section of these courses as seen from in front, while text-figure 8 shows the same thing sketched in relief from the side. This topography will presently be described. a Origin of Electric Tissues of Gymnarchus Niloticus. 183 The electric-motor cells at this age must of course be considered as still very young, and descriptions from the adult are desirable. We will begin by examining the 9-day-old embryo once more to see if they have started. In this cord (fig. 8, plate 2) there are but very faint traces of any nerve- Fic. 9.—Diagram to show position and relations of ‘nerve elements to electric spindles. S.C., spinal cord, showing central canal and four motor elec- tric nerve-cells. Processes from these cells pass out through ventral roots, and distribution of these roots to electric spindles and muscle-masses is in- dicated. M,some of the muscle-masses. D, U.M., -LLN. 2L.M., and V show electric spindles on one side. S»p.G., spinal ganglion; L.L.N., lateral line nerve. p-G cell development, and some neuroblast mitosis is still going on. In the position to be later occupied by electric cells a little enlargement of nucleus and cell-body is visible. One cell here is of great interest, and that is one of the now well-known ‘“Hinterzellen”’ or giant cells, first described by J. Beard (37), Rohon (29), Studnitzka (33), and others, in the embryos of Salmo and Raja, and later by Fritsch (21), as a different sort of cell, in the adult Lophius. Still later, such cells were described by the writer (36) in the embryos and adults of various Pleuronectide and in Pterophryne, where he showed it to be the same cell in both embryo and adult, and described the relations of anterior and posterior branches of the neuraxon. From its size, shape, and position in the present specimen, it seems that this dorsal cell might be, in some way, related to the electric-motor cells, but that question is easily settled when one examines the 12-day embryo and finds that all of the dorsal cells have effectually disappeared before the electric- motor cells begin to differentiate. Besides, the well-known fiber courses of the dorsal cells as worked out by Fritsch in Lophius, the writer in Ptero- phryne, Harrison in Salmo, and Johnson in Catostomus are sufficient proof that the two have nothing in common. 184 Papers from the Marine Biological Laboratory at Tortugas. The 42-day larva shows the best development of these electric-motor- nerve cells until an older fish can be described. Here we find a heavy, rounded, cytoplasmic body of about 40 microns in diameter (fig. 14, plate 4, A, B, and C). Its outline is usually pear-shaped, with the axis-cylinder process given off at the pointed end. Dendrites are not visible in the specimen at hand. While most of these cells are of the same size, a few are noticeably smaller. The nucleus is large, round, and placed nearly always in a very eccentric position. This position is most frequently on the side away from the neurite, but sometimes it may be very close to the neurite. Its diameter is about 16 to 19 microns in the largest cells of this age, and its outline is hard and round. The nucleolus consists of a single, or rarely double, plas- mosome, which is a very little less than 5 microns in diameter, but which, when double or multiple, is of proportionally smaller size. As I have shown in the nucleus of the electric cells of several torpedoes, when a plasmosome is multiple its several parts, collectively, are larger in mass than is a single normal or usual plasmosome. Of course, the nucleus was carefully examined as to any possible ori- entation of its nuclei, particularly the plasmosome, with reference to gravity or to the electric current or to the axis of the cell. Nothing of this sort could be found, although this does not preclude such a condition in the grown fish. It is known that in Torpedo, where such an orientation does exist, this same orientation is not found in the embryonic or larval stages. From these cells thin, delicate axis-cylinder processes were traced down, as has been described, and into the ventral roots of the spinal nerves in a sufficient number of cases to assure the observer that they all ran in this direction. The process was thinnest shortly after leaving the cell, and became thicker as it approached the nerve-root. When it once entered the root it again became very thin, although now invested with a connective tissue and a medullary sheath. A large number of connective-tissue nuclei, nerve-sheath nuclei, and some unknown elements cause the motor-root of the nerve to swell to some size just after leaving the cord. It decreases again in size before it enters the foramen and leaves the vertebral canal in company with the much smaller dorsal root. Just outside the canal the dorsal root ends in the spinal ganglion, and the motor-root traverses the inner side of the ganglion, from which it emerges on the lower edge, and at once divides into a dorsal and a ventral branch. The ventral branch passes backward and downward (see text-figs. 8 and 9) to a point at a level with the lower middle spindle, and here it gives off a considerable group of fibers which pass caudad just inside of the spindle capsule. These fibers are joined by similar groups from others of the spinal nerves and the whole mass forms the lower middle electric nerve (text-fig. 8, V3). At the posterior level of each electroplax this nerve gives off a few fibers (text-fig. 8, £) which branch out and innervate the posterior surface of this electroplax. Origin of Electric Tissues of Gymnarchus Niloticus. 185 The remainder of the ventral branch passes farther down and again gives off a branch which goes to form part of the ventral electric nerve (text-fig. 8, Ns). This latter sends off little branches to furnish the posterior surfaces of the ventral electroplaxes with motor-fibers (text-fig. 8, £). Going back to the anterior root, we find that its dorsal branch leads directly upward and, passing between the spinal ganglion and neural arch, slopes gently backward to give off a large branch which becomes a part of the upper median electric nerve (text-fig. 8, Ne). Its remaining fibers reach upward and furnish the dorsal spindle with a part of the fibers that form its dorsal electric nerve (text-fig. 8, N;). Of course, there are muscle-motor elements in both these branches of the anterior nerve-root, the number depending upon the position, back- ward or forward, at which we examine the arrangement. Two examples of the motor-nerve cells of the muscle-tissue from the cord in region D are shown in figure 14, plate 4, D and Z. At the level of the anterior parts of the electric organ, when there are large quantities of muscle and the electro- plaxes are very small, the muscle branches are large, particularly the dorsal branches, which have to supply the muscle-bundles for the large dorsal fin. In the posterior region of the tail, on the other hand, the muscle is almost entirely absent and the muscile-branches of the anterior roots are not even easily seen. The eight (four on each side) longitudinal electric nerves are interest- ing in that they form a morphological buffer between the conflicting segmentation of the nervous, skeletal, and muscular systems, and the inde- pendent segmentation of the electric system. Were the electric segmenta- tion to correspond with that of the others, we should not find these nerves in this recognizable form. In fact, in the anterior part of the electric organ we do not find them as continuous nerves, in many places, for more than two or three neuromeres at a time. And even posteriorly where they do form continuous nerves, the fibers that enter them from any given spinal nerve do not pass very far back in them before leaving to innervate one of the electroplaxes. Each nerve cell probably lies but a very short distance in front of the electroplax which it supplies. This was decided upon by plotting the relative positions of all motor-electric cells and all electroplaxes in the larger part of the organ. While doing this it was also determined how many nerve-cells sent their nerve-processes to each electroplax. Thus, in region E of the 42- day-old larva there were easily counted 414 of the electric-motor cells, while in the same part there were 81 electroplaxes. This makes it quite sure that, on an average, about 5 of the nerve-cells were used for each electro- plax. It was attempted to count the nerve-fibers as they left the electric nerve to branch out over the electroplax, but the elements were too small and the fixation not just what was needed to do this. It could be easily done in a grown specimen. I shall now take up the structure of the nerve-fibers as they leave the 186 Papers from the Marine Biological Laboratory at Tortugas. electric nerves to pass to the electroplaxes, assuming that they have no peculiarities of interest during their course through the nerve-tracts between the cord and the electric chambers. (The electric chambers are not divided as in Mormyrus by a transverse septum of heavy, opaque, white connective tissue; this can be explained by the fact of their secondary segmentation.) As the few fibers destined for any particular electroplax ‘first leave the electric nerve they are directed caudad and are very small in diameter and invested with a fine connective-tissue sheath. This covering is probably medullated in the grown fish. They quickly turn in a gentle curve whose diameter is about half that of the spindle and pass forward through the electric connective tissue to the posterior surface of the electroplax. At the beginning of this curve, or just after leaving the electric nerve, the axis cylinder or neuraxon enlarges to a considerable diameter and becomes very wavy and irregular in diameter. Its course is no longer straight, and it is not possible to find a section of any considerable part of its length, except in some very few cases. Its sheath of connective tissue is loose fitting and the inner side shows a loose reticulum of fine fibrils and plates that form a weak connection between the axon and sheath. It is at once apparent in the maze of fibers which approach the electro- plax that the few neuraxones which first entered the compartment are now many in number. Still, it is difficult to see where they branch in the thin sections, owing to their sinuous and irregular courses and to the large numbers of transverse and oblique sections present. In several places this branching was seen, however, and recognized as the same multiple branching which has often been described in the terminal part of nerve-paths and, in particular, in the same comparative region of the electric tissue of Raja by Ballowitz (5) and Retzius (26), in Torpedo by several authors, and especially in Malopterurus by Ballowitz (4), who refers in this article to many other cases. The writer has also seen and figured it in the electric tissue of Astroscopus in a paper soon to appear. In the present case, one section, as can be seen in figure 16, plate 5, exhibits two cases of this branching, one of them showing a single fiber dividing into at least three or four branches; also, figure 25, plate 9, in which there is to be seen a well-defined single fiber dividing into two branches just before they end in two club-shaped nerve- endings in the electric layer of the posterior surface of an electroplax. The abundant nodes of Ranvier, described by Schlichter (30) in the case of Mormyrus, were not observable here, probably on account of the lack of osmic-acid fixation. I have no doubt that they are present and they must be present at the points at which the fibers divide. As has already been stated, the fibers now approach and end on the posterior surface of the electroplax, as well as on the lower sides of some of the papille that arise from it. The mode of ending is not difficult to see, apparently, although this much studied and controverted question should not be approached lightly, especially where the material is merely a subli- mate-acetic fixation stained with iron-hematoxylin and eosin. Origin of Electric Tissues of Gymnarchus Niloticus. 187 The nerve-process, carrying its connective-tissue sheath until it actually reaches the surface of the electroplax, ends in a blunt and somewhat thick- ened knob which is embedded in a hollow or invagination in the substance of the posterior cytoplasmic or electric layer of that structure (see figs. 18 and 109, plate 6; also fig. 16, plate 5; also figs. 24 and 25, plateg). This knob may be quite elongate in form and in some cases appears to be branching. The nerve fiber becomes very narrow and apparently dense just before entering the cavity, but it quickly broadens out, to as much as or more than its previous width, to fill the cytoplasmic cavity of the electroplax. Its substance becomes very light-staining, more so than any other part of its length is, and this light-staining quality is most apparent at its extreme distal end in the cavity. The nerve-fibrils, faintly visible in the outer courses of the axon and rather more so in the denser neck just before entering the cavity, can be seen in the light-colored club-shaped ending to be running in an irregular reticulum instead of in their previously almost straight and parallel manner. They could not be traced into the protoplasmic bridges between the club-shaped nerve-ending and the surface of the protoplasmic cavity in which it lies. In very few instances did the nerve-substance fill the recess in the electric layer tightly, probably owing to some shrinkage in both of the tissues. The small space between the two surfaces is crossed by the numerous fine processes or strands, mentioned above, of some of the nerve-tissue remnants, probably of an original closer contact. The question now arises as to whether the cavity or depression which the nerve-ending occupies is to be considered as an invagination of the surface of the electric layer or as a real penetration of the electroplax by the nerve. I shall consider it as the former, because the surface of the electroplax appears not to be interrupted by the opening but to continuously follow the inner edge of the cavity all the way around. The presence of a perceptible cell-membrane or electrolemma would assist in the solution of this question, but such an organ could not be demonstrated. The edge of the cytoplasm was sharp and definite, but no membrane was visible, either by its staining properties or by refrangibility. It possibly will be found in the adult organ. Nor was it possible to see the “‘Stabchen”’ or rodlets which have been described in other electric organs by Ballowitz (3, 5) and his pupils (30). Some striated arrangement of the granular electrochondria or granules described previously was observ- able, but this constituted a fibrillar secretory striation, such as is seen in the surfaces of most cells that are undergoing exchanges of any kind. Here, again, we must await examinations of the adult electroplax before we can say if such ‘“‘Stabchen”’ are present; also if they are homologous with the well-defined rodlets found by Ballowitz in Torpedo and in Raja, and whether they are specific structures of any importance in the production of electricity. Returning to the relation of nerve-ending to electroplax, we have practically decided that the nerve-club is embedded in an invagination of the surface rather than in a cavity in the substance of the electric layer. 188 Papers from the Marine Biological Laboratory at Tortugas. In doing this it has formed a much more intimate connection than may be seen in other places where the nerve-fibers touch the electroplax very closely, even being partly embedded in it, or running through the fundus that lies between two papille, where the fiber is closely pressed on three sides by the surface of the electroplax. In these fiber contacts the con- nective-tissue sheath persists, while in the club-shaped or heavily rounded endings the connective-tissue sheath is lost, left at the surface, and the little protoplasmic bridges, shown when slight shrinkage has taken place, testify to the intimacy of the contact. In addition a slight amount of fine, golden-colored granules surrounds the nerve-ending, lying on its surface, between it and the substance of the electroplax (figs. 24 and 25, plate 9). These granules are not found on any other part of the nerve surface. Naturally the paper of Schlichter (30) was carefully examined to see what he had found as to the ending of the electric nerves on the related form, Mormyrus oxyrhynchus. He had adult material, but otherwise was no better off than the writer in the possession of material which had been treated especially for neurological study. He describes the nerve-fibers with their medullary sheaths as coming in contact with the large process of the electroplax and then suddenly ending just as they reach certain large inden- tations of the surface of this electroplax. He found in these indentations only a little coagulated material and some nuclei. The writer has no doubt that the slight coagulum represents what remains of a club-shaped nerve-ending similar to that which he finds in Gymnarchus. The nuclei, from their position in Schlichter’s picture, are evidently the nuclei of terminal connective-tissue coverings. If this idea be correct we will have a very simple but interesting form of nerve-ending, much larger in size than that found on any muscle or any other electric organ and one in which it will be, apparently, easier to study the intimate contact of nerve-substance with motor-substance from a cytological point of view than in any other form. In particular, we should try to stain these endings with the nitrate of silver and methylene blue methods devised for neuro-cytological studies. This work, however, can be undertaken only on the ground, with good laboratory facilities and with an abundance of fresh material. The writer has published observations on some peculiar horizontal, pointed rods, or pointed threads, found imbedded in the electric layer of the electroplax of Astroscopus. At that time he suggested that they might be in some way homologous with or related to the ‘‘Stabchen”’ because of the absence of any other well-defined ‘“‘Stabchen”’ in this fish. Such structures are not found in the present Gymnarchus larva, but they have been seen and described in Raja, in a paper soon to be published. Their presence in Raja, in addition to the ‘‘Stabchen,”’ proves them to be entirely different cell organs. One word in regard to certain possibilities for the physiological study of the electric organ in Gymnarchus. A recent paper by Bernstein and Origin of Electric Tissues of Gymnarchus Niloticus. 189 Tschermak and another by Bethe go to show that the electric discharge of Torpedo and other fishes is produced by different concentrations of sodium chloride in the electroplaxes and in the intervening electric connective tissue. Since there is a long series, theoretically, of these alternate segments of higher and lower degrees of concentration of the electrolyte, and since all the higher concentrations are presumably equal, and all the lower concentra- tions are equal, we would have a series of equal potentials alternately opposed to each other, and the result would be zero or else only the strength of one concentration current, in case there was one more or one less of either of the concentrations. To obviate this difficulty a membrane has been imagined, on one surface of each electroplax, presumably the electric or nerve-ending surface (pos- terior surface in this case), which will be permeable only to one kind of the ions, either negative ions or positive ions, and by which the current is thus rendered integral in one direction. Two things remain to be proved in connection with the above theory— the fact of different concentrations and the presence of such a membrane. This can be done in any fish in which it is possible to effectively separate the segments in a fresh state, so as to submit them to delicate chemical tests. In Gymnarchus we have a fish whose elements are larger than those in any other one of the seven electric types—large enough to be cut apart, I believe, and analyzed separately in the chemical laboratory; also large enough to submit anterior and posterior surfaces to physical tests that may show its permeability to either positive or negative ions and its imperme- ability to the other kind in one direction. This experimental work can most certainly be done if the proportion between the bulk of electric con- nective tissue and electroplax remains the same in the adult as in the larva (see fig. 23, plate 9). Fritsch shows much less of the electric connective- tissue segments in his figures of the adult organ. Numerous other anatomical features of Gymnarchus have caused it to be classed with the other mormyrid fishes. This fact makes it of interest to compare its electroplax with the very different electroplax in these fishes. That found in Mormyrus oxyrhynchus will serve as a type and its general plan has been well shown by Ogneff (25) and Schlichter (30). Here it is evident that a number of consecutive and entire myotomes have been converted into electroplaxes and that the middle layer of each electroplax is composed of unaltered and clearly striated myofibril bundles. The large number of these fibril bundles, and their distribution, indicate that the whole electroplax in Mormyrus is a syncytium composed of all or most of the cells which would otherwise have gone to make up the single myotome. In this we find an agreement with the electroplax of Gymnarchus which is also formed from several cells. In the one case, however, all the cells in the myotome have been used (Mormyrus); in the latter only those lying in eight particular localities (Gymnarchus). (See paper by Dahlgren (36).) 190 Papers from the Marine Biological Laboratory at Tortugas. Further homology is seen in the disposition of the probably superfluous myofibrils. In both forms they are relegated to a middle position in the electroplax, while the apparently more important cytoplasm forms layers on the anterior and posterior surfaces of the structure. Also, in both, the now useless myofibril bundles are packed out of the way at right angles to the axis of the electroplax, which remains the same as the former axis of the muscle-cells that were used to form it. The only difference lies in the fact that the striation of the fibrils is retained in the Mormyrus forms, while it is lost in Gymnarchus, the dark- staining anisotropic substance apparently dissolving away. From what little can be predicted concerning the possible origin of the electric tissue in the other teleost forms, it is probable that the Mormyridze (including Gymnarchus) are the only fish in which the electroplax is formed as a syncytium from more than one cell. In Astroscopus, Electrophorus (formerly Gymnotus), and Malopterurus the structures show every evidence of having been developed from single myoblasts with the exception of Malopterurus, where it is a question as to whether they are not evolved from gland-cells instead. SUMMARY. In conclusion it may be stated that— (1) The electric tissues of Gymnarchus are developed by the differ- entiation of certain portions of its normal, striated muscle-tissues during an embryonic or larval period extending from the ninth day to the forty- second day of embryonic life. The critical period of this change takes place in the neighborhood of the eleventh, twelfth, to fifteenth days. (2) The muscle fibers which go to form the electric tissue give up their usual segmentation into myotomes, and first form eight long and con- tinuous spindles which afterward segment, each into a lesser number of masses, the electroplaxes. (3) The myofibrils lose their transverse striation and form a large inner core for each electroplax. They lie in a wavy mass, mostly at right angles to their former course. The active cytoplasm forms an outer layer which is denser and stains deeper on the posterior than on the anterior surfaces. It contains granules. (4) Each electroplax is made up of several muscle cells, 12 to 20 or more. This is different from the two elasmobranch fishes, in which each muscle cell forms only one electroplax. (5) The nerve is distributed on the posterior surface and ends in blunt knobs that lie in cavities formed by the invagination of the surface of the electric layer. The current probably runs from tail toward head, as in Mormyrus. (6) The development of the tissue gives a strong clew to the probable development of the electric tissues in the other mormyrid fishes. _ \O On! 10. Il. 12. 23: 14. TS 16. 17. 18. 19. 20. 2I. 22. 23; 24. 25. 26. 27; 28. 29. 30. St. 32; 33: 34. 35- 36. 37: A VE HN LITERATURE. . ANDREWS, Mrs. G. F. The living substance as such and as organism. Journ. of Morph., vol. x11, 2, Supplement. . ASSHETON, RICHARD. (See No. 24.) . BaLLowitz, E. Zur Anatomie des Zitteraales mit besonderer Beriicksichtigung seiner elektrischen Organe. Arch. f. Mik. Anat., Bd. 50, 1897. Ueber polytome Nervenfaserteilung. Anat. Anz., Bd. xvi, s. 541, 1899. —. Ueber den feineren Bau des elektrischen Organs des gewohnlichen Rochen, Raja clavaita. Anat., Hefte, Merkel und Bonnet, 1897. Das elektrische Organ des afrikanischen Zitterwelses, Malopterurus electricus Lac. Anatomische Untersuchs., Fischer, Jena, 1899. . BIEDERMAN, W. Elektrophysiologie. G. Fischer, Jena, 1895. . BupGett, J.S. (See No. 24.) . BUTscHELI, O. Untersuchungen iiber mikroskopischen Schaum und das Protoplasmas. London, 1894. DAHLGREN, U. The giant ganglion cells in the spinal cord of the order Heterasomata. Anat. Anz., Bd. x11, pp. 281-293. —. A new type of electric organ in an American teleost fish, Astroscopus (Bre- voort). Science, vol. 23, p. 469. , AND F, W. SILVEsTER. The electric organ of the ‘‘Stargazer,” Astroscopus (Brevoort). Anat. Anz., Bd. 29, p. 387. ENGLEMANN, TH. W. Die Blatterschicht der elektrischen Organe von Raja in ihren genetischen Beziehungen zur quergestreiften Muskelsubstanz. Pflug. Arch., Bd. 57, p. 149, 1894. Erpi, M.. P. Ueber Gymnarchus niloticus. Bull. d. Koénigl. Bayer. Akad. d. Wiss. (Gelehrte Anzeiger), No. 69. Munchen, Oct., 1846. Ueber eine neue Form elektrischen Apparates bei (Gymnarchus niloticus). Idem., Nr. 13, April 13, 1847. Ewart, J.C. The electric organ of the skate. Phil. Trans. Roy. Soc. of London, vol. 179, P- 399, 1889. f F The electrical organ of the skate: Observations on the structure, relations, progressive development, and growth of the electric organ of the skate. Phil. Trans. Roy. Soc. of London, vol. 183, pp. 398, pl. 26-30, 1893. Fritscu, G. Weiterer Beitrage zur Kenntnis der schwach elektrischen Fische. Sitz.- ber. d. Preuss. Akad. d. Wiss. Juni bis Dez., 1891. —. Zur Organization des Gymnarchus niloticus. Sitz. d. k6nig. preuss. Akad. d. Wiss. zu Berlin, Philos.-Math. Class. von 5 Feb., 1885. Die elektrischen Fische. Leipzig, Veit & Co., 1890. Ueber einige bemerkswerthe Elemente des Centralnervensystems von Lophius piscatorius. Arch. f. Mik. Anat., Bd. 27, s. 13-31. GARTEN, S. Die Production von Elektricitat. In Winterstein’s Handbuch der ver- gleichenden Physiologie. 8 bis tote Lieferung. Jena, Fischer, 1910. IwanzorF, N. Das Schwanzorgan von Raja. Bull. de la Société impériale des Nat- uralistes de Moscou, p. 53, 1895. Kerr, J. GRaAwAM. ‘‘The work of John Samuel Budgett.”” Cambridge Univ. Press, 1907. (Articles by Richard Assheton, Budgett and Arthur Shipley.) OcneEFF, J. Einige Bemerkungen iiber den Bau dess chwachen elektrischen Organes bei den Mormyriden. Zeits. f. Wiss. Zool. Bd. 64, p. 565, 1898. Retzius, G. Ueber die Endigung der Nerven in dem elektrischen Organen von Raja clavata. Biologisches Untersuchungen, N. F., Bd. vu, 1898. RHUMBLER, L. Versuch zu einer mech. Erklarung der indirecten Kern und Zell- teilung. Arch. Entwick. Mechanik, Bd. m1. RUpPEL, Ed. (Reported by Fritsch. See. No. 19.) Rouon, V. Zur Histogenesis des Riickenmarks der Forelle. Sitz. d. Math. Phys. Klass ed. Akad. d. Wiss. in Miinchen, Bd. 14, 1884. SCHLICHTER, HEINRICH. Ueber den feineren Bau des schwach elektrischen Organs von Mormyrus oxyrhynchus. Zeit. f. Wiss. Zool., Bd. 84, 1906. SCHNEIDER, C. Lehrbuch der Histologie. G. Fischer, Jena, 1904. STRASSBURGER, E. Ueber Cytoplasmastrukturen, Kern, und Zellteilung. Jahr. Wiss. Bot., Bd. xxx. StupNiTzKA, F. K. Ein Beitrag z. vergl. Histologie und Histogenesis des Riickenmarks. Sitz. d. Kgl. Bohm. Ges. d. Wiss. in Prag, 1895. Witson, E. B. Protoplasmic structure in the eggs of echinoderms and some other animals. Journ. Morph., Boston, vol. 15, supp., 1900. Harpy, W.B. Onthestructure of cell protoplasm. Journ. of Physiol., vol. xxIv, 1899. DAHLGREN, U. Origin of the electricity tissues of fishes. American Naturalist, vol. 44, pp. 195-202. — is $ ’ ' : BEARD, J. The transient Ganglion Cells and their Nerves in Raja batis. Anatomischer Anzeiger, Bd. 7, 1899. I9I RIG: <2. RiGs 2. FIGs 3- FIG. 4. Fic. 5. EGO: Fig. 7: BGs 3s. Fic. 9. FIG. 10. Eres pie FIG. 12. EXPLANATIONS OF THE PLATES. PLATE I. Transverse section through body of 9-day-old embryo of Gymnarchus in region C. L.M, position of the large muscle cells that will eventually change into the lower median electric spindle; V.F, location of large vertebral fibers; XXXX, locations of electric spindles in an adult fish with reference to bony structures, blood-vessels, and spinal cord; WN, canalis centralis; Bi, and Bs, caudal blood-vessels. X IIO. Muscle-tissue in transverse section from 12-day-old embryo of Gymnarchus to show typical muscle cells. A, one of oldest fibers with two strong fibril bundles; B, growing muscle cell with fibrils being laid down ina ring; C, extreme upper edge of myotome with youngest cells at top. XX II50. Transverse section through group of larger muscle cells found at Z.M. in fig. 1. Under greater magnification. For explanations see text. X I150. Longitudinal vertical section of C region in 9-day-old embryo, ventral portion. Between marks O O are seen same parts of seven myotomes as are shown at L.M. in preceding figure in transverse section. Note heavy development of these cells. A line drawn between marks @ @ would indicate location of seven groups of vertebral fibers in next few mediad sections. PLATE 2. Transverse section of body of 12-day-old Gymnarchus embryo in region CC). D, dorsal electric spindle; U.M, upper median spindle; Z.M, lower median spindle; V, ventral spindle; Conn.T, connective-tissue septum; JV, canalis centralis. B,; and Bs, caudal blood-vessels. X II0. Three degenerating muscle cells from another (3) zone in fig. 12. XX 1440. Electric spindle showing first stage where electric muscle cells can be called an electroplax. Complete coalesence has taken place. Nuclei are in peripheral layer. Myofibrils are concentrated in middle to form central fibrillar core of structure. XX I150. Transverse section of upper dorsal part of spinal cord in a Gymnarchus, embryo 9 days old. Neuroblasts, in region that will later show electric motor-cells, show some little differentiation and growth. Near, and above, this region is one of the transitory giant ganglion cells. X 940. PLATE 3. Transverse section of left ventral spindle from region C in embryo of 12 days. To left of dotted line are normal muscle cells, to right are degenerating muscle cells, inthe midst of which, and surrounded by dotted circle, are a group of muscle cells going through the first steps of transformation into electric tissue. X I150. Transverse section of ventral electric spindle in process of formation in region B, of 12-day-old embryo of Gymnarchus. At left of figure two muscle cells begin to degenerate, while just to right of upper one of this pair a muscle cell is seen in advanced stage of atrophy. Connective-tissue capsule beginning toform. Electric muscle cells rather more coalesced than infig.9. X I150. Transverse section of upper middle electric spindle in same section as fig. 10. Rapidly degenerating muscle cells to left. Further coalescence of electric muscle cells than in fig. 10. Myofibrils weakening but still in separate groups. XX II50. PLATE 4. Longitudinal vertical section of dorsal part of region D of 12-day-old embryo of Gymnarchus, showing relations of muscle tissue and electric tissue. Bracket I indicates zone of degenerating muscle fibers on inner median edge of five myotomes, on ventral side of median connective tissue; zone 3 shows atrophy- ing muscle fibers on inner median edge of myotomes, dorsad of median con- nective tissue; zone 4, rudiment of upper median electric spindle; zone 5, same of dorsal electric spindle; zone 6, muscle tissue from ventral spindle to dorsal outer edge; zone 7, connective tissue; zone 8, dorsal epithelium. X 140. 192 FIG. 13. Fic. 14. FIG. 15. FIG. 16. FIG. 17. Fic. 18. FIG. 19. FIG. 20. FIG. 21. Fig. 22; FIG. 23. 13 Origin of Electric Tissues of Gymnarchus Niloticus. 193 Longitudinal vertical section of ventral part of region C of 12-day-old embryo of Gymnarchus. Zone 1 indicates ventral epithelium; zone 2, ventral outer portions of five myotomes; zone 3, lower median electric spindle which is dividing into three parts in five vertebral segments; zone 4, inner degener- ating muscle fibers of five myotomes; zone 5, projected outline of vertebral segments. XX 140. Motor nerve cells from spinal cord of 42-day-old larva of Gymnarchus. A, B, and C, three electric motor nerve cells. D and E, two muscle motor nerve cells. From region ZH. XX 1500. PLATE 5. More highly magnified part of fig. 12. Brackets indicate same zones. XX 760. Highly magnified part of transverse section through an electroplax from C region of 42-day-old larva of Gymnarchus. Curving bundles of modified myofibrils are seen in core and cut at various angles. Outer or electric layer is fixed and stained to show finer outer granules or electrochondria. Vacuoles as well as larger meshes of the linom are visible on inner edge of electric layer. Elec- tric nerves are cut in sections at many angles and show medullary or connec- tive-tissue sheath and its nuclei. At two points branching of these nerve fibers is visible and at four points its contact with electric layer of electroplax is to be seen. XX 1500. PLATE 6. Longitudinal section of a young electroplax from B region of 12-day-old embryo. Segmentation of electric tissue is now complete into unit electroplaxes of which figure is good example. Central fibril bundle beginning to lose its transverse striation. Structure is as long as three muscle segments or bony segments, as near-by spinal processes show. This represents oldest stage in embryo of this age. XX 210. Longitudinal section of young electroplax from C region of 42-day-old larva of Gymnarchus. Fibrils have lost all striation and become waved in middle part. Cytoplasm shows strong differentiation at two poles (ends). Anterior end strongly vacuolized. Papille begin to grow from ends of middle part and nerve-endings have become established at junction of posterior third with middle third. This shows youngest stage in 42-day-old larva and is a successively later step in differentiation than fig. 17. XX 210. Longitudinal section of an electroplax from E region of 42-day-old larva of Gym- narchus. Fibral bundle much ‘‘waved”’ in middle third, which now begins to show form of electroplax. Papille more developed. Nucleus and some cytoplasm inclosed in fibrillar core. XX 210. PLATE 7. Longitudinal section of another electroplax from further caudad in £ region of same larva of Gymnarchus. This represents oldest stage accessible and is regarded as almost adult in its histological characteristics. Two nuclei shown in fibrillar core; one of these shows characteristic swelling. PLATE 8. Transverse section of body of 42-day-old embryo of Gymnarchus from D region. Greater part of muscle tissue is here transformed into electric tissue. Section passes through very last part of dorsal fin. Electroplaxes cut at different levels and in one case section passes through right ventral spindle showing only electric connective tissue. D, dorsal spindles; U.M, upper middle spindles; Z.M, lower middle spindles; V, ventral spindles; Bi; and By, caudal blood-vessels; JV, rele canal. Ni, Ne, Nz, and N4, electric nerves; NV;, lateral line nerve. X II Part of transverse section from middle part of an electroplax in E region of same larva of Gymnarchus. This figure serves well to show larger granules that lie deeper in electric layer, also connective-tissue sheath or wall of electric spindle. X 1500. PLATE 9. Longitudinal section from E region, showing parts or whole of 11 electroplaxes lying in dorsal, upper middle, and lower middle spindles. Some remaining muscle fibers are visible at sides. Figure serves to show that there is no heavier, transverse connective-tissue wall lying in electric connective tissue, as is found in Mormyrus. This emphasizes secondary segmentation found in Gymnarchus. X 110. 194 Papers from the Marine Biological Laboratory at Tortugas. Fic. 24. Small part from edge of basal part of a papilla on posterior surface of an electroplax from C region of 42-day-old Gymnarchus. Shows a single nerve, ending ina somewhat elongate and branched end-plate in electric layer. Outer crowded mass of granules stained deeply with eosin; larger inner granules not stained. 1500. Fic. 25. Small part of edge of another part of anterior surface of same electroplax. Shows a nerve fiber dividing and each branch ending in typical club-shaped nerve ending of Gymnarchus. X 2200. DAHLGREN PLATE 1 ee ee oe es ee eee eer - : SONA rh ie ‘ | ; | 3 4 R - > - m > is - 3 { 4 ee ie : , ri rs - ae ; e > if } ic. » + oa = . = = Tt. ‘sth 4 © “ ha ee Se | i * it = ee port i? oe “ ; ; n ve ek * . F Bh ey 4 ‘ * t : : ’ : host - hy Sa 2 < — > sigh an= {¥ ee a lal ; bit Se fT SES OE ie? als 2 Pogheeniiseie oer = ’ i i e 2 i ‘ i ee ws tee = mom Sacnntd Sy Me RN eee Ma See ree mem ip 6“ - = ry ‘ ard py see Sos Lk cade a a en ee oe SE AA Oe r DAHLGREN PLATE 2 ConT LM. a i? fj ; Ps ae , ‘ DAHLGREN PLATE 3 * Ay ae aah, Cem” 6 ~~ , Re) | th, Sie & @ wp Bag “it net / 6 ae ; ? 0,:#0 S & ’ eT oe i @ P @ x r ‘ fe ¢ i. 4 t yer : ng ‘ PLATE 4 DAHLGREN fa et sates Sc aa 14 DAHLGREN PLATE 5 see e esses ey patie oS ee a il 6 says PLATE 6 - ~~” @ GS Soe Moo SS a eee / “~~ © e®°, rh 2o S&S ef og ® © eee = é Th a pages a eee a3 : Sees aa 9 rs) ee & ee ie Sse ap S29 2 Oe Se°° ' Dee D6 — ark iat ae i ~ ‘ - oe eo oO : Q .e@ e832. / z/ ‘ / e é — DAHLGREN —> ae ‘ PLE te meee ett ime ies Leite ay eee bg Sey ety . DAHLGREN PLATE: 7 { ' j i 1 k nea ath! he se ake A DAHLGREN PLATE 8 22 DAHLGREN 24 i) 3; 4. NET) e949 iyi PLATE 9 XI. THE DEVELOPMENT OF THE APYRENE SPERMATOZOA OF STROMBUS BITUBERCULATUS. BY EDWIN: E.REINKE: Proctor Fellow in Biology, Princeton University. Seven plates, 195 Ae CONTENTS, ee AM AOE ISILCRIZCRD 2 ict WOO Resi payin sree 12's 48° Slain ee Me etek die canter ote obese eater Observations upon the living ‘cells of the testis...) 2 .e'45 saul oat = aR Growth of the apyrene spermatoblast 1. .5tuedee cue. vais’ cee eee ee Growth and differentiation of the apyrene spermatosome PIGEHSSIOD fs 5 o's salt td ere cee wee tele oa don BRI Ratt ote OE eh CR as ee Re NS. 196 THE DEVELOPMENT OF THE APYRENE SPERMATOZOA OF STROMBUS BITUBERCULATUS. By Epwin E. REINKE. INTRODUCTION. The fact that there is a dimorphism in the male sex cells of many Proso- brachs is well known at the present day. This is due to the great number of investigations that have been directed to the subject, but, in spite of them all, we are still as much in the dark as regards the function of the atypical spermatozoa as were the earlier investigators who did their work more than half a century ago. Furthermore, while the list of Prosobranchs in which this dimorphism obtains has been extended to a great many forms, in only a few of them do we have careful descriptions of the structure of ‘ the atypical spermatozoa, and in only one case (Paludina) is there a complete account of their development. Obviously the solution of the problem is to be sought in a study of as many forms as possible, for I believe that each form has not only its own peculiarity in structure, be it great or small, but also its peculiarity in development. By peculiarity I mean not only the differences in form and size and in development between the typical and the atypical spermatozoa of any one genus, but also the differences between the atypical spermatozoa of different genera. In a comparative study of these peculiarities and differences may be found the explanation of the existence of the dimorphic spermatozoa. Comprehensive work such as this has already been begun; Kuschakewitsch (’11) has reported briefly upon the first two of a series of forms under investigation by him, and the writer has recently done the same for Strombus. My attention was first called to the question of the dimorphic sper- matozoa in Prosobranchs by Professor E. G. Conklin. At his suggestion Urosalpinx cinerea was made the main object of the investigation, supple- mented by comparisons with the fresh-water snails Gontobasts virginica and Paludina vivipara (the Japanese variety). Material was collected and considerable work was done on these forms. But in May 1g11, while at the temporary laboratory of the Carnegie Institution of Washington established at Port Royal, Jamaica, I was afforded the opportunity of studying and collecting material of various Prosobranchs, including Strombus bituberculatus. This form proved so interesting and so favorable for research that my immediate attention was centered upon it. My work was further 197 198 Papers from the Marine Biological Laboratory at Tortugas. advanced by another visit, in March 1912, to the temporary laboratory of the Carnegie Institution, this time at Montego Bay, Jamaica. Here I was able to carry on with partial success some experimental work and also to verify in the living cells most of the stages in the development of the apyrene spermatozoa that I had traced in preserved material. My thanks are due to the Carnegie Institution for the privileges extended to me on both occasions, and in particular to Dr. A. G. Mayer, the Director of the Department of Marine Biology. Dr. H. A. Pilsbury of the Academy of Natural Sciences of Philadelphia very kindly identified the species for me. I am under great obligation to Dr. L. R. Cary, of the Department of Biology, Princeton University, for turning over to me his sectioned material of the testis and oviduct of Goniobasis. Finally, I wish to express a very sincere gratitude to Professor Conklin, not only for his many invaluable suggestions and criticisms, but also for the keen interest which he has at all times taken in the work. HISTORICAL. In 1837 v. Siebold discovered the fact that there are two kinds of sper- matozoa in the testis of Paludina vivipara. ‘The typical ones he called hair- shaped and the others worm-shaped spermatozoa. Since then the question of this dimorphism has been investigated by several writers nearly all of whom have given their attention mainly to Paludina. Paasch (’43) and Kolliker (’41 and ’47) denied the existence of the so-called worm-shaped spermatozoa of v. Siebold, the former interpreting them as being bundles of the hair-shaped spermatozoa and the latter claiming them to be different stages in the development of a single kind of spermatozoon. Leydig (’50), however, not only agreed with v. Siebold as to the existence of a dimorphism in the spermatozoa of Paludina, but also corroborated his description of the development of the worm-shaped kind. His views were very generally accepted until Duval (’80) investigated the question. He described an independent origin for the two kinds of sper- matozoa, but in other respects his work was shown by v. Brunn (’84) to be quite inaccurate. Although entirely confused in the successions of cell-generations, v. Brunn made a nearer approach to the facts in the case than anyone had done previously. He also described the dimorphic spermatozoa in five genera besides Paludina and later in the same year extended the entire list to include the following forms: Paludina vivipara, Ampullaria (sp.?), Murex brandaris, M. erinaceus, M. trunculus, Cerithium vulgatum, Nassa mutabilis, Fusus syracusanus, Marsenia (sp.?), Aporrhats pes pelicant, Cassidaria echin- ophora, Dolium galea, Tritonium cutaceum, T. parthenopeum, Vermetus gigas. Brock (’87) added to this list by describing the atypical spermatozoa in three exotic Prosobranchs, Péteroceras lambis, Strombus lentiginosus, Cyprea annulus, C. caput serpentis, and C.lurida. As regards Strombus, his descrip- tions and observations are quite accurate as far as they go, but his figures are lacking in detail. Papers from the Marine Biological Laboratory at Tortugas. 199 There then followed two papers, one by Kohler (’89) on Murex, and the other by Auerbach (’96) on Paludina. These authors established the fact that there is an almost complete loss of chromatin during the develop- ment of the worm-shaped spermatozoa, but they too were unable suc- cessfully to trace the succession of cell-generations and thus they made little or no material advance over the work of v. Brunn. To these obser- vations Erlanger (’97) added a minute description of the adult worm- shaped spermatozoon in Paludina. It is unnecessary here to go into a detailed discussion of the literature mentioned above, as it has all been discussed in the work of Meves (’03). This author has given a detailed and correct description of all the various stages in the development of both kinds of spermatozoa in Paludina. Furthermore, he has given careful consideration to the findings of the more important of the earlier writers and has shown wherein they were in error. Meves describes the earliest differentiation between the two kinds of spermatozoa as occurring during the growth period of the spermatocytes, the spermatogonia of both being alike. He denies the statement of v. Brunn that the basal nuclei reproduce by direct division. This seems to the writer to be a debatable ground and forms the only weak point in the work of Meves. v. Brunn, of course, was quite wrong in supposing that the cells formed by this direct division of the basal nuclei give rise eventually to both kinds of spermatozoa; he failed to recognize the sper- matogonia as distinct primitive elements of the testis. According to Meves these cells undergo unequal growth and thus give rise to both the hair-shaped and the worm-shaped spermatozoa. At the suggestion of Waldeyer he uses the terms eupyrene and oligopyrene to describe the two kinds of spermatozoa. The words have reference to their adult condition, 1. e., those provided with the ordinary amount of nuclear material and those with but a little; similarly, apyrene spermatozoa are those in which the entire nucleus disappears. According to Meves, the oligopyrene spermatocytes of Paludina grow very much larger than the eupyrene and are clearly distinguishable, not only on account of their increased size, but also on account of their peculiar pattern and the presence of a large centriole and sphere. Preparatory to cell-division, this centriole divides and a little later both halves fragment into many smaller centrioles. The division which now occurs is abnormal. Two groups of very small centrioles form the poles of an ill-defined spindle. The daughter chromosomes pass to both poles, but before the division of the cell-body is completed the majority of them scatter again throughout the cytoplasm; four of the chromosomes, however, remain at the poles to form the new nuclei. These become vesiculated and may all unite with one another or some may unite and others remain single or all may remain single. Thus there may be four, three, two or only one nuclear vesicle in the secondary spermatocyte. The remaining ten chromosomes undergo a peculiar vesiculation in that the chromatin becomes lumped as a crescent on one side of the vesicle. 200 Papers from the Marine Biological Laboratory at Tortugas. The second division follows immediately and is even more abnormal than the first. The centrioles scatter to the periphery and from each there pass out radiations. With the dissolution of the nuclear membrane or membranes, as the case may be, four new chromosomes make their appearance; these all show a longitudinal split. Very soon, radiations from a centriole are attached to each of the new chromosomes and they are drawn to various points near the periphery of the cell. By this time the ten vesiculated chromosomes remaining from the first division have been inclosed in a sort of capsule and are thus prevented from participating in the ensuing division. This is introduced by a slight elongation of the cell. Next, the centrioles move to the two ends of the cell and in so doing they form a-spindle; one daughter chromosome then passes to each pole, while the other chromosomes remain scattered in the cytoplasm. With the division of the cell-body, the vesiculated chromosomes (inclosed in their capsule) pass into one or the other of the daughter cells, where they are set free by the dissolution of the capsule. As a result of this division, Meves holds that the new nucleus is formed out of only a single chromosome, which very quickly becomes vesiculated and increases in size; all of the other chromosomes are gradually dissolved in the cytoplasm. Before the process of division is completed, cilia grow out from the centrioles, forming a brush at each pole of the dividing cell. After the cell has divided, the centrioles divide and their distal halves begin to move across the cell. A connection in the form of a fiber is always maintained between the proximal and the distal halves of the centrioles, the former remaining attached to the cell-wall. The nucleus in the mean- while decreases in size and its chromatin forms a layer under the nuclear membrane. It moves closer to the plate of dividing centrioles. The con- tinued movement of the distal centrioles across the cell results in pushing the nucleus against the cell-wall and in forming the bundle of axial fibers between the proximal and distal centrioles. Further growth of the axial fibers is accompanied by a gradual lengthening of the cell until the typical worm-shaped condition is reached. At the very end of the spermatozoon the nucleus forms a thimble-like cap over the end of the bundle of axial fibers. In formulating an hypothesis to explain the problematical function of the oligopyrene spermatozoa, Meves discusses the theories of v. Brunn and Brock. v. Brunn believed them to be rudimentary eggs and therefore functionless, and this for two reasons: first, after careful investigation he failed to find them either fertilizing the ovum or even present in the oviduct; second, he believed their development bore resemblances to that of the ova. These views Brock attempted to refute on the grounds that any such rudi- mentary structure must have a proved homology with some still functioning part or organ and its departure in structure and development must be shown to be due to retrogressions; furthermore, if the dimorphism of spermatozoa is to point to an ancestral hermaphroditism, as v. Brunn Papers from the Marine Biological Laboratory at Tortugas. 201 claimed, the lower Prosobranchs should show this peculiarity, which they do not. He comes to the conclusion, then, that they are functionless. Now Meves has shown that the oligopyrene spermatozoa of Paludina in their development show a marked parallelism—not to the ova, but to the true spermatozoa, and that they are produced in numbers equal to the latter. He suggests that there may be certain conditions under which they do fertilize the eggs; granting that they do and since the nucleus is accepted to be the bearer of heredity, then an embryo resulting from an egg fertilized by an oligopyrene or apyrene spermatozoon would show pre- ponderating female characteristics. Such a case would prove the hypothesis that the nucleus is the bearer of heredity. The spermatozoon would then have, either primarily or solely, the mere function of stimulating the egg to division. According to Boveri, this stimulus is due to the introduction of a centrosome into the egg. But if it should be shown that actually the oli- gopyrene spermatozoa never fertilize the egg, then Meves would take the view, with Brock, that they are functionless. Stephan (’03a) investigated the apyrene spermatozoa of Cerithium vul- gatum, Murex trunculus, M. brandaris, Triton nodifer, and Nassa mutabilis. All of these forms he reported to have apyrene spermatozoa whose devel- opment is comparable to that of the oligopyrene of Paludina. All of the chromatin disappears in the atypical spermatozoa of these forms and in every case except Cerithium the bundle of axial fibers never comes into contact with an existing nucleus, as is the case in Paludina. In Cerithium the advancing bundle of axial fibers does touch a small nucleus which is the equivalent of a chromosome. Before the fibers have completed their growth, however, this small nucleus degenerates and disappears completely. Again, in Cerithium, as in Paludina, the cilia persist, while in the other forms they are retracted before the spermatozoon has fully matured. Upon these facts and also upon the positions of the anterior centrosomal structure in the adult atypical spermatozoa of the different forms under investigation, he concludes that gradations may be established between the different types of atypical spermatozoa, with Cerithium occupying an intermediary position between Paludina and the others. In another paper Stephan (’030) described briefly the development of the apyrene spermatozoa of Murex brandaris. The development here agrees in general with the description given by Meves for the oligopyrene sper- matozoa of Paludina. It differs in the facts that there is an intense vacu- olization of the cytoplasm and that there is no nuclear element in the adult spermatozoon.! Basing his conclusions on certain observations of Popoff which were published later, R. Hertwig (’05) expressed his belief that the oligopyrene and apyrene spermatozoa serve to fertilize the egg, but such fertilization would not be a uniting of the male and female nuclear components, and would stand on the boundary between parthenogenesis and true fertilization. 1A third paper by Stephan (’03c) onthe development of the apyrene spermatozoa of Cerithium vulgatum and Nassa mutabilis, with two figures, was inaccessible to me. 202 Papers from the Marine Biological Laboratory at Tortugas. This fertilization would result, according to Hertwig, in the production of males: ‘‘The sex-determining capacity which is indicated in the female by the formation of large eggs that are rich in yolk and small rudimentary ones, in this case would be transferred to the male sex.”’ Retzius (’06) described and figured the atypical spermatozoa of A porrhais pes pelicani, Turritella terebra, Murex trunculus, Fusus despectus, and Buc- cinum undatum. Popoff (’07) made a series of very careful and important observations upon Paludina. He examined the seminal receptacles of a great number of females at all times during the period of their sexual activity (May to November). He found that any one of the following conditions may obtain: (1) The receptacle may contain normal eupyrene and oligopyrene spermatozoa; (2) it may contain normal eupyrene and degenerating oli- gopyrene spermatozoa; (3) it may contain only eupyrene spermatozoa; (4) it may be completely empty. Of the 476 females examined during the whole period, in 176 cases the receptacle contained normal eupyrene and oligopyrene spermatozoa; the proportion of the cases in this condition to the total number examined at any one time showed a decided fall after the period of greatest sexual activity was reached (July). The condition which obtained in the next greatest number of cases (152) was the fourth, in which the receptacle was empty, and here the percentages of cases increased as the season advanced. In 89 cases the receptacle was found to contain only eupyrene spermatozoa, 7. e., the third condition; the percentage of these cases fell during August and then held fairly constant. Finally, only 39 individuals were found to be in the second condition (normal eupyrene and degenerating oligopyrene spermatozoa in the receptacle); here the percentage of cases gradually increased until August and then fell in Sep- tember. These facts indicated that the oligopyrene spermatozoa are shorter-lived than the eupyrene. In order to ascertain this definitely, Popoff segregated a number of females from the males immediately after copulation. From time to time the seminal receptacles of several of these females were ex- amined and as a result he was able to establish the fact that after Io to 12 days the oligopyrene spermatozoa begin to degenerate, while this does not occur among the eupyrene until after about 25 days. In addition to this, Popoff found both kinds of spermatozoa in equal numbers in sections of the oviduct; only in the case of animals whose receptacles contained no oligopyrene spermatozoa did he fail to find these forms in the oviduct. This would indicate that the function of the oligo- pyrene spermatozoa is to participate in the fertilization of the egg. His results, however, in attempting to obtain eggs at the time of fertilization were practically negative. In three eggs, which constituted the only clear cases, Popoff found that the spermatozoa which had entered were eupyrene, so that nothing could be stated with certainty concerning the function of the oligopyrene spermatozoa. He thinks, however, if they may be said to have Papers from the Marine Biological Laboratory at Tortugas. 203 any function with regard to the egg (as is indicated by their presence in the oviduct), that this function can only be of a sex-determining character. Lams (09)! reported briefly that the development of the atypical sper- matozoa in Murex confirms the description given by Meves for Paludina. In sectioned eggs he was able to find only eupyrene spermatozoa and there- fore came to the conclusion that they alone serve to fertilize the egg. The experiments of Kuschakewitsch (’10) have led to some very inter- esting results. By obtaining the ripe eggs of Aporrhais pes pelicant from the oviduct and mixing them with a mass of both kinds of spermatozoa which had been diluted with sea-water, he was able to observe the matura- tion and the first two cleavages of from 10 to 40 per cent of the eggs thus treated. In many instances he could sce the formation of an entrance-cone and the entrance of the spermatozoon itself into the egg, but in no case was he able to observe the same processes for the atypical spermatozoa. But in sections of eggs which had been killed 20 minutes after mixing with spermatozoa, he found that the apyrene spermatozoa can and do enter the eggs, either alone or in addition to the eupyrene spermatozoa. Here the apyrene forms undergo a series of degenerative changes and are finally extruded. In later stages, during and after the first maturation spindle, their presence in the egg has never been detected. Here it is seen, then, that the apyrene spermatozoa were extruded before any nuclear changes had taken place in the egg. Again, in some Proso- branchs the apyrene spermatozoa are very large and totally immotile, so that one can hardly imagine that they can in any way get into the egg. In view of these facts Kuschakewitsch has come to the conclusion that any sex- determining function in the apyrene spermatozoa is to be doubted. Finally, Kuschakewitsch (’11) has briefly described the development of the eupyrene and the apyrene spermatozoa in Conus mediterraneus and of the apyrene spermatozoa in Vermetus gigas. In both cases the development of the apyrene spermatozoa shows many differences from that of the oligo- pyrene of Paludina. All three forms, however, have in common the greatly increased growth of the spermatocytes as the starting-point of the differentiation between the typical and the atypical spermatozoa. Kuschakewitsch finds that in Conus no divisions occur in the develop- ment of the apyrene spermatozoa and that the spermatocyte is transformed into the spermatozoon by the disappearance of all the nuclear material and the subsequent growth of two fibers and two flagella from the centrioles. The nucleus may either gradually dissolve as a whole or it may fragment. The process of fragmentation may take place in either of two ways: the nucleus may become vesiculated and then break down into two and later into several compact and vacuolated spheres, or it may become compact and break down into a number of dense chromatic bodies which probably represent chromosomes. In either case the end result is the disappearance of all the chromatin. The two fibers eventually come to lie at the surface 11 have been unable to obtain this publication and have therefore relied upon the review of it given by Kuschakewitsch. 204 Papers from the Marine Biological Laboratory at Tortugas. of the cell and the flagella with the pair of centrioles belonging to them at one of the ends of the spermatozoon. The conduct of the mitochondria is very striking; they appear first as granules, but eventually form fibers. During the course of development one or more vacuoles appear in the cell, which are quite characteristic of the adult spermatozoon. In Vermetus, too, he finds that there is a direct breaking down of the nucleus of the spermatocyte. The chromatin forms into karyomerites and upon the dissolution of the nuclear membrane these lie free in the cytoplasm and are all gradually absorbed. In the spermatocyte there is to be seen a pair of centrioles lying in a thicker mass of cytoplasm; these form a group of secondary centrioles at one pole of the cell at the time of the dissolution of the nuclear membrane or immediately afterwards. The secondary centri- oles give rise to a number of threads which eventually form a bundle of axial fibers whose middle portion passes through the center of the cell and whose ends extend beyond it in either direction. In the cytoplasm of the cell chere arise a great number of thick-walled chambers whose cavities are filled with a clear substance. In each of these chambers a sphere of albumen is gradually formed. In the spheres themselves there appears either a large central vacuole or several smaller peripheral ones. Afterwards the contents of these vacuoles seem to become greatly condensed. In his conclusions Kuschakewitsch points out the fact that the develop- ment of the atypical spermatozoa in Conus and Vermetus does not present the same parallel to the development of the typical spermatozoa as Meves established in the case of Paludina. He thinks, however, that we are justi- fied in speaking of these structures as ‘‘spermatozoa”’ when the fact is borne in mind that their origin is the same as that of the oligopyrene spermatozoa of Paludina and Murex and that it is hardly conceivable that, in the well- defined group of Prosobranchs, structures with a similar origin would have a different morphological significance in different members of that group. Thus the mode of the development of the apyrene spermatozoa in Conus and Vermetus can in no wise be regarded as supporting the hypothesis of v. Brumm, that these structures are rudimentary ova. On the contrary, in Conus true eggs appear in the testis in addition to the apyrene spermatozoa.! As regards Strombus the writer has shown (Reinke ’12) that here, too, the development of the apyrene spermatozoa takes place without any maturation divisions, and furthermore that there are no processes, such as the formation of chromosomes and abortive spindles, which could be described as surviving from a condition in which those divisions did take place. In addition there is the fact (which will be shown in the succeeding pages) that the apyrene spermatozoa are not derived from the spermato- gonia. But in spite of these facts, there are good reasons for still applying the term ‘‘apyrene (oligopyrene) spermatozoa” to these structures. In the first place, the name is well established and has been used commonly, and 1Since the above was written there has appeared another paper by Kuschakewitsch giving a detailed account of the development of the eupyrene and apyrene spermatozoa in Conus and Vermetus. In it he states that the apyrene spermatocyte of Conus sometimes undergoes a division after the nucleus has disappeared. —e Papers from the Marine Biological Laboratory at Tortugas. 205 the words ‘‘apyrene”’ and ‘‘oligopyrene’’ are sufficiently descriptive to distinguish them from the true or ‘‘eupyrene”’ spermatozoa. Then again, they are spermatozoa in the sense that they are products of the testis and are found in the seminal fluid along with the true spermatozoa. In retaining these names the writer does not mean to imply any homology between the apyrene and the eupyrene spermatozoa. Indeed, on account of the absolute lack of homology the terms ‘‘apyrene spermatocyte’’ and “‘apyrene spermatid’’ must be abandoned. A spermatocyte is a cell which is derived directly, by a mere process of growth, from a spermatogonium, and a spermatid is one of the four cells resulting from the two maturation divisions which such a spermatocyte has undergone and which is then transformed into a spermatozoon. The cells which give rise to the apyrene spermatozoa of Strombus go through two distinct periods of development and at no time can they be homologized with either the eupyrene spermatocytes or spermatids. The first of these periods starts at the time when the cell is first recog- nizable in the testis.' This period continues throughout all the stages of the gradual but great growth of the cell and ends with the breaking down of the nucleus and the simultaneous scattering of the centrioles. The second period may be said to begin with the division of the centrioles; it is a period of marked cytoplasmic differentiations and it ends, of course, with the formation of the fully matured apyrene spermatozoon. For the cell during the first period it is proposed to use the name “‘apyrene spermato- blast’’ and during the second period to designate it as the ‘“‘apyrene sper- matosome.”’ Both of these words, ‘‘spermatoblast”’ and ‘‘spermatosome,”’ have been used hitherto in various connections; neither of them has the restricted meaning of ‘‘spermatogonium”’ or “‘spermatocyte’”’ and, as used here, they imply no homologies with either of those structures. METHODS. Several fixatives were used for preserving the testis and the uterus of Strombus, namely, Bouin’s, Flemming’s, Helly’s, and Zenker’s fluids. Of these, Flemming’s fluid (strong solution) gave by far the best results. Small pieces of the tissue were fixed for from 6 to 8 hours and were then washed for at least the same length of time, usually over night. Asa result of such fixation the osmic and acetic acids did not penetrate very strongly into the inner portions of the tissue, and here the preservation of the mitochondria, the spheres, and the fibers was all that could be desired. Sections were cut 5 micra thick and were stained with iron hematoxylin, followed by a weak solution of erythrosin in 70 per cent alcohol, which had 1It may be well to state at this point that the writer has succeeded in tracing the origin of these back to the so-called basal nuclei of Platner. He has found that the nurse-cells in the testis of Littorina have a similar origin, although the developmental changes are less complicated. It was deemed best, however, not to present the evidence of this origin at this time, but to wait until the history of the basal nuclei had been followed, in all its details, not only in these forms but also in other Prosobranchs, both those which have dimorphic spermatozoa and those which do not. The scope of the present paper, therefore, is confined to a description of the develop- ment of these cells from the time when their nuclei have been completely differentiated from the basal nuclei and have been established in the syncytium of the testis along with the true spermatogonia, each one surrounded by its own definite cytoplasmic body. 206 Papers from the Marine Biological Laboratory at Tortugas. been tinged to a purple color by the addition of a small quantity of a con- centrated solution of gentian violet. The gentian violet was added to provide a nuclear stain for such structures as had not retained the iron hzmatoxylin; it gave a differential stain to the spheres and thus proved to be of considerable value. This stain has been used subsequently by other . workers in this laboratory with fair success. Sections of the uterus and the seminal receptacle were stained in the same way. Smears from the testis, the vas deferens, and the seminal receptacle were prepared by making thin films of the contents of those organs diluted with sea-water. The films were allowed to dry over osmic vapor, were washed in fresh water, and then stained with Delafield’s or iron hematoxylin. Sea-water was found to be the best artificial medium for studying the living cells. The contents of the testis or sperm-duct were mixed with sea- water and then a drop of the mixture was placed on a slide and surrounded by a complete ring of glycerine-gelatine. In this way, a cover-glass placed upon the drop was not only supported, but it was also given enough rigidity to permit the use of an oil-immersion lense. If care were taken to see that the ring of glycerine-gelatine was unbroken, practically no evaporation of the mixture would take place. Under these conditions the various cells of the testis would remain alive for at least 2 hours. Sometimes methylene blue was used as an intra vitam stain, but this was found to be quite unneces- sary and to cause an earlier breakdown of the cells than would otherwise occur. OBSERVATIONS. ADULT SPERMATOZOA. The eupyrene spermatozoa of Strombus bituberculatus resemble in general those of other Prosobranchs. They are relatively short when compared with the spermatozoa of Fasciolaria, Littorina, or Neritina. The perfo- ratorium is short and conical, the head thick and spindle-shaped (fig. 7). The head and the middle-piece, which is very slender and passes almost imperceptibly into the tail, constitute about half the total length of the spermatozoon. In marked contrast to them stand the apyrene spermatozoa (fig. 6). The term ‘“‘worm-shaped”’ can in no wise be applied to these, as they are totally different from the oligopyrene spermatozoa of Paludina. They were first described by Brock (’87) in the case of S. lentiginosus and again more recently and in some detail by the present writer (’12) in the case of S. bituberculatus. While Brock’s brief description is correct as regards the general facts, his figure of the adult apyrene spermatozoon is not in accord with the facts as they have been observed by the present writer in S. bituberculatus, and there is no reason to suppose that any very great vari- ation occurs between the apyrene spermatozoa of the two species. In their structure, the apyrene spermatozoa of S. costatus are identical with those of .S. bituwberculatus, the only difference being in their respective lengths. Apparently Brock did not fully understand the structure of the undulating Papers from the Marine Biological Laboratory at Tortugas. 207 membranes and he failed to observe the very evident centrosomal crescent at one end of the spermatozoon. The adult apyrene spermatozoon of Strombus is composed of a central spindle-shaped body, which is long and narrow and slightly flattened dorso- ventrally, on the sides of which are attached two broad undulating mem- branes (fig. 6). At the anterior end of the cell-body there is a crescent- shaped plate of centrosomal origin; starting from the horns of this crescent the undulating membranes round out quickly to their maximum width, while posteriorly they gradually become narrower and finally end in a short, sharply pointed tail-piece. In the membranes, here and there, can be seen indications of the fibers of which it is composed, but both these and the centrosomal plate are brought out more distinctly in fixed and stained specimens. The interior of the cell is filled with a number of large secreted polygonal bodies composed of an albuminous substance or substances, possibly of a deutoplasmic nature like the yolk granules in the nurse-cells in the testis of Littorina (Reinke ’12)._ When treated with iodine these bodies fail to give the reaction for starch or glycogen. When a mass of apyrene spermatozoa are separated from the eupyrene and are subjected to the xanthoproteic test, there is a very strong and clear reaction indicating the presence of albumen. In sections stained with iron hematoxylin the secreted bodies show the characteristic “‘Spiegelfarbung”’ of yolk granules. These bodies are more or less regular in shape and position but they decrease in size toward either end of the cell. In the living spermatozoon, at its posterior end, there can be seen a small aggregation of mitochondria. No traces of nuclear material can be found except in fixed specimens; here they appear as occa- sional darkly staining granules of degenerating chromatin lying between the albuminous bodies. The orientation of the apyrene spermatozoon has been established upon morphological grounds. In Paludina, that end of the oligopyrene sper- matozoon which contains the remains of the nucleus, 7. e., the head, has been designated as anterior and this, too, is the end toward which the axial fibers have grown. As there is no nuclear head in the apyrene spermatozoon of Strombus, the direction of growth of the axial fibers must determine the poles of the cell. The blunt, rounded end is the one toward which the fibers have grown and is therefore anterior, while the narrow pointed one is posterior. In movement it is the morphologically posterior end of the spermatozoon which is usually directed forward, although occasionally the reverse is seen to occur. The movement of the spermatozoon is caused by contraction waves which pass alternately down the two membranes. At first these contraction waves are long and slow, propelling the spermatozoon with an even, steady motion. They usually pass down the membranes in a postero-anterior direction, but occasionally a reversal takes place; it is then that the sper- matozoon is seen to move with its anterior end directed forward. The 208 Papers from the Marine Biological Laboratory at Tortugas. movement of the spermatozoon is comparatively slow and is not long continued, for it soon attaches itself, by means of its tail-piece, to the glass slide or other object on which it is being observed. As soon as this occurs, the contraction waves pass down both the membranes simultaneously and they then become much shorter and faster. With the spermatozoon attached in this way, the membranes may continue to be active for a time which varies with the medium used. In sea-water, contractions of the membranes were still seen after two hours, in isotonic NaCl after five hours, but in isotonic MgCle they were very quickly inhibited. It very frequently happens, however, that the tail-piece breaks off and the spermatozoon swims away with a very much more rapid movement than it had at first. When the tail-piece begins to break it can be seen to be composed of a number of fused flagella; sometimes, as the spermatozoon moves away a few of these flagella still adhere to it. This probably explains the statement of Brock to the effect that a tuft of flagella, which is invisibile at first, is to be seen after the spermatozoon has been swimming about for a while. The first indication of breakdown in the spermatozoon occurs among the albuminous bodies, in that they gradually disappear. The process is undoubtedly one of liquefaction, for they leave in their stead a brownish semi-fluid substance in which, however, may still be seen the outlines of these bodies. When this has occurred one can observe for the first time the myoneme-like striations that are present on the cell-wall. Accompanying the disappearance of the albuminous bodies, there is a general flattening of the cell-body and the membranes. Undulations of the membranes con- tinue a long time after this flattening has taken place and lead eventually to a partial separation of the membranes from the cell-wall. After this their disintegration is rapid. The shape of the apyrene spermatozoa of Strombus bituberculatus is ery constant and the variations in size that occur are very slight. From the centrosomal plate to the tip of the tail-piece the spermatozoa measure about 90 micra; the width of the membrane is usually about 5 micra and the greatest diameter of the cell-body averages about 8 micra. In S. costatus the apyrene spermatozoa are much longer, measuring 120 to 125 micra, but the other dimensions are not increased proportionately ; according to Brock’s statement they attain a length of 180 micra in S. lentiginosus. The apyrene spermatozoa in S. bituberculatus and in \S. costatus are greatly outnumbered by the eupyrene. Brock has estimated the numerical relation between them to be about 1 to 500 for S. lentiginosus. Because of the failure to obtain ripe eggs of Strombus at the time of the year when the material for this investigation was collected (March to the middle of May), it was impossible to d.termine experimentally whether or not the apyrene spermatozoa have any function with regard to the eggs. Ripe eggs of several sea-urchins were to be had in great abundance, how- ever, and those of Tripneustes were mixed with some of the contents of the sperm-duct of Strombus. The presence of the eggs had no effect whatever Papers from the Marine Biological Laboratory at Tortugas. 209 upon either kind of spermatozoa except in a few instances where the egg- membrane had been ruptured by the pressure of the cover-glass. In such cases the eupyrene spermatozoa were strongly stimulated and swarmed around the eggs in great numbers, but the apyrene spermatozoa in the immediate vicinity of the eggs were not affected in the slightest. At the same time of the year the seminal receptacle of the female was found to be quite swollen and full of a whitish mass. When some of the contents was diluted with sea-water and examined it was seen that in addition to a large number of oval granules only the eupyrene spermatozoa were present. These latter were normal and very quickly became active. Of a dozen females which were examined, not one showed a single structure in the seminal receptacle which could possibly be identified as the remains of an apyrene spermatozoon. The explanation of this phenomenon was found only when sections of the seminal receptacle had been made and studied. The uterus and seminal receptacle constitute that portion of the vagina which lies just beyond the orifice of the oviduct, beneath the mantle. The vagina itself is simply a groove running beneath the mantle, at its base and on the right side, from the orifice of the oviduct proper to the foot; it corre- sponds to the genital groove of the male. Just below the orifice of the oviduct, the upper fold of the vagina is much broadened and thickened and is covered by the lower wall which is correspondingly broadened, forming a contractile sheath or mantle. This thickened portion of the vagina is the true seminal receptacle; it is a gland comprised of a large number of con- voluted tubules lined with a secretory epithelium. The space between the seminal receptacle and its sheath may be spoken of as the uterus. Lying around the seminal receptacle in a semicircle is the main portion of the nidamental gland; one branch of the latter extends down along the vagina for a certain distance. The ducts of the nidamental gland open into the vagina, just below the seminal receptacle. Further on down, and at the point where the lower branch of the nidamental gland ends, is the opening into the vagina of a large diverticulum, the bursa seminalis. Beyond this point the vagina traverses the foot as an unmodified groove. In the sections mentioned above, it was found that there was a dense detritus in the uterus composed of eupyrene and apyrene spermatozoa. Apyrene spermatozoa with relatively few eupyrene scattered amongst them formed the inner portion of this detritus. Segregating themselves from the apyrene spermatozoa, the eupyrene had moved for the most part to the periphery of the detritus. A clear homogeneous substance separated them from the apyrene spermatozoa. Other eupyrene spermatozoa had moved on into the tubules forming the seminal receptacle and had there arranged themselves with their heads against the outer surface of the epithelial cells. Here they were undoubtedly receiving nourishment, as was shown by the vacuoles which had taken the place of the distal granules of the secretory cells. In five series of sections no traces of apyrene spermatozoa were 14 210 Papers from the Marine Biological Laboratory at Tortugas. found in the tubules nor was either kind of spermatozoa found in the upper portion of the oviduct. The changes that the apyrene spermatozoa had undergone were quite remarkable. In some cases the spaces between the albuminous bodies were packed with very small darkly staining granules; in others these granules had replaced the albuminous bodies in a portion of the spermato- zoon; and in others the albuminous bodies had entirely disappeared and the whole body of the spermatozoon was filled with the small granules. In a few instances the body of the cell presented the appearance of an empty, shrunken shell lying between the two undulating membranes. Clearly the process going on in the apyrene spermatozoa was concerned with the breaking down and ultimate exhaustion of the substance stored in the cell. It is probable, then, that in the examination of those dozen females only the tubules of the seminal receptacle had been tapped, and not the uterus, a contingency that is highly possible, depending upon the point of the incision and its depth. If that were so, then the granules observed were from the secretory granules of the tubules. But it is also possible that the uterus itself had been tapped and that the apyrene spermatozoa had undergone katabolic changes to such an extent that they were no longer recognizable.! OBSERVATIONS UPON THE LIVING CELLS OF THE TESTIS. By treating the contents of the testis in the manner already described, one can make a very comprehensive, if not detailed, study of the develop- ment of the apyrene spermatozoa of Sirombus. ‘This is particularly true of the later stages, where such structures occur as are represented by fig- ures I to5. These five figures, together with figure 6, show four very char- acteristic stages of the developing spermatosome from the time the axial fibers have grown across the cell up to and including the adult spermatozoon. They give a general idea of the appearance of the living cells and of the extent to which details of structure can be made out in them with the aid of an oil-immersion lens (Zeis 2 mm. apochromatic); furthermore, they bring out contours and the size relations existing between various stages to the greatest possible advantage. In studying the living cells it was the hope of the writer not only to verify the stages which had been traced in fixed and sectioned material, 1 While a guest at the Laboratory of Marine Biology of the Carnegie Institution at Dry Tortugas during June 1913, the writer had the opportunity of examining some 25 females of Strombus gigas, all of which had copulated. While it is too late to present a detailed account of this examination, it may be stated that the results obtained give ample confirmation of the statements made above. The fates of the eupyrene and of the apyrene spermatozoa after copulation are very different. The eupyrene spermatozoa reach the seminal re- ceptacle and are stored there in great numbers, remaining alive for an unknown period. Some of the apyrene spermatozoa move into the uterus along with the eupyrene, but there a sharp separation of the two kinds takes place, the eupyrene spermatozoa moving into the seminal receptacle while the apyrene clump together and undergo degenerative changes. Eventually the uterus becomes rid of them. But the greatest portion of the apyrene spermatozoa never even reach the uterus; instead, together with some eupyrene spermatozoa, they move into the bursa seminalis and there undergo degeneration until they are completely broken down. Gradually the whole mass is encapsulated in a secretion from the walls of the bursa seminalis and is finally extruded. In some instances it was found that the contents of the extruded capsule had become infected and that putrefaction had set in. There are some grounds for the writer’s belief that some of the eupyrene spermatozoa which get into the bursa seminalis find their way out again and, passing up the vagina, reach the seminal receptacle. The most significant point is that oviposition had not yet begun. The description given above of the relations of the various parts of the female genitalia is based upon dissections of Strombus gigas. Papers from the Marine Biological Laboratory at Tortugas. 2i% but also, if at all possible, to watch the processes of growth and differ- entiation as they were taking place. All attempts, however, to observe the breaking down of the nucleus and the division of the centrioles proved to be unsuccessful. These processes are completed very rapidly and it is doubtful whether the centrioles with their radiations could be distinguished intra vitam. On the other hand, however, such processes as the vesiculation of the nuclear fragments, the formation of the albuminous bodies, and the growth of the axial fibers take place so gradually that they showed no appre- ciable progress in cells which were under constant observation for two or three hours. Nevertheless, many different stages in the development of the apyrene spermatozoa could be distinguished and observed most satisfactorily. The large nucleated cell, corresponding to the oligopyrene spermatocyte I of Paludina, was easily recognized. Earlier stages in the growth of that cell could not be distinguished from the eupyrene spermatocytes. The next stage (fig. 1) which could be easily picked out was one in which the axial fibers had grown well across the cell and many of the nuclear frag- ments had become vesiculated. Figure 2 shows a stage of approximately the same age, but viewed at an angle of 90° from the other. In these stages the cell is not much larger than it was just previous to the breakdown of the nucleus. Later on, the continued growth of the axial fibers results in the elongation of the cell; this makes it appear larger, but no real increase in its volume has taken place. This is shown in figures 3 and 4, which represent stages following closely one upon the other and viewed in planes at right angles to each other. In figure 3 the bundle of axial fibers is seen just in the act of splitting apart, while in figure 4 the process is represented as being probably a little more advanced. This splitting of the bundle of axial fibers is followed by the further separation of the halves until they reach the cell-wall on both sides. Very soon they begin to project edgewise beyond the cell, giving the first indi- cation of the formation of the undulating membranes (fig. 5). Beyond this point there is not a very great increase, if any, in the actual volume of the cell; it merely becomes gradually longer and narrower until the adult con- dition is reached (fig. 6). It will be noticed that as the cell approaches maturity it becomes more and more filled with the albuminous bodies, while the nuclear fragments degenerate and disappear. When first formed, the albuminous bodies are spherical in size and are scattered irregularly throughout the cell, first filling up its anterior portion. Later on, as a result of their increase in number and the lateral compression of the cell due to its elongation, they become polygonal and arrange themselves in an orderly fashion. Until the undulating membranes have been formed the cell is immotile; the flagella at its base never beat, but instead undergo fusion to a certain extent. As will be shown later on, the myoneme-like striations are not present until after the formation of the membranes. The study of the living cells of the testis is very valuable, because by this method all possibilities of abnormality due to fixation and subsequent 212 Papers from the Marine Biological Laboratory at Tortugas. treatment of the tissue may be eliminated. However, it leaves great gaps in the stages of development of the apyrene spermatozoa and affords very little or no chance of determining the finer details of the protoplasmic differentiations and the structural changes that occur in the cell. For this purpose the study of material that has been well fixed and stained is alone available. GROWTH OF THE APYRENE SPERMATOBLAST. In Strombus the large nucleated cell which becomes differentiated into the apyrene spermatosome can be traced back through a period of unin- terrupted growth to a small cell, the apyrene spermatoblast, which is distinct from either the eupyrene spermatocytes or spermatogonia. This cell (fig. 8) lies close to the walls of the lobules of the testis and may be surrounded by older cells of the same nature or by the general syncytium of the testis or partially by both. Two or three of these cells, lying in the syncytium and isolated from others of their kind, are frequently in close proximity to a nest of eupyrene spermatogonia and would be taken at first glance to belong to that nest. Closer and more careful examination, however, reveals certain distinct, if minute, differences between them and the spermatogonia. In the first place, there is a difference of position. Where they occur, the young apyrene spermatoblasts always lie between the wall of the lobule and the spermatogonia. If the group of apyrene spermatoblasts is a larger one, it is seen that the outer cells are larger and older than those lying next to the wall of the lobule. Such a group as that may be surrounded by several nests of spermatogonia or by spermatocytes. A comparison of all the nuclei of all these various cells shows that those of the outer spermato- blasts are as large or larger than the spermatogonial nuclei, while those of the inner spermatoblasts are smaller. But the cytoplasmic bodies of even the youngest spermatoblasts are larger than those of the spermatogonia and they are inclosed by definite membranes. The cytoplasm of the spermato- gonia forms a thin sheath around their nuclei and the cell-walls are so delicate that very often it is impossible to distinguish them in the cytoplasm of the surrounding syncytium. The nuclei of the apyrene spermatoblasts and of the eupyrene spermato- gonia differ in pattern. In the latter, during the resting stages, the chro- matin forms an irregular and dense network, the strands of which show many free ends and, when seen in cross-section, have very frequently the appear- ance of granules. Once the spermatogonia have begun to grow, the nuclear pattern changes gradually until the typical spireme of the spermatocyte is formed. The nuclear pattern of the spermatoblasts, on the other hand, remains constant until a late period in the growth of these cells, when they can be easily distinguished from the eupyrene cells by their increased size. These nuclei are relatively rich in nuclear sap and poor in chromatin, the latter being present in the form of irregular granules which are partially con- nected by a chromatic material (figs. 8-14). At least one karyosome is always present in the nucleus. + Papers from the Marine Biological Laboratory at Tortugas. 213 Finally there are two cytoplasmic structures which aid in distinguishing the young apyrene spermatoblasts. The first of these is a very evident, darkly staining granule, whose derivation is unknown. Its position varies; it may lie just beside the nucleus or nearer the periphery, but it is always found in the proximal half of the cell. It usually appears to be surrounded by a clear area (figs. 8 to 12). Occasionally a cell is found which has two oc three of these granules in it. The other structure is the centrosome. This appears at first as a distinct centriole with a few faint rays and it lies between the nucleus and the base of the cell (fig. 8). In other spermato- blasts, which are still so young that they might be confused with the sper- matogonia, it may be seen that the centriole has divided into two and that these are surrounded by a sphere (fig. 9). That centrosomes do exist in the spermatogonia is proved by their appearance in mitoses, but the writer has never been able to identify them positively in the resting cells. The distinct appearance of the centrosomes (centrioles) in the young spermato- blasts is not at all surprising when it is borne in mind that their later history is quite complicated and that they play a very important part in the develop- ment of the apyrene spermatozoa. The other bodies are never found in the cytoplasm of the spermatogonia. The eupyrene spermatogonia greatly outnumber the apyrene spermatoblasts and, with the basal nuclei, they occupy the space next to the wall of the testis except here and there where they have been displaced by groups of the spermatoblasts. These groups occur at irregular intervals and the cells composing them vary not only in number but also in age. The development of the apyrene spermatozoa from these spermatoblasts takes place, as has already been indicated, without a single division of the cell. The first period of this development covers a gradual growth of the spermatoblast, resulting in a tremendous increase in the size of the cell. This statement applies not only to the nucleus and cytoplasm, but also to the centrosome; the latter comes to be the most striking and interesting element in the cell. The structure of the primitive spermatoblasts has been described. They are irregular in outline and are not pear-shaped or stalked as are the oligopyrene spermatocytes of Paludina (fig. 8). Usually, as they increase in size and are crowded away from the wall of the lobule by the formation and growth of younger spermatoblasts, they are set free and tend to become round (fig. 12). Occasionally a cell which has almost attained its full growth is seen to have retained a connection with the syncytium by means of a stalk, even after it has been crowded away (fig. 13); such a cell presents a close resemblance to the oligopyrene spermatocytes of Paludina. In some few instances two or three fully grown spermatoblasts may be seen lying side by side close to the wall of the lobule and only separated from it by a thin layer of syncytium. These cells were the last ones to be formed at that point and have undergone their entire growth without suffering displacement. Both in the cases of these and of the stalked cells there is 214 Papers from the Marine Biological Laboratory at Tortugas. no relation between them and the syncytium other than one of close con- tact. Thesyncytium has become exhausted and has come to form a narrow fibrillar layer between them and the wall of the testis. This shrinkage of the syncytium beneath the spermatoblasts is noticeable in much earlier stages (figs. 9, 10, and 11). The growth of the nucleus keeps pace with that of the cytoplasmic body. In the very young spermatoblasts the chromatin has the appearance of irregular granules lying in a matrix of diffuse achromatic material. There is always a karyosome present (fig. 8). Practically the same nuclear structures are shown in figures 9 and Io except that in the former the karyosome was just above the plane of focus and in the latter the greater portion of the nucleus was in the next section. The karyosome is very persistent and constant in its appearance; in sections which have been destained until the other structures of the nucleus have lost their stain, the karyosome presents the appearance of a thick ring inclosing a lighter colored substance. As the nucleus grows larger, there is an increase in the amount of achromatin present and in the number of chromatin granules (compare figs. 10, II, and 12). The increase in the number of these granules is due, probably, to their division. The subsequent growth of the resulting halves keeps them all of a fairly constant size. When the nucleus has reached its maximum size the chromatin granules begin to grow at the expense of the achromatin (fig. 14). Their growth continues until they are as large or larger than the karyosome, so that the latter becomes indistinguishable. Eventually, practically all of the achromatin is absorbed and the chromatin granules are retracted to the periphery of the nucleus, just beneath the membrane. Here they undergo fusion to a certain extent and form large lumps of chromatin, or karyomerites, of various sizes and shapes (figs. 16, 17, and 18). The interior of the nucleus is filled with a colorless nuclear sap (fig. 17). It is now ready to break down, but before describing this process it will be well to consider the growth of the centrosome. The first indication of a centrosomal structure in the spermatoblast appears at a very early stage, in the form of a distinct, darkly staining centriole, from which two or three rays (fig. 8) pass out. The position of this centriole, near the nucleus and in the proximal half of the cell, is in- variable. The formation of a sphere around the centriole is not noticeable until the latter has divided (fig. 9). In such a stage as this the sphere is quite distinct and rays from both centrioles can be seen. It is impossible to say whether the next stage (fig. 10) is reached by a division of the two centrioles into four or by their fragmentation into a greater number. The centrioles are very much smaller than they were and can not be counted with any degree of certainty. The sphere has grown no larger and the rays from the centrioles do not extend beyond its periphery. The centrioles now lie in a homogeneous substance which fills the center of the sphere and is more deeply staining than the latter. This substance represents the first formation of the centrosome proper. Papers from the Marine Biological Laboratory at Tortugas. 215 Further multiplication of the centrioles is accompanied by their separa- tion from each other and by the spreading of the substance of the centro- some (figs. 11 and 12). It becomes very difficult to distinguish the indi- vidual centrioles from the substance in which they lie, but several of them can usually be seen lying at various points on the edge of the centrosome. The outer sphere, too, has increased in size and about this time begins to resolve itself into an outer dark court and an inner clear court (fig. 12). A little later on, the centrosome, which is now considerably larger, becomes spherical and acquires a definite membrane and the rays become stronger and stronger (fig. 13). It reaches its full growth at approximately the same time as does the nucleus (fig. 14). From it there passes out in all directions a great number of strong rays, which extend to all parts of the cell; at the base of each lies a small centriole. Lying around the centrosome on all sides, but more particularly on the sides away from the nucleus, is a mass of mitochondrial granules (figs. 14 and 15). They are irregular in shape and size and stain deeply with iron hematoxylin. Figure 14 does not show all the mitochondria that were seen in the cell. The formation of the mitochondria is rather sudden. They appear in the sphere shortly after the formation of the inner court and are found at first on all sides of the nucleus. Figure 13 represents approximately the stage at which they first appear; it was drawn from a cell which showed the effects of osmic acid to some extent and therefore does not represent either the mitochondria or the rays as fully as they might be seen in other cells. As the mitochondria increase in number and size, they gradually obliterate the outer court of the sphere, while the inner court grows wider and wider. The last traces of the outer court are shown in figure 15. With the appear- ance of the mitochondria, one loses sight of that body which was mentioned as being found in the cytoplasm of the young spermatoblasts (figs. 8 to I2). This body gradually grows larger and very pronounced, but later on it is impossible to distinguish between it and any of the larger mitochondrial granules. Judging from the great number of cells which are found with a fully formed centrosome, the last stage in the growth of the spermatoblast must extend over a very long period. Gradually the mitochondria withdraw from the space between the centrosome and the nucleus and that half of the centrosome which lies toward the nucleus becomes more lightly staining and seems to lose its definite outline. On the other hand, the opposite half of the centrosome takes a deeper stain and it can actually be seen that the centrioles are massing themselves upon that side (fig. 14). At this time, one or two comparatively large granules are to be found in the centrosome; they are very constant in their appearance but no suggestion can be made as to their significance. Finally, those rays which traverse the space between the centrosome and the nucleus attach themselves to the nuclear wall (fig. 14); then follow, in quick succession, the retraction of the chromatin to the wall of the nucleus, the dissolution of the latter on the side toward 216 Papers from the Marine Biological Laboratory at Tortugas. the centrosome, the scattering of the karyomerites, and the disappearance of the centrosome with its rays. The actual beginning of dissolution of the nuclear wall is shown in figures 15, 16,17, and 18; these go in pairs, figure 15 with figure 16 and figure 17 with figure 18, representing consecutive sections through two different cells. In figure 15 the mitochondria lie for the most part above the centrosome, while the latter in turn is at a slightly higher focus than the karyomerites. These last, represented in the figure as large black bodies, lie free in the cytoplasm, into which also some of the nuclear sap has escaped. Many of the mitochondria have grown into short stout rods (chondrioconts). It will be noticed that the rays from the centrosome are becoming less evident. In figure 16 isshown the remainder of the nucleus; some of the nuclear sap and a few of the karyomerites are moving out into the cytoplasm. The further and complete scattering of the karyo- merites is usually accompanied by the total disappearance of the centro- somal rays. But in figures 17 and 18 a cell is shown in which absolutely no traces of the centrosome or the rays could be found and as yet none of the nuclear material had escaped into the cytoplasm. These figures were drawn very carefully and show all the structures visible in the cell. The mitochondria were rendered somewhat indistinct by the action of the osmic and acetic acids, but the same can hardly be said of any surviving centrosomal structure; for a younger spermatoblast lying beside this cell showed a distinct centrosome. The nuclear membrane has become very much thinner and at one place, on the side towards the mitochondrial apparatus, it has been dissolved (fig. 18). It would therefore seem that the same process, undoubtedly a chemical reaction, which brings about the disappearance of the centrosome also leads to the dissolution of the nuclear membrane. Of the two phenomena, the latter is completed the more slowly. The karyomerites are carried to various parts of the cell with the diffusing nuclear sap. The greater part of the chromatin remains in what was originally the distal half of the cell, although some of the fragments move into the region occupied by the mitochondria. This distribution of the chromatin is illustrated in figures I9 and 20, which represent two consecutive sections through the same cell. That part of the cell shown in figure 19 constitutes the proximal half of the cell, while the distal half is shown in figure 20. The larger lumps of chromatin now break down into several smaller parts (figs. 20 and 22), and as a result of this fragmentation there is a more equal distribution of the chromatin throughout the whole cell (figs. 24 and 25, consecutive sections through the same cell). The ultimate karyomerites show an extreme variation in size, some of them being extremely small (fig. 22). Kuschakewitsch (’11) has described a somewhat similar formation of karyomerites in the apyrene cells of Vermetus. Here the karyomerites apparently are all fully formed before the dissolving of the nuclear mem- brane. They are of a fairly uniform size and shape and many of them show Papers from the Marine Biological Laboratory at Tortugas. 217 a longitudinal split, thus leading to the belief that they are true chromo- somes. Kuschakewitsch states that ‘the chromatin of the nucleus gathers into larger elements, which almost entirely fill up the interior of the nucleus (Kernraum). In some cases there are formed, transitorily, true chromo- somes which exhibit a clear longitudinal split. Then the nuclear membrane disappears and the chromatin elements (karyomerites) lie free in the plasma.’”’ The outstanding difference between the processes relative to the breaking down of the nucleus in the two forms is the fact that in Strombus the ultimate karyomerites are not formed within the nucleus while the latter is yet intact. Furthermore, when finally established, they have an extreme variation in size, little uniformity in shape, and never show any indications of a longitudinal split. In short, unlike those of Vermetus, the karyomerites of Strombus can not be said to be chromosomes, nor is it at all certain that they represent an attempt on the part of the chromatin to form chromosomes. The further history of the karyomerites is one concerned with processes leading to their vesiculation and, later, to their gradual degeneration and ultimate disappearance. When first scattered into the cytoplasm the karyomerites lie in close contact with it on all sides. Very shortly after- wards, a clear space, one which gradually grows larger, separates them from the surrounding cytoplasm (figs. 20 to 25). The substance filling this space is essentially like nuclear sap and is probably formed by a differ- entiation of the cytoplasm immediately surrounding a karyomerite. A little later a definite membrane can be seen to have been formed around the clear area, separating it and the inclosed karyomerite from the cytoplasm (figs. 26 to 32). It is quite possible that this membrane is formed asa pre- cipitation membrane between colloids of opposite electrical charges. If this explanation is the true one, then the differentiation of the cytoplasm must have been caused by a chemical reaction between itself and the karyomerite, the substance formed acquiring the acid properties of the chromatin. The reaction must continue until the interposition of the formed substance causes too great a separation of the reacting substances. Then, when the reaction has ceased, a precipitation membrane is formed between the acid and therefore positively charged substance resulting from the reaction and the basic and negatively charged cytoplasm. After the formation of the membrane, the inclosed karyomerite breaks down into several smaller pieces, each of which takes up a position just beneath the membrane (figs. 26 to 32). In this way a number of chromatic vesicles arise which are very characteristic for the ensuing stages, and which, as will be shown later, persist for a long time before their actual degeneration sets in. Most of the smaller chromatic fragments, however, do not form vesicles but are absorbed without any further changes. The disappearance of the centrosome takes place very suddenly and the nature of the process is not entirely understood. Absolutely no struc- tures were found which could be interpreted with any degree of assurance 218 Papers from the Marine Biological Laboratory at Tortugas. as being phases of the centrosome subsequent to the stage shown in figure 15. One cell alone showed an indistinct structure which might be taken for the remains of a rapidly dissolving centrosome, but this can not be asserted to be the case (fig. 19). The body was round and apparently denser at its periphery than in the middle and from it there passed off a number of short and rather vague rays; it was partially hidden by what is taken to be a large mitochondrial body. The figure of it is not presented because much importance is attached to it, but merely because it happened to occur in a cell which was valuable for other features which it showed. All that is definitely known to occur is that while the substance of the centrosome is dissolved and disappears very rapidly, the centrioles themselves are un- affected by the action. Having been freed, they scatter in small groups. The movements of the centrioles are hidden to a great extent, as it is very difficult to establish their identity among the many and very small mitochondrial and chromatic granules present. They eventually reach the cell-membrane, however, either singly or in small groups, and here some of them can be discerned quite readily (figs. 19, 24, and 25); others of perhaps questionable identity can sometimes be seen in the interior of the cell (fig. 19). A little later, secondary rays arise from those centrioles which have reached the membrane, causing them now to become very evident (figs. 21, 22, 26, and 27). Figures 21, 22, and 23 are consecutive sections through the same cell, and that portion of the cell represented by figure 21 showed the majority of the centrioles; those lay beneath the plane of the focus figured. A few others were found in the next section of the cell (fig. 22), but none were found in the upper section (fig. 23). It is thus seen that the centrioles come to lie on the periphery in one half of the cell and this half is the one which was originally proximal to the wall of the testis.1 The distribution of the centrioles was seen to better advantage in the cell represented by figures 26 and 27. Here they were found entirely confined to the left half of the cell. Those shown in the figures constitute probably less than one-quarter of the total number in the cell, but only those were figured concerning whose identity there could not be the slightest doubt. The same thing may be said of figures 28 and 29 which represent a slightly older cell. The centrioles are seen to have gathered into larger groups and have moved closer together. From each centriole, too, there has grown out a short flagellum. The rays from the centrioles were some- what obscured on account of the ‘‘osmication”’ of the cell. These, however, are well shown in figures 30, 31, and 32, which represent consecutive sections matoblasts as have never been forced away from the syncytium. Four facts must be borne in mind: (1) in the young spermatoblasts the position of the centrioles is invariably in the proximal half of the cell; (2) matured spermatoblasts which lie close to the wall of the testis are independent and are not a part of the shrunken syncytium; (3) in such spermatoblasts the centrosome occupies a position distal to the wall of the testis; (4) they undergo their later development in the same position. It is impossible to believe that the centrosome and nucleus of these spermatoblasts have reversed their position during the growth of the cells. On the other hand, in view of the fact that these cells usually lie free except at the point of contact with the syncytium, it will be readily granted that the whole cell has probably revolved. Furthermore, a few instances of spermato- blasts in various stages of growth have been observed, the position of whose centrosomes would indicate a turning of the cell. Therefore, in these spermatoblasts, that portion which appears to be the distal half is really the originally proximal one. After the disappearance of the centrosome in such spermatoblasts as lie close to the wall of the testis, the centrioles come to lie beneath the membrane in the half of the cell directed toward the lumen, that is, in the originally proximal half. 1 The fact that this half of the cellis the originally proximal’one is easily established in the case of such sper- = Papers from the Marine Biological Laboratory at Tortugas. 219 through a cell of approximately the same age as the preceding one. This cell and others similar to it present the appearance of having polyasters on its one side, while on the other lies the majority of the now vesiculated karyomerites. This appearance is not maintained for any length of time, for the centrioles, with the lengthening of their rays, draw together and form a somewhat oval plate just beneath the membrane (fig. 33). The flagella now form a short, thick brush on the outside of the cell. The rays which have now grown very strong, traverse the cell for a short distance as a compact column and then, diverging, they extend to all parts of the cell. A little later the centrioles divide and then the distal halves begin to move across the cell, forming fibers between themselves and the proximal halves. The latter, which are perhaps smaller than the others, flatten themselves against the membrane, each one forming a node at the base of its flagellum. The mitochondria can now be seen lying above the centrioles and encir- cling the rays which pass off from them. This irregular ring of granules is shown in cross-section in figures 33 and 35 and in surface view in figure 34. After the dissolution of the nuclear membrane and the dispersal of the karyomerites the mitochondria become obscured for a time, owing to the great difficulty of distinguishing between them and the smaller chromatic elements. They apparently remain in much the same position they occu- pied before the dissolution of the centrosome. With the vesiculation of the karyomerites they can once again be distinguished, but they now appear much more irregular in outline and size (figs. 28 and 29). Later they form the ring mentioned above. The spermatoblast has now reached a condition in which it is ready to undergo the growth and differentiations which lead directly to the forma- tion of the adult apyrene spermatozoon. With the completion of the pri- mary growth of the cell, the nucleus has broken down to form a number of small chromatic vesicles. As will be shown later, these are concerned with the secondary growth of the cell, after which their degeneration sets in. By the accumulation of the centrioles at one point of the cell, and their subsequent division, the orientation of the matured apyrene spermatozoon is established; for, as has been pointed out, the antero-posterior axis of the spermatozoon is determined by the direction in which the distal halves of the centrioles traverse the cell. For these reasons the cell may now be spoken of as the apyrene spermatosome. Before proceeding with the description of the growth and differentiation of the spermatosome, two facts must be considered concerning the spermato- blast: The first has to do with the polarity of the cell and the second with the lack of a strict time correlation between the breaking down of the nucleus and the disappearance of the centrosome, and between the phe- nomena immediately subsequent to both. It has already been explained that the orientation of the apyrene sper- matozoon is a morphological one, based upon the direction of growth of 220 Papers from the Marine Biological Laboratory at Tortugas. the bundle of axial fibers. It remains to be pointed out that this direction of growth is determined at the time of the appearance of the first centro- somal structure (the centriole) in the very young spermatoblast (fig. 8) and is parallel with an imaginary line passing through the centriole and the center of the nucleus. This line is the chief axis of the cell and it establishes not only the anterior and posterior poles of the adult spermatozoon, but also the plane of its bilateral symmetry. The centrosome is known to lie in the proximal half of the cell and upon its disappearance the centrioles come to be pretty evenly distributed on the periphery of the same half (figs. 26 and 27). If, insassembling, the centrioles move toward a central point, as there is every reason to believe, then the center of the plate formed by them will lie in the chief axis of the cell. After the division of the cen- trioles the distal halves move across the cell in a direction vertical to the plate and therefore parallel to the chief axis. That the chief axis of the cell establishes a bilateral symmetry is shown by comparing figure I with figure 2 and figure 3 with figure 4. The first and third of these figures represent two different stages of the apyrene spermatosome, while figures 2 and 4 represent stages almost corresponding to them, but viewed at right angles to them. In the first case, a plane passing vertically through the longitudinal axis of the bundle of axial fibers divides either of the cells into like halves, while in the second case a similar plane divides either of the cells into unequal and dissimilar portions. The same bilaterality is established by the chief axis in the spermatoblast, indeed in very early stages (figs. 12, 13, and 14). Thus it appears that the polarity of the apyrene spermatozoon is determined at a very early stage in its development, at the time when the chief axis is established. The distal side of the spermatoblast marks the anterior end of the spermatozoon and the proximal side the posterior end. The latter may therefore be spoken of as the base of the spermatoblast or spermatosome. While the dissolution of the nuclear membrane and the disappearance of the centrosome are initiated simultaneously, the phenomena immediately concerned with these two processes and also those subsequent to both do not keep step in different cells. It has already been pointed out that while in one cell (figs. 15 and 16) the centrosome was still perfectly visible after some of the karyomerites had escaped into the cytoplasm, in another (figs. 17 and 18) no traces of the centrosome could be found even before this had occurred. Similarly, in one cell (figs. 19 and 20) all of the centrioles have not yet reached the periphery and the larger karyomerites are frag- menting, while in another (figs. 21 to 23) the centrioles already show their secondary rays, although we are dealing with what is clearly an earlier stage as regards the chromatin. But ina third cell (figs. 24 and 25), and here the greatest variation occurs, we have a condition in which the centrioles have reached the periphery and their secondary rays have not as yet appeared, while the formation of the ultimate karyomerites has already taken place. Papers from the Marine Biological Laboratory at Tortugas. 221 The occurrence of variations such as these seems to indicate that the processes concerned with the nucleus have great independence from those concerned with the centrosome and vice versa—greater than is known to exist in both normal and abnormal mitosis. In abortive mitoses, such as the so-called oligopyrene spermatocytic divisions of Paludina, there is always maintained a constant time correlation between the phenomena which occur in the nucleus and those which occur in the centrosome. An interpretation of the processes described above as being an abortive mitosis of the apyrene spermatoblast of Strombus seems to be unwarranted by the facts in the case. GROWTH AND DIFFERENTIATION OF THE APYRENE SPERMATOSOME. At first there is very little, if any, growth of the spermatosome; whatever growth there is, takes place slowly and for a long time but little difference in size exists between the spermatosome and the spermatoblast previous to the breaking down of the nucleus. The changes that occur are concerned chiefly with the chromatic vesicles and the centrioles. By the time the centrioles have divided, the vesiculation of the karyo- merites has been completed, or nearly so (figs. 33 and 34). The chromatic vesicles now acquire a very definite position in the cell. They lie in the outer portion of the cell, not far beneath the cell-membrane, and are con- fined almost entirely to the distal half of the cell. In a section passing through the chief axis of the cell, the chromatic vesicles are seen to lie in a semicircle, the open side of which points towards the base of the cell. This arrangement is shown particularly well in figure 39. At about this time, too, there is a very noticeable swelling of the chromatic vesicles due to the gradual increase in the amount of nuclear sap within them (fig. 41). In many instances an achromatic network is also visible in them (figs. 36, 39, 40, and 41). Ina word, the chromatic vesicles now present the appearance of so many small but active secondary nuclei whose reconstruction is not yet completed; and such, indeed, they are considered to be by the writer. It will be remembered that the secondary rays of the centrioles, at the time of the division of the latter, pass through the ring of mitochondria and then, diverging, they reach to the furthest parts of the cell. There is very little evidence in support of any belief that they attach themselves to the different chromatic vesicles; on the contrary, the ends of many of them can be clearly seen to pass on beyond the vesicles and lie free in the cyto- plasm (figs. 33, 34, 35, 36, and 39). After the division of the centrioles, the connection of the rays with the distal halves of the former, or the distal centrioles as they will now be designated, becomes less distinct. In front of the advancing distal centrioles there is formed a dense and homogeneous substance which very frequently obscures those portions of the rays which pass through it (figs. 34 and 35). Into this substance, too, from the distal centrioles, there grow out slender fibrillar processes which become more pronounced as the centrioles advance (figs. 34, 35, 36, and 39). These fora time are probably continuous with the rays, but as the latter become less 222 Papers from the Marine Biological Laboratory at Tortugas. distinct whatever connection there may be between them is gradually lost (figs. 36 and 39). As the distal centrioles move away from the base of the cell, there is formed between each pair of daughter centrioles an axial fiber (figs. 34, 35, etc.). These fibers are comparatively stout, more so than the flagella, and they stain very intensely with iron hematoxylin. In common with all the other centrosomal structures, they are brought out best in those portions of the tissue where the action of the acetic and osmic acids has not been too strong. In their journey across the cell, spinning out the axial fibers behind them, the distal centrioles come to pass through the ring of mitochondria (figs. 36 and 39). As they approach this ring they undergo a lateral com- pression in two opposite directions, so that they draw closer together and eventually fuse to form a solid, very flatly oval plate, whose greatest diam- eter is no less than that of the group of proximal centrioles at the base of the cell. The compression of the distal centrioles with the corresponding flattening of the bundle of axial fibers is shown in figures 36 and 39. It is shown to better advantage, perhaps, in figures 37 and 38, which represent two consecutive cross-sections through a bundle of axial fibers at about the stage shown in figure 39. The first of these cuts the axial fibers near the base of the cell and the second right through the ring of mitochondria; they bring out the difference in contour of the bundle of axial fibers at those two points and give some idea of the enormous number of fibers which constitute the bundle. In front of the advancing centrosomal plate lies the area of dense material which was mentioned before and which is now becoming more and more evident (figs. 36 and 39). Around the anterior portion of the bundle of axial fibers lie other masses of denser cytoplasm, the inner margins of which are lined by the mitochondria (fig. 38). At this stage, or a little later, in many instances the bundle of axial fibers becomes somewhat twisted and the centrosomal plate is thrown a little out of its horizontal position. Further growth of the axial fibers is accompanied by a further flattening in its anterior portion and a more complete condensation of the centrosomal plate, so that the whole structure, when viewed from the side, has a sharply pointed appearance (fig. 40), while its upper and lower surfaces are as broad anteriorly as they are at the base of the cell (fig. 41). The dense substance in front of the centrosomal plate is pushed on ahead, while the other masses of dense cytoplasm, together with the mitochondria, spread out along the bundle of axial fibers behind the centrosomal plate, forming a sort of sheath around the former. The mitochondria are shown in longitudinal section in figure 40 and in surface view in figure 41. As is indicated in figure 40, the spreading of the mitochondria and of the dense cytoplasm associated with them results in their separation into two parts, the posterior one of which comes to form a secondary ring around the base of the bundle of axial fibers. As before, this secondary ring is com- posed of the dense and homogeneous cytoplasm, the inner surface of which Papers from the Marine Biological Laboratory at Tortugas. 223 is bounded by the mitochondria. A cross-section through the posterior region of the spermatosome shows that the mitochondria of the secondary ring are closely applied to the bundle of axial fibers (fig. 42). The completed separation of the mitochondrial apparatus into an anterior and posterior portion is shown in figure 43. The peculiar structure of the secondary ring is maintained for a long time and the mitochondria connected with it persist in the adult spermatozoon. The mitochondria of the anterior portion, on the other hand, disappear very suddenly, at least no traces of them can be found after the stage represented by figure 43. It will be noticed that here already there has been a diminution of the denser cytoplasm around the anterior end of the bundle of axial fibers; by the time the centrosomal plate has reached the cell membrane practically all of it has disappeared (figs. 44 and 45). While it is possible that the anterior mitochondria may be dis- solved im situ along with the other substance, two other explanations may be offered for their disappearance: (1) they may move in among the axial fibers and there give rise to occasional mitochondrial granules that are later found scattered throughout the length of the bundle or (2) they may be drawn back to the base of the cell and augment the secondary ring that has been formed there. By comparing figure 42 with figure 37, it will be seen that after the forma- tion of the secondary mitochondrial ring the posterior portion of the bundle of axial fibers has become more nearly circular, that its diameter is decreased, and that the fibers themselves have drawn closer together. The reduction of the diameter of the bundle is also shown in figures 41 and 42. The process seems to continue proportionately with the gradual growth of the fibers (figs. 44 and 46). At first the growth of the axial fibers is very much slower than that of the flagella. It will be remembered that the flagella begin to grow out before the centrioles have divided. By the time the distal centrioles have traversed one-third the distance across the cell, the flagella have reached their maximum growth and are more than twice as long as the axial fibers (figs. 33, 34, 35, 36, and 39). From this point on they gradually become shorter and more ragged in appearance, eventually fusing at a very late stage to form the posterior tip of the spermatozoon (figs. 40, 41, 43-49, 63, and 64). This shortening of the flagella is probably due to their partial retraction into the cell, although their ragged appearance would suggest that they have been worn down by attrition with other free cells. v. Brunn (’84) has described a similar action on the part of the flagella for the oligopyrene spermatozoa of Paludina. His observations were con- firmed by Meves (’03), who explains the phenomenon on the ground that the flagella are the free ends of the axial fibers, each one passing through a ring formed by one of the proximal centrioles upon the cell-membrane. The shortening of the flagella is caused by the downward growth of the cell substance around and between them, so that, eventually, a portion of each flagellum comes to lie within the body of the cell. The facts that have 224 Papers from the Marine Biological Laboratory at Tortugas. been observed in the case of Strombus do not support this view. In the first place, there is not the slightest morphological evidence that the proximal centrioles form rings; they clearly form nodes just beneath the membrane and each one lies at the base of a flagellum. Then again, the flagella have begun to shorten before there has been any appreciable lengthening of the spermatosome in either direction; if any has occurred, it is certainly not sufficient to account for the very noticeable shortening of the flagella that has taken place. It remains to be pointed out that there exists a striking similarity between the apyrene spermatosome (as regards its centrosomal structure at the time of the early growth of the axial fibers) and the well-known ciliated epithelial cells of the Mollusca. In figure 35, for example, the proximal centrioles can well be compared with the so-called basal granules of the ciliated epithelial cells and the distal centrioles with the inner granules. The slender fibers growing out from the distal centrioles are analogous to the fibers which extend across the epithelial cell. With the progressive growth of the axial fibers across the spermatosome the resemblance dis- appears to a great extent. While a strict homology between the two structures can not be maintained, we have here a strong, if indirect, support for the Lenhossék-Henneguy theory of the centrosomal origin of the basal granules found in a ciliated cell. Several changes occur in the spermatosome about the time that the distal plate of centrioles reaches the cell-membrane. At this time or perhaps a little earlier, the chromatic vesicles begin gradually to lose their sap, so that they decrease in size and the chromatic granules come closer together (figs. 43-45). As the process continues the chromatin gathers on one side of the vesicle which is now becoming more and more reduced in size (figs. 46, 47, 48, and 49). Soon the membrane disappears, either by being dis- solved or by the total loss of the sap, so that the whole structure shrinks into a solid mass. In either case, what were formerly active chromatic vesicles are now reduced to degenerating lumps of chromatin (fig. 63); eventually, all of these are dissolved in the cytoplasm and disappear, leaving no traces in the adult spermatozoon. The whole process of the degeneration of the vesicles can be followed through in the living cells (figs. 1-6). The direct effect of the degeneration of the chromatic vesicles is found in the appearance at this time of the secreted albuminous bodies. Just as in some epithelial cells where the formation of the secreted granules is accompanied by the degeneration of the nucleus, so here the processes leading to the eventual disappearance of the chromatic vesicles go hand in hand with the secretion of the albuminous bodies. A similar corre- lation between the degeneration of the nucleus and the secretion of the albuminous bodies has been described for the so-called nurse-cells of Littorina (Reinke, ’12). In Strombus the albuminous bodies appear first in the anterior portion of the spermatosome; a few can always be found around the advancing end of the bundle of axial fibers by the time it has Papers from the Marine Biological Laboratory at Tortugas. 225 grown across the cell. In later stages they are formed in increasing numbers and gradually come to fill up first the middle and then finally the posterior portion of the spermatosome (figs. 48, 49, 63, and 64; also figs. I-6). By the time that all the chromatin has disappeared from the cell, practically all of the albuminous bodies have been formed (fig. 64). The formation of the albuminous bodies takes place in very much the same way as Kuschakewitsch (’i1) has described for the apyrene sperma- tozoon of Vermetus and which has been briefly reviewed on page 204. In Strombus there is not a general differentiation of the cytoplasm into a large number of thick-walled chambers. Instead, there is a continual for- mation of occasional vacuoles throughout the cell, those that are first formed filling up rapidly with the secreted albumen before the later ones have appeared. For this reason one never sees more than three or four vacuoles in any one given cell (figs. 43, 47, and 49). The extreme vacuolization shown in figure 46 is due to imperfect fixation and is nota natural condition. The process of the secretion of the albuminous bodies can be followed most readily in the posterior region of the spermatosome. In a vacuole such as one of those shown in figure 49, a matrix is laid down composed of a substance which has a marked affinity for nuclear stains. The appearance of this matrix differs to some extent with the degree of “‘osmication’”’ which the cell has undergone. In the inner portions of the tissue it is seen to be composed of a number of very irregular, darkly staining masses (fig. 52), while in the outer portion it appears to have the form of a network (fig. 63). It must be borne in mind that the magni- fication in figures 63-65 is less than in the preceding figures. Between these two extremes there stands a condition in which the network is made up of granules. Around this substance a clearer, homogeneous material is secreted, which gradually fills up the vacuole and effaces the matrix. In many instances narrow strands can be seen connecting this substance (an albumen?) with the surrounding cytoplasm (figs. 46, 49, and 54). When the vacuole has been completely filled, a membrane is formed around the secretion, the body assuming a spherical shape. The bodies now give up their stain (iron hemotoxylin) very readily; in the places where the spermatosome has been strongly ‘‘osmicated,’ they either appear unstained and yellow, or else they show the characteristic ‘‘Spiegelfarbung”’ of yolk granules. As has been stated, their change to the hexagonal shape seen in the adult condition is due to the pressure they exert upon one another after they have completcly filled the spermatosome and the latter has undergone its final elongation and constriction. By far the most striking changes that begin to take place at the time when the distal centrioles reach the cell membrane (figs. 44 and 45) are those which result from the continued growth of the axial fibers, viz., the formation of the undulating membrane and the resulting changes in the size and shape of the spermatosome. Having come to extend entirely across the cell, the bundle of axial fibers, by their continued growth, begins 15 226 Papers from the Marine Biological Laboratory at Tortugas. to protrude beyond the original boundary of the cell, causing the spermato- some to acquire a flask-shaped appearance (figs. 3 and 46). In the ‘“neck”’ of the spermatosome, those yolk bodies which were the first to be formed come to lie between the broad surfaces of the bundle of axial fibers and the cell-membrane; the edges of the bundle, on the other hand, lie in close contact with the cell-membrane (figs. 3, 4, and 48). At this stage, in the living spermatosome, there is always to be seen a very noticeable protuberance on one side of the brush formed by the flagella. This structure, although it is only a transitory one, is very constant in the living cells. The first indication of it is shown in figure 2, where it appears as a swelling at the base of the cell. Later it becomes very much longer and flattened in one axis (figs. 3 and 4) and appears finally to be withdrawn into the cell-body (fig. 5). The only trace of this protuberance that could be found in sectioned material is shown in figure 40. The point at which this protuberance occurs must mark the weakest part of the cell membrane; this is shown by the changes that occur when degeneration sets in. Aftera spermatosome in the stage represented by figures 3 or 4 has been in an artificial medium for two or three hours, the protuberance begins to become very much distended. This is followed by a shifting, to that quarter, of all the free contents of the cell, with the result that the bundle of axial fibers, instead of passing through the cytoplasmic body, lies tangentially to it and is surrounded by the cell membrane alone. This protuberance may be interpreted as a precocious lengthening of the cell in a posterior direction, the completion of which is prevented by the actual attachment of the fibers to the cell-wall. As will be shown presently, such a lengthening does take place at a very much later stage, but this time it is accomplished through the direct agency of the axial fibers themselves. After a spermatosome has reached the stage represented by figure 46, a longitudinal split occurs in the bundle of axial fibers, dividing it into two approximately equal halves (fig. 47). This is caused by the continued growth of the axial fibers without a corresponding lengthening of the cell. Growth of the spermatosome does take place, to be sure, resulting in its becoming longer and narrower, but this growth is not commensurate with that of the axial fibers, nor is it accompanied by an increase in the volume of the cell. If these facts are borne in mind, then the following steps in the formation of the undulating membranes will be easily understood. After the split occurs in the bundle of axial fibers, the two halves, or the secondary bundles, as they may now be called, move apart from each other, and gradually approach the cell-membrane from opposite sides. A lateral view of one of the secondary bundles, shortly after this movement has commenced, is shown in figure 48. By comparing figures 47 and 48, it will be seen that the albuminous bodies already formed are so placed that they do not obstruct the path of the fibers; those that eventually come to lie in the middle and posterior portion of the spermatosome are not secreted until after the fibers have reached the membrane on each side of the cell Papers from the Marine Biological Laboratory at Tortugas. 227 (figs. 49 and 63). As the secondary bundles separate, the individual fibers composing them fold back upon each other, so that several twists are formed in the bundles (figs. 49 and 63). Owing to the shape of the spermato- some, the secondary bundles reach the membrane in the anterior part of the cell long before they do so in the posterior region. Correspondingly, the differentiation of these structures to form the undulating membranes takes place progressively from the anterior to the posterior end of the spermato- some. Consequently, a series of cross-sections passing in the opposite direction through a spermatosome in the stage represented by figure 49 should show all the changes that take place both before and after the secondary bundles reach the surface. Such a series is represented by figures 50 to 58. In figure 51 it is seen that the secondary bundles, in moving apart, have broken the ring of mitochondria which was mentioned before as having been formed around the base of the primary bundle. Afterwards, the mitochondria come to lie between the secondary bundles, while the dense cytoplasm which has been associated with them is gradually dissolved (figs. 49 and 63). It will be noticed that posteriorly the secondary bundles are wider than they are anteriorly. As they approach the surface the fibers draw closer and closer together; as a result, the bundles stain more and more deeply and it becomes increasingly difficult to distinguish the individual fibers. It can be seen, however, and very clearly too, that there is a modification taking place in that those fibers which lie nearest the cell-membrane tend to draw slightly away from the others and to fuse with one another (figs. 52 and 53). By the time the bundles have reached the surface, the fusion of these outer fibers has been completed and they are so closely applied to the cell-membrane that the latter becomes indistinguishable from them (fig. 54). The next figures show successive stages of the process by which the secondary bundles push out the cell-membrane and come to extend laterally beyond the body of the cell. Figure 58 represents a section passing just behind the centro- somal plate and shows that the fibers have arranged themselves into definite rows, a fact also indicated in figures 56 and 57. As will be shown presently, this arrangement of the fibers is necessary for the complete formation of the undulating membranes. At a somewhat later stage, after the secondary bundles have reached the cell-membrane throughout their entire length, the myoneme-like striations are formed. This process is shown in figures 59 and 60. Some of the fibers which lie next to the cell-membrane now separate from their fellows and begin to move out across the surface of the cell just beneath the mem- brane. From this it would seem that there is not an actual fusion of the exterior fibers of the secondary bundles, but that they lie in very close contact with each other. It is probable, too, that the exterior fibers do not participate in the twisting of the secondary bundles spoken of above. Since the fibers are attached at both ends, the twisting of the secondary bundles must take place at their middle portions and doubtless proceeds in 228 Papers from the Marine Biological Laboratory at Tortugas. one direction and then in the other; it is simply an accompaniment of the separation of the secondary bundles and after these have reached the cell- membrane such twists as have taken place are apparently undone. At all events, as the secondary bundles begin to push out beyond the cell-body, a rearrangement of the inner fibers occurs in that they form a number of horizontal layers (figs. 56, 57, 58, and 60). Some of the inner fibers, appar- ently, crowd into the layer of exterior ones, taking the place of those that have moved away, while the others continue to form a decreasing number of horizontal layers. The result is that the bundles become wider and flatter. This is shown in cross-section in figures 61 and 62 and in surface view in figure65. The first of these figures is an oblique section through the anterior region of a spermatosome at about the stage represented by figure 64. The section represents a cross-section just behind the centrosomal plate of another spermatosome of the same age. Figure 65 is a longitudinal section through a spermatosome which was almost identical with the one repre- sented by figure 64. In this last figure the myoneme-like striations are also shown in surface view. As is shown here, the same force (that is, the continued growth of the axial fibers without a fully compensating increase in the length of the cell-body) which causes the secondary bundles to reach the surface of the cell and then to extend beyond it, eventually causes several folds to occur throughout their length. This is also indicated in figure 61. After the secondary bundles have reached the cell-membrane, the space which has been vacated by them is filled with a substance whose consistency is greater than that of the cytoplasm surrounding the albuminous bodies (fig. 53). Indeed, there seems to be a general flowing of the denser, more granular ctyoplasm to the center of the cell, leaving the homogeneous enchylema to fill the spaces between the albuminous bodies (figs. 59 and 60). Eventually this denser part of the cytoplasm becomes differentiated in such a way that it forms a fibrillar core passing down through the center of the cell (fig. 64). The changes in the shape of the spermatosome, as a result of the splitting of the bundle of axial fibers and the subsequent growth of these fibers themselves, are shown in figures 47, 49, 63, and 64. At first, accompanying the spreading of the secondary bundles to the surface of the cell, there is a slight elongation of the spermatosome in an anterior direction (fig. 49). This is continued until the secondary bundles begin to protrude beyond the cell-body, when it practically ceases (fig. 63). It will be noticed that the centrosomal plate has been pushed right up to the very anterior tip of the spermatosome. The further growth of the fibers results next in the lateral extension of the secondary bundles beyond the cell-body throughout their entire length and then in the elongation of the spermatosome in a posterior direction (fig. 64). This fact is evidenced not only by the change in the contour of the posterior end of the spermatosome, but also by the redistribution of the mitochondria. The base of the cell has become pointed and the flagella have fused to form the extreme end of the tail-piece. About Papers from the Marine Biological Laboratory at Tortugas. 229 this time, too, the folds that were mentioned before are formed in the second- ary bundles. A spermatosome which is in the stage represented by figures 64 and 65 is now ready to undergo the final change which transforms it into the fully formed apyrene spermatozoon (fig. 6). It has been completely filled with the albuminous bodies and all the chromatin has been dissolved in the cytoplasm; the secondary bundles have come to form folded ridges on both sides of the cell, already indicating the final shape of the undulating mem- branes. The length of the spermatosome, however, exclusive of the flagella is only about 70 micra. A general lengthening and constriction of the cell now take place which result in its becoming spindle-shaped, with an increase of from 15 to 20 micra in its length. At the same time the folds of the secondary bundles are somewhat straightened out and they them- selves become much broader and flatter, assuming the shape and appearance of typical undulating membranes. The fibers are now probably arranged into two outer layers where they are in close juxtaposition to one another, if not actually fused, and a single inner layer where they are more widely separated from each other. The base of the cell is drawn out narrowly to a point, ending with the fused flagella. It is quite possible that the final change in the shape of the cell is to be attributed to the continued growth of the fibers. DISCUSSION. There are two outstanding differences between the development of the apyrene spermatozoa of Strombus and that of the oligopyrene spermatozoa of Paludina. The first of these is in regard to their origin. It has been shown by Meves (’03) that the oligopyrene spermatozoa of Paludina arise from the spermatogonia, certain ones of which undergo an unusual growth and become the oligopyrene spermatocytes. In Strombus the apyrene spermatozoa arise from certain cells which are different from the spermato- gonia and which have been called the apyrene spermatoblasts. Secondly, in their later development, the apyrene spermatozoa of Strombus do not show nearly as close a parallelism to the development of the eupyrene as do the oligopyrene spermatozoa of Paludina. In both forms the develop- ment of the bundle of axial fibers in the atypical spermatozoon can be closely homologized with the growth of the axial fiber in the true spermato- zoon, but the two divisions of the oligopyrene spermatocyte which take place in Paludina are entitely lacking in the apyrene spermatoblast of Strombus. Again, in the adult oligopyrene spermatozoon of Paludina the equivalent of one chromosome is retained and forms its nucleus, while the apyrene spermatozoon of Strombus contains no chromatin at all. Finally, it might be pointed out that the oligopyrene spermatozoon of Paludina shows a much closer approximation to the form of the eupyrene than does the apyrene spermatozoon of Strombus. . On the other hand, there exist certain fundamental similarities of development and structure between the atypical spermatozoa of Strombus 230 Papers from the Marine Biological Laboratory at Tortugas. and Paludina and indeed of other forms in which they have been described. The first of these is the relatively immense growth of the cell which ulti- mately gives rise to the atypical spermatozoon. Second, the history of the centrosome and the development of the axial fibers of the atypical spermatozoon present close parallels in Strombus and Paludina. The centrosome fragments and gives rise to a number of centrioles; the division of these with the subsequent separation of the daughter centrioles results in the formation of the bundle of axial fibers. In Strombus the differenti- ation is carried further, in that the bundle of axial fibers splits and the two halves move laterally to the surface of the cell and there form the two undu- lating membranes. Very similar phenomena have been described by Stephan (’03b) and by Lams (’0g) for Murex and by Kuschakewitsch (’11 and ’13) for Vermetus. In fact, in every form in which the development of the atypical spermatozoa has been described we find the differentiation of the centrosomal element into axial fibers. Perhaps the most generally constant phenomenon in the development of the atypical spermatozoa of different forms is the fate of the chromatin. In Paludina, according to Meves (’03), one chromosome is retained and forms the nucleus of the adult oligopyrene spermatozoon, while all the rest of the chromatin degenerates. A similar condition is held by Lams (’o9) to be the case in Murex, although Stephan (’03)) had previously reported that in this form all of the chromatin degenerates. In all other forms in which the matter has been investigated, the nucleus breaks down and an atypical spermatozoon is developed which contains no chromatin. Another characteristic held in common by the atypical spermatozoa of certain forms is the fact that the products of metabolism are stored in the cell. Thus, we find large secreted bodies in the atypical spermatozoa of Strombus, Pteroceras (Brock, ’87), and Vermetus (Kuschakewitsch, ’11 and ’13). It is probable, tco, that the mitochondrial apparatus, so beautifully described and figured by Retzius (’05), in the oligopyrene spermatozoa of Paludina is of the same nature, as well as the refractive bodies described in the atypical spermatozoa of Cyprea by Brock (’87), of Murex by Koehler (’88), and other writers and the vacuoles in the apyrene spermatozoa of Conus described by Schiemenz (’96) and Kuschakewitsch (’11 and ’13). Both Stephan (’03a@) and Kuschakewitsch (’11) have suggested the possibility of arranging the atypical spermatozoa of different forms into a gradated series based upon differences in their development and structure and representing further and further retrogressions from the typical sper- matozoa. At the head of such a list would be placed Paludina and Murex in the order named. Next would come Conus, for here Kuschakewitsch (13) has recently found that one division normally occurs in the develop- ment of the apyrene spermatozoon. After these there might follow in succession Cerithium, Vermetus, and Strombus. ‘The list can not safely be extended further than this, since we are not sufficiently acquainted with the phenomena which take place in other forms; and even as it stands the list Papers from the Marine Biological Laboratory at Tortugas. 231 is full of contradictions. Thus, for instance, Cerithiwm should perhaps precede Conus, since in the former one chromosome is retained in the apyrene spermatozoon for a longer time than the others (Stephan, ’03@), while in Conus all the chromatin of the apyrene spermatocyte degenerates simul- taneously. Again, when considered from the standpoint of the differ- entiations of the centrosome and the development of a motor apparatus, Conus should be placed at the end of the list, for here the apyrene spermat- ozoa are totally immotile. Vermetus has been placed before Strombus because of the formation of chromosomes and of an abortive spindle in the development of the apyrene spermatozoa of the former; but the order could well be reversed when it is borne in mind that, in the development of the bundle of axial fibers of the atypical spermatozoon, Strombus resembles Paludina much more than does Vermetus. The truth of the matter is that we do not have sufficient data upon which to formulate such a series. In the writer’s opinion, the evidence is not clear enough for us to draw correct inferences as to whether or not we are dealing with a series of retrogressions from the atypical spermatozoa. While it has been shown that the atypical spermatozoa, in all those forms in which they have been studied, show certain fundamental similarities, each one also has certain differences peculiar to itself. One of the greatest of these differences, perhaps, is that in regard to the origin of the apyrene spermatozoa of Strombus. Nevertheless, in spite of such differences, it is the writer’s opinion that they are all essentially the same and are derived from similar elements of the testis—that is, they are modifications of the accessory cells of the testis, just as are the nurse-cells in the testis of Lit- torina. It is suggested that the explanation of the differences between them is to be found in the manner in which the differentiation of the primi- tive cells of the sexual gland takes place into the various elements of the testis. So, in the Pulmonates there is a sharp differentiation of the primi- tive sex-cells into the true sex-cells and into the accessory cells, the basal nuclei with their surrounding syncytium. Here the latter never become differentiated into independent cells, although they undergo changes in their size and shape (Platner, ’85). Among the Prosobranchs, in Littorina, the writer has never observed the same tremendous swelling of the basal nuclei as may be seen in the ovotestis of Planorbis, but, instead, certain of them give rise directly to the nurse-cells to which the spermatozoa are attached. In Strombus those elements which give rise to the atypical spermatozoa are differentiated from the true sex-cells along with the basal nuclei. The latter never give rise to nurse-cells of any kind. In forms such as Paludina and Murex, the elements which give rise to the atypical sper- matozoa are not differentiated from the true sexual elements along with the basal nuclei, but only after the spermatogonia have been formed. In Paludina the formation of nurse-cells has never been observed by the writer nor have they been described by Meves, but the basal nuclei undergo changes which are very similar to those which occur in the ovotestis of 232 Papers from the Marine Biological Laboratory at Tortugas. Planorbis. In Vermetus and in Conus, too, according to Kuschakewitsch (’13), the atypical spermatozoa are not differentiated until after the forma- tion of the spermatogonia. But in both these forms nurse-cells are also developed and apparently from the basal nuclei; whether or not their development is the same as that of the nurse-cells in Littorina is not known by the writer. In all these forms there is a period in the development of the testis when the true sex-cells are undifferentiated from the accessory elements. It may be that certain differences in the plasma of the primitive cells lead to differences in the manner of differentiation of the accessory elements and these in turn lead to the production of adult structures which differ from each other to a greater or less degree, but all of which show certain fundamental resemblances. Besides the characteristics held in common between the atypical sper- matozoa of Strombus and those of other forms, the former show certain resemblances to the nurse-cells in the testis of Vermetus and of Littorina. While there are no large cells in the testis of Strombus which could possibly be taken for anything except the apyrene spermatoblasts, there exist a striking resemblance between certain of the developmental stages of these and some of the figures given by Kuschakewitsch (’13) for the nurse-cells in the testis of Vermetus (figs. 104 and 105, Taf. xxiv, and figs. 169 and 170, Taf. xxv, Arch. f. Zellforsch., Bd. x). The writer has shown that nurse- cells are developed in the testis of various species of Littorina which, when they have reached their maximum size, resemble the oligopyrene spermato- cytes of Paludina. Later on, the cells become filled with large secreted bodies, while their nuclei gradually degenerate. In the final stages the latter may still be seen in a very shrunken, degenerated condition lying just beneath the cell-membrane, It is in connection with the formation of these secreted bodies and the corresponding changes in the nucleus that the apyrene spermatozoa of Strombus resemble the nurse-cells of Littorina. It will be remembered that in the apyrene spermatoblasts, after the nucleus breaks down, chromatic vesicles are formed which gradually increase in size until the formation of the albuminous bodies is begun. From this point on, with the continued formation of the albuminous bodies, they gradually degenerate and are finally all absorbed in the cytoplasm. A comparison of the processes which occur on these cells and those which take place in normal secretory cells lead to some interesting consider- ations. Mathews (’99) has shown that secretory activity or, to use his term, hylogenesis, is accompanied in various epithelial cells by certain changes in the nucleus, in that the latter becomes irregular in outline and is displaced towards the base of the cell. In these cells the process is cyclic; after the cell has been exhausted the nucleus returns to its normal condition and the activity begins over again. The changes which the nurse-cells of Littorina undergo at the commencement of their secretory activity are very similar to those which take place in secretory epithelial cells. In the former case, however, the process is not repeated and in every instance it is ac- Papers from the Marine Biological Laboratory at Tortugas. 233 companied by the gradual degeneration of the nucleus. It seems to be not unlikely that these two facts stand in a causal relation to each other. The same condition, that is, the unrepeated secretory activity of the cell accompanied by the degeneration of the nucleus, holds true in the case of the apyrene spermatozoa of Strombus as well as in the nurse-cells of Littorina. But here the question might well be asked, why does not the nucleus degenerate directly, instead of undergoing the complicated history that has been described? The primary function of the nucleus is concerned first with the growth of the spermatoblast and then with the secretion of the albuminous bodies, and possibly its presence is essential for the develop- ment of the great centrosome with its many centrioles. The differenti- ations which occur in connection with the centrioles lead to the formation of the undulating membranes and they are initiated by the disappearance of the centrosome. It is believed that the impulse, whatever it may be, which is responsible for the disappearance of the centrosome causes also the dissolution of the nuclear membrane. That impulse may be the same as that which causes a cell to divide mitotically, but, as has been shown, there is no real evidence to support such an assumption. Stated a little differ- ently, when considered in regard to the subsequent development of the apyrene spermatosome, it is the change that occurs to the centrosome that is of prime importance at this time and that which occurs to the nucleus is incidental to it. Were this not so, we should expect either that the chromatic fragments would degenerate without first becoming vesiculated and the secretion of the albuminous bodies be commenced at the same time, as is the case in the apyrene spermatozoa of Vermetus,! or else that the nucleus would remain intact until the secretory activity of the cell is begun and then degenerate as a whole, which is the case in the nurse-cells of Littorina. Instead, the karyomerites become vesiculated and form a great number of secondary nuclei which increase in size until the first albuminous bodies are formed, when they begin to degenerate. Moreover, if the secretory activity of the cell is enhanced by an increase in the surface contact between the nucleus and the cytoplasm, then the formation of the secondary nuclei is of great advantage. The conditions observed in the seminal receptacle and uterus of Strombus bituberculatus and S. gigas after copulation and at a time before oviposition has commenced, and indeed before the ova are fully matured, throw a certain amount of light upon the nature of their function. The fact that eupyrene spermatozoa are found in those parts at such a time indicates that they, or at least some of them, must remain there for a protracted period— that is, until and during the time of oviposition. The greater part of that time is no doubt spent in the seminal receptacle, where they can and do receive an independent supply of nourishment. But, judging from the 1 In this connection Kuschakewitsch (’13) has made the suggestion that in Vermetus the retention of the chromatic elements as such almost until the end of spermatogenesis is to be explained on the grounds that the chromatin participates in the formation of the albuminous bodies in the protoplasm of the apyrene spermatozoon, 234 Papers from the Marine Biological Labora:ory at Tortugas. fact that in nearly every case examined they have also been found in the uterus along with the apyrene spermatozoa, it must take a long time for them to move into the seminal receptacle. In the uterus a separation of the two kinds of spermatozoa occurs, the eupyrene being retained in good condition while the apyrene may be seen in all conditions of katabolic changes leading to the total exhaustion of the secreted bodies. It may be that the apyrene spermatozoa serve to nourish the eupyrene until the latter are stored in the seminal receptacle, where they may obtain an adequate and independent food supply. Whether this is so or not, it is believed that the apyrene spermatozoa of Strombus subserve their purpose in the uterus and possibly also in the vagina and bursa seminalis immediately after copu- lation and that their function is connected with their breaking down. There are two other ways of explaining the behavior of the apyrene spermatozoa after copulation has taken place. The first of these was suggested to the writer by Professor Conklin and is to the effect that the apyrene spermatozoa in breaking down may liberate a substance which stimulates the eupyrene spermatozoa or the eggs or both during fertilization. The possibility of their being an aid in the final disposition of the eupyrene spermatozoa was indicated by a series of experiments recently performed by the writer. It may be that the substance liberated by the apyrene sper- matozoa as a result of the changes which they undergo causes the eupyrene spermatozoa to withdraw from them and eventually to enter the seminal receptacle. The fact that the apyrene spermatozoa of Strombus break down after copulation and are discarded, while the eupyrene are retained and kept in a healthy condition, together with the results obtained by Kuschakewitsch (’10) in his study of the fertilization of the eggs in A porrhais pes pelicani— these facts, in the writer’s opinion, preclude the view that the atypical spermatozoa may participate as such in the fertilization of the egg. 1 While at the Tortugas Laboratory during June 1013, cultures of the contents of the sperm-duct of Sirombus gigas were kept alive and observed under varying conditions. The spermatozoa lived longest in cultures made from sterilized sea-water, in some cases as long as 72 hours. The length of life of the sper- matozoa apparently depended upon the extent to which the cultures had been kept free from bacterial infection. In such cultures, there was always a separation of the two kinds of spermatozoa, the eupyrene remaining active for a long time after the apyrene had ceased their movements. In very many instances a clear drop of some substance could be seen attached to the side of an apyrene spermatozoon whose contents had partially or com- pletely broken down. Very striking results were obtained in a few cultures in which a small piece of decaying tissue was included in order to test the reaction of the two kinds of spermatozoa to COz. In every case the eupyrene spermatozoa moved further and further away from the source of the CO: until they reached the edges of the cover glass. The apyrene spermatozoa, on the other hand, were not greatly stimulated except in one instance, where they gave a distinct, positive reaction, gathering around the pieces of decaying tissue. One effect of COz upon the apyrene spermatozoa was to cause a comparatively rapid and complete breakdown of their contents. Unfortunately, for lack of time, these experiments could not be repeated under conditions of greater accuracy, Papers from the Marine Biological Laboratory at Tortugas. 235 SUMMARY. 1. The apyrene spermatozoa of Strombus, in the adult condition, are devoid of any nuclear material. The body of the spermatozoon is spindle- shaped and is filled with large hexagonal albuminous bodies. Locomotion is brought about by means of the undulations of two broad membranes attached to each side of the body. 2. Together with the eupyrene spermatozoa, they are found in the uterus of the female, where they undergo certain katabolic changes resulting in the total exhaustion of the albuminous bodies. They are never found in the seminal receptacle, even when the latter is crowded with eupyrene spermatozoa. 3. The differentiation between the eupyrene and the apyrene spermato- zoa can be traced back further than the growth period. In fact, apyrene spermatozoa develop from cells which can be distinguished from the sper- matogonia of the first generation and which have been named the apyrene spermatoblasts. 4. The apyrene spermatoblasts are believed to arise from the basal nuclei. Confirmation of this phenomenon in other related forms is required before it can be considered an established fact. 5. The very early appearance of a centriole in the apyrene spermato- blast establishes the polarity of the apyrene spermatozoon. 6. The growth of the apyrene spermatoblast is marked by a great in- crease in the volume of the cytoplasm and nucleus and by the development, from the original centriole, of a very large centrosome containing many secondary centrioles at its periphery. No division of any kind takes place during the development of the spermatoblast. 7. After the apyrene spermatoblast has acquired its full growth, the dissolution of the nuclear membrane and the disappearance of the centro- some take place simultaneously. The chromatic fragments (karyomerites) are scattered throughout the distal half of the cell and the centrioles become grouped beneath the membrane at the base of the cell. From each of the latter a flagellum grows out. The cell may now be spoken of as the apyrene spermatosome, for the ensuing changes give rise directly to the adult apyrene spermatozoon without any appreciable increase in its volume. 8. The centrioles divide and the distal halves move across the cell in the direction of the chief axis, forming a bundle of axial fibers between themselves and the proximal halves. At the same time the karyomerites become vesiculated and remain active for a time, gradually becoming larger. 9. The mitochondria which have been formed around the centrosome during the growth period now constitute the inner margin of a ring of dense cytoplasm which lies around the bundle of axial fibers at its base. 10. When the bundle of axial fibers has come to extend entirely across the cell, the secretion of the albuminous bodies is begun in the anterior por- tion of the spermatosome. This process continues hand in hand with the degeneration of the chromatic vesicles, until, successively, the middle and 236 Papers from the Marine Biological Laboratory at Tortugas. posterior portions of the cell have been completely filled with them and all of the chromatic vesicles have disappeared. 11. In the meantime, the bundle of axial fibers continue to grow until the cell has been elongated in an anterior direction, along its chief axis, by about one-third its original diameter. Afterwards, the continued growth of the axial fibers is not accompanied by a fully compensating increase in the length of the cell; the changes which ensue are all the direct result of this fact. 12. A longitudinal split occurs in the bundle of axial fibers, dividing it into two equal halves, each of which moves in an opposite direction towards the surface of the cell. In doing so they rupture the ring of dense cytoplasm which lie around the base of the bundle. This substance gradually disap- pears, but the mitochondria associated with it persist in the adult apyrene spermatozoon. 13. Having reached the surface of the cell, the secondary bundles of axial fibers now come to extend laterally beyond it. The further growth of the fibers causes a number of folds to occur throughout the length of the bundles. Finally a general elongation and constriction of the cell occur, with the result that the body of the spermatosome acquires the spindle-like shape of that of the adult apyrene spermatosome and the secondary bundles on both sides of it become straighter and more flattened, forming the undulating membranes. 14. In the writer’s opinion the apyrene spermatozoa of Strombus sub- serve their purpose after copulation has taken place and their function is connected with the katabolic changes which they undergo. Their behavior may be explained on the grounds of the following three suggested possi- bilities: (a) They may serve as nurse-cells to the eupyrene spermatozoa after copulation and before the latter reach the seminal receptacle; (b) they may, by the liberation of some substance, stimulate the eupyrene sperma- tozoa or the eggs or both during fertilization; or, by the liberation of some substance to which the eupyrene spermatozoa are negatively chemotactic, they may act as an aid in the final disposition of the latter. 15. The conditions met with in the uterus and seminal receptacle of the female after copulation has taken place and before oviposition has com- menced show that the apyrene spermatozoa do not participate as such in the fertilization of the egg. BIBLIOGRAPHY. AUERBACH, L. 1896. Untersuchungen iiber die Spermatogenese von Paludina vivipara. Jenaische Zeitsch. f. Naturw., Bd. xxx, N. F. xxii. Brock. 1887. Ueber die doppelten Spermatozoen einiger exotischer Prosobranchier. Zool. Jahrb., Bd. 11. v. Brunn, M. 1884a. Untersuchungen iiber die doppelte form der Samenké6rper von Paludina vivipara. Arch. f. mikr. Anat., Bd. xxtt. 18845. Weitere Funde von zweierlei Samenk6rper in demselben Tier. Zool. Anz., Bd. vi. Papers from the Marine Biological Laboratory at Tortugas. 237 DUVAL. 1880. Etudes sur la spermatogénése chez la Paludine vivipare. Journ. de Micrographie, T. 1v, Nos. 8-9, 10-II. v. ERLANGER, R. 1897. Bemerkungen iiber die wurmf6rmigen Spermatozoen von Paludina vivipara. Anat. Anz., Bd. xiv. HERTWIG, R. 1905. Ueber das Problem der Sexuellen Differenzierung. Verh. deutsch. Zool. Ges., 15. Jahresversamml., 1905. KOEHLER. 1888. Recherches sur la double forme des spermatozoides chez le Murex. Recueil zool. Suisse, T. v, 1892. KOLLIKER. 1841. Beitrage zur Kenntnis der Geschlechts-verhaltnisse und der Samenfliissigkeit wirbellose Thiere nebst einem Versuch iiber das Wesen und die Bedeutung der sog. Samenthiere. Berlin. 1847. Die Bildung der Samenfaden in Blaschen als allgem. Entwicklungsgesetz. Neue Denkschriften der Allg. Schweitzer Ges. f. d. ges. Naturw., Bd. viii. KUSCHAKEWITSCH, S. 1910. Zur Kenntnis der sogenannten ‘‘wurmférmigen” Spermien der Prosobranchier. Anat. Anz., Bd. XxXxXvII. 1911. Ueber die Entwickelung der Spermien bei Conus mediterraneus Brug. und Ver- metus gigas Biv. Biol. Centralblatt, Bd. xxx1, Nr. 16 u. 17. 1913. Studien iiber den Dimorphismus der mannlichen Geschlechtselemente bei den Prosobranchia. I. Arch. f. Zellforsch., Bd. x. Lams. 1909. Recherches concernant le dimorphisme des éléments séminaux chez le Murex Annales Soc. Médec. Gand, T. LXxxIx. LEyDIG, F. 1890. Ueber Paludina vivipara. Zeitschr. f. wiss. Zool., Bd. 11. MatTHEWS, ALBERT. 1899. The Changes in the Structure of the Pancreas Cell. Journ. of Morph., vol. xv. MEVES, FR. 1900. Ueber den von v. la Velatte St. George entdeckten Nebenkern (Mitochondrien- kérper) der Samenzellen. Arch. f. mikr. Anat., Bd. LvI. 1903. Ueber oligopyrene und apyrene Spermien und iiber ihre Entstehung nach Beobacht- ungen an Paludina und Pygera. Arch. f. mikr. Anat., Bd. LXxI. PAASCH. 1843. Ueber das Geschlechtssystem und _ tiber die harnbereit. Organe einiger Zwitters- necken. Wiegmann’s Arch. f. Naturgeschichte, Bd. 1. PLATNER, G. 1885. Ueber die spermatogenese der Pulmaten. Arch. f. mikr. Anat., Bd. xxv. 1889. Beitrage zur Kenntnis der Zelle und ihrer Teilungserscheinungen. II. Samen- bildung und Zellteilung bei Paludina vivipara und Helis pomatia. Arch. f. mikr. Anat., Bd. XXXII. Pororr, M. 1907. Eibildung bei Paludina vivipara und Chromidien bei Paludinaund Helix. Arch. f. mikr. Anat., Bd. Lxx. REINKE, E. E. 1912. A Preliminary Account of the Development of the Apyrene Spermatozoa in Strombus and of the Nurse-cells in Littorina. Biol. Bull., vol. xx11. REtTz1us, G. 1905. Zur Kenntnis der Spermien der Evertebraten. II. Biol. Unters., N. F., Bd. x11. 1906. Die Spermien der Gasteropoden. Biol. Unters., N. F., Bd. xim. SCHIEMENZ. 1896. Beitrage zur Kenntnis der Coniden. R. Bergh. Nova Acta Acad. Caes, Leop.- Carol. Germ. Nat. Curios., Bd. Lxv, S. 95. v. SIEBOLD. 1837. Fernere Beobachtungen iiber die Spermatozoen der wirbellosen Tiere. 2, Die Spermatozoen der Paludina vivipara. Miiller’s Arch. f. Anat., Physiol. und wiss. Med., Jahrg. 1836. STEPHAN, P. 1903a. Le développement des spermies oligopyrénes et apyrénes de quelques Proso- branches. C. R. Soc. Biol. Paris, T. Lv. 1903b. Le développement des spermies apyrénes de Murex brandaris. C. R. Soc. Biol. Parise. Gy. 1903c. Le développement des spermies apyrénes de Cerithium vulgatum et de Nassa mutabilis. Bibliogr. Anat., T. XII. EXPLANATION OF FIGURES. All figures were drawn at table height with the aid of a camera lucida. With the exception of those on plate I and figs. 63-65, all figures were drawn with the use of the Zeiss 1.5 mm. apochromatic objective and the 12 compensating ocular; the figures on plate I were drawn with the use of the Zeiss 2 mm. apochromatic objective (N.A. 1.30) and the 6 com- pensating ocular and were then enlarged to 1825 diameters, while figs. 63-65 were drawn with the use of the Zeiss 1.5 mm. apochromatic objective and the 8 compensating ocular. In studying the cells the Zeiss 1.5 mm. apochromatic objective as well as the 2 mm. (N.A. 1.40) were used in combination with the 8 compensating ocular. Figs. 8-65 were drawn from sections, 5 micra thick, of material fixed in Flemming’s fluid (stronger solution) and stained with iron hematoxylin followed by gentian violet and erythrosin. PLATE. i: (Initial magnification 1825 diameters, not reduced. g by) 1 to 5. Three progressive stages in the development of the apyrene spermatosome, drawn from living material. Figs. 1 and 2 show corresponding stages viewed in planes at right angles to each other; figs. 3 and 4 are similar views of a later stage. 6. Adult apyrene spermatozoon, drawn from a specimen fixed with osmic vapor and stained with iron hematoxylin. 7. Adult eupyrene spermatozoon, drawn from a specimen fixed with osmic vapor and stained with iron hematoxylin. PLATE 2, (Initial magnification 3450 diameters, reduced one-third.) 8 to 14. Various stages in the growth of the apyrene spermatoblast from the earliest appearance of the centrosomal structure until a time just before it disappears and the nuclear membrane is dissolved. 15 and 16. Two consecutive sections through a single cell showing the initiation of the changes which occur to the nucleus and the centrosome. The mitochondria are shown lying around the centrosome (fig. 15). 17 and 18. Two consecutive sections through a single spermatoblast in the same stage as the preceding one. Note that the centrosome has disappeared. 19 and 20. Two consecutive sections through a single spermatoblast showing the scattering of the chromatic masses after the dissolution of the nuclear membrane. PEATE 3: (Initial magnification 3450 diameters, reduced one-third.) 2I, 22, and 23. Three consecutive sections through a single spermatoblast showing the appearance of groups of centrioles at the periphery of the cell. 24 and 25. Two consecutive sections through a single spermatoblast showing the completed fragmentation of the chromatic masses to form the ultimate karyomerites. 26 and 27. Twoconsecutive sections through a single spermatoblast showing the appearance of the centrioles on the periphery of the originally proximal half of the spermato- blast and the vesiculation of the karyomerites. 28 and 29. Two consecutive sections through a single spermatoblast, showing the grouping of the centrioles and the growth of the flagella. 30, 31, and 32. Three consecutive sections through a single spermatoblast of the same age as the preceding one, showing the rays which extend from the centrioles into the interior of the cell. PLATE 4. (Initial magnification 3450 diameters, reduced one-third.) 33. A section through a very young spermatosome. The centrioles have gathered at the base of the cell but have not yet divided. 34. A section through a spermatosome in the stage in which the centrioles are dividing. 35- A section through a spermatosome after the division of the centrioles showing the manner in which the axial fibers are formed. 238 Papers from the Marine Biological Laboratory at Toriugas. 239 36 and 39. Sections through two spermatosomes showing consecutive stages in the growth of the axial fibers. 37 and 38. Two consecutive sections through the bundle of axial fibers of a spermatosome in a stage corresponding to that shown in figure 39. Note the almost round base and the flattened anterior portion of the bundle and the mitochondrial apparatus surrounding the latter. 40 and 41. Sections through two spermatosomes of approximately the same age; the sec- tions pass in planes which are at right angles to each other. Figure 41 gives a surface view of the bundle of axial fibers and the mitochondria surrounding it; note the increase in the size of the chromatic vesicles shown here. 42. A cross-section through the posterior region of a spermatosome which is a little older than the preceding ones. Note how closely the mitochondria are applied to the bundle of axial fibers. PLATE «be (Initial magnification 3450 diameters, reduced one-third.) 43. A secuoe through a spermatosome showing the further growth of the bundle of axial ers. 44 and 45. Sections through two spermatosomes showing the bundle of axial fibers extending across the cell and the early formation of the secreted bodies; the sections pass in planes which are at right angles to each other. 46. A section through a spermatosome showing the continued growth of the bundle of axial fibers until it extends beyond the cell. 47 and 48. Sections through two spermatosomes of slightly different ages showing the splitting of the bundle of axial fibers; the sections pass in planes which are at right angles to each other. PLATE 6. (Initial magnification 3450 diameters, reduced one-third.) 49. A longitudinal section through a spermatosome in a well-advanced stage. The splitting of the bundle of axial fibers has been completed and the cell is partially filled with the secreted bodies. The nuclear vesicles are rapidly degenerating. 50 to 58. Consecutive cross-sections through a single spermatosome a little younger than that shown in figure 49. The sections are taken in order from the posterior to the anterior end of the spermatosome and show the modifications of the two halves of the bundle of axial fibers as they reach the cell-wall on each side. 59 and 60. Cross-sections through the posterior and anterior halves of two spermatosomes of the same age showing some of the axial fibers moving out beneath the cell-wall. PLATE. 7: (Initial magnification 2450 diameters, reduced one-third.) 63. A longitudinal section through a spermatosome in which the halves of the bundle of axial fibers have reached the cell-wall on both sides throughout their entire length. The secretion of the albuminous bodies has not yet been completed nor has all the chromatin disappeared. 64. A longitudinal section through a nearly adult apyrene spermatozoon showing the total disappearance of the chromatin and the cell filled with the secreted bodies. 65. A longitudinal section through an apyrene spermatozoon of the same age as that shown in figure 64 giving a surface view of one of the two undulating membranes and the longitudinal fibers beneath the cell-wall. sits Seer yale eta Fay! + i chery ae REINKE | ones Se i ee ht AS) 1 ee es Fa on = es PEP me spt 2 an ‘ar, oan" ¥ pa ey im “ue 7 i? REINKE 16 PEATE 2 REINKE PLATE 3 REINKE PLATE 4 bh] f i: cS Ls ieAteccel divs epee i 4 : j ; ‘i PLATE 5 REINKE PLATE 6 REINKE PLATE 7 REINKE XII. HISTORY OF THE SPOTTED EAGLE RAY, AETOBATUS NARINARI, TOGETHER WITH A STUDY OF ITS EXTERNAL STRUCTURES. BY EY Ws GUDGER, State Normal College, Greensboro, North Carolina. Ten plates and nineteen text-figures. 241 CONTENTS. The Spotted, Eagle Ray, Aciobaius nardnnes 50 550d «cing sea ce ale Maen oe a eee ka Oe 243 General Deneriptrious aie. <2 ash crores heuedets Ps = Sg 5-5 se ew tete SM wees ae tes) oe 244 SPECS DESCEND ELOMS ctor eevee creore ate tee nee. Pelee revan eid taseS ca le neh eel ae RE Rac oe eae ee 261 (olOr Amie IDES: 5 one wie Os aoe tie 6 oie ho nd oe ie wee mamas Oe ree 261 aim Peed SPINES: cok sighs Rath setametdee Ay dad cbs es oe ee ee ee ee Be 274 He gudyenoue ak. oie: RE SE LEL: CECPLE TOR ee Os COP AEE bee 280 Niaiwerandebeeth i rrcscud ane Sparen’ geass one suese /sy age arenes Saeko eae ee ts ro encore oe Rene 283 Dorsal sensory pore Syeneid 244 sos PAS bt ath At awh olka aba ae mane 294 RAGES eer nde Sak Ge ea rote gah do Pine Cale oe ate wale stabi A oe ences emer ECE 296 CEU ae ORARNES: cS vs Ae code saa SS anak ee ee ye Gee x tee Ren ene oe Reet ae ae eee 296 Olensine and defensive habits. 44.5 os Sg sh ae Se ceca Shah ae Sele dee Rae 298 Breeditio Gabits., 2 o-.s0.0:5 egy pease fe es oe eR oo asin eos oe lala eee 300 |: [2,0 <1 Pe fg ee Wren er PSE Ene rteas Pie dary het ae iy iG te) 303 PCONOMIC “Values Mite Rate SRR ma UTR Le cds eae, SVS atone Matetenersl ena ede oie aveea 305 Genera and’ Species of “Angle-toothedWRaysy cc.ci: Ks make-up. Near the tail it has two fins about the size of one’s hand, rounded in outline and of equal length. Its head, which is thick, compressed, and / furrowed in the middle, is about as large as that of a good-sized pig. The mouth rounded underneath is triangular, compressed a little, and terminates in a snout. The opening of the mouth is on the ventral sur- face, 5 inches from the end of the snout. The mouth is 2)4 inches wide, toothless, but having in the place of teeth a lower jaw in the shape of a tongue. This is 4 inches long, 114 inches wide, and reaches to the external opening of the mouth. Likewise, there is an upper jaw placed crosswise, 2 inches long and as many wide. The lower jaw consists of 17 hard white bones having the shape of the letter U and firmly joined to the membranes. Underneath there lie TEXT-FIG. 1.—Narinari, from 17 other bones, one under each, of spongy ap- Marcgrave. pearance but not so hard. The upper jaw con- sists of 14 bones, shaped like the letter J and also joined together by membranes. Likewise 14 other bones lie above these. Moreover the two jaws are joined to the other bones of the head by membranes [cartilages ?]. The cavity of the skull, wherein the brain lies, is about 6 inches long and hardly 2 wide. The snout is wholly cartilaginous. The fish has two small eyes about the size of a nummus misnicus. Behind these eyes, on each side, is a large breathing-hole capable of holding an apple of ordinary size. Within these holes the leaves of the gills lie hidden. On the lower side at the [hinder] end of the head are five oblong incisions. The whole upper surface of the body is of a steel (ferreus) color with white spots the size of a nummus misnicus scattered over it, while the under part is entirely white. The skin is everywhere smooth and without scales. The length of the body from the end of the snout to the root of the tail is 114 feet; the width between the extremities of the triangular wings is 3 feet 10 inches. The length of the fins near the tail is 7 inches, the width 4. The length of the head is 10 inches, the 9 ————= = ee) = = 0: The Spotted Eagle Ray. 247 width 7, and it is 114 feet thick. The tail is 4 feet 3 inches long and its thickness at the beginning is 5 inches, but it gradually becomes thinner. A little behind the beginning of the tail, there is a small, short fin, a little more than an inch long; and just behind this, standing erect, are two little hooks, curved like fish-hooks and 3 inches long. Its flesh has a good flavor and is sufficient to feed 40 men. This description needs little or no comment here. It is excellent and would enable one to identify the fish to-day. The statement that the head of the ray is 1% feet thick, is, inasmuch as the length of the animal is only 18 inches, evidently an error (either editorial or typographical, since the work was published four years after Marcgrave’s death). So, likewise, is the statement that the tail is 5 inches thick at the root. Text-figure 1 is Marcgrave’s wood-cut made from the oil painting in the collection referred to. The figure in Marcgrave’s text has been colored by hand in one copy of the work found in the Library of Congress, but in the other copy, as in that owned by the Bureau of Fisheries and in the copy belonging to the present writer, the figures are uncolored. The upper surface in the colored figure is a dark steel-blue (ferreus Marcgrave calls it) dotted with white or bluish-white spots. Attention should here be called to the discrepancy between the description and the figure in the matter of spines. The former has them curved like fish-hooks, but the figure has them straight, with one large barb on each. As a matter of fact, they are of the ordinary ray type, straight, flat, and multibarbed on the edges. Figure 3, plate I, is a photograph of the water-color painting to which reference has been made. When it is recalled that this painting was made some 275 years ago, on the wild and inhospitable coast of Brazil, one hardly knows whether to marvel most at the accuracy of the work or the excellency of its preservation. If this figure be compared with text-figure 1, it will be seen how poorly the wood engraver has copied the original painting. Concerning these figures, see Gudger (1912, pages 265-272).!_ The marginal notes on the original painting are presumably those of Count Johann Moritz. Figure 4, plate 11, is a photograph of the oil painting in the larger collection described above. The deficiencies of this latter figure, when compared with the former or with my photograph (fig. 1, plate 1), are so marked as to call for no comment here. The older writers seem to have contented themselves with merely quoting Marcgrave and copying his figures. So did Piso himself, in his folio volume entitled ‘De Indiz Utriusque re Naturali et Medica,” etc., Amsterdam, 1658, though he abridged Marcgrave’s original description somewhat. However, he does one good deed in giving us a clue to the meaning of the Brazilian word Narinart. He tells us that the tail is armed with two stings “approaching in form the figure of an arrowhead, and therefore these fishes are properly called by us Pylstaert, and by the Brazilians Narinari.”” Now since the Dutch word Pylstaert means both a water bird and a sting ray, we might think that the allusion was either to the wing-like pectorals or to 1 For these photographs and for much information about these drawings, I am indebted to the courtesy of Dr. Perlbach, of the Royal Library of Berlin. 248 Papers from the Marine Biological Laboratory at Tortugas. the spined tail, had not Piso made it clear that the latteris meant. Itseems then that the word Narinari means sting ray (see Martius, 1867).} Willughby, in his ‘‘De Historia Piscium,” published after his death by Ray at Oxford in 1686, reproduces both figure and text with great exactness. The fish, however, is represented as if it were lying on its left and were drawn from the right side. The description and figure are credited to Marcgrave. The next reference to this interesting fish is to be found in Jonston’s “Historie Naturalis de Piscibus et Cetis.’”’ This work passed through several editions. The first, according to Walbaum, was published at Frankfort in 1649. Three editions are dated at Amsterdam in 1657, 1660, and 1718. In the Latin version in the Library of Congress, bearing date of 1767, Jonston reproduces Marcgrave’s figure, but reverses right and left sides, as does Willughby, and takes out some of the crudities of the drawing, and quotes Marcgrave’s description almost verbatim, giving him credit. Henry Ruysch, in 1718, published his ‘‘'Theatrum Universale Omnium Animalium, Piscium,” etc., which seems to be a compilation of the works of various writers from Aristotle to Marcgrave and Jonston, but mainly of the latter. Ruysch has rearranged Marcgrave’s description and figures, prob- ably following Jonston therein. He has copied Marcgrave’s description of Narinari in the minutest detail, but without a word of reference or acknow- ledgment. He reproduces Marcgrave’s figure, but has improved it, espe- cially in the shading. In 1697 Hans Sloane figured and described in volume x1x of the ‘‘ Phi- losophical Transactions’’ the jaws of a Pastinaca marina from Jamaica, which he says is identical with Marcgrave’s Narinari. In his brief descrip- tion of the fish he says that it is ‘‘smooth, blue; covered with white spots.” Later (1725), in volume 11 of his Natural History of Jamaica, he describes this fish more fully. This [whip ray] was about two foot over from corner to corner, and all blue even the flesh itself with white spots on it, the under side or Belly was white, as in others of this kind, the tail was six foot long, black, small and smooth, of which are made whips, whence the name Whip-Ray, beyond the Pinna at the end of the body or in the beginning of the tail lie one, two or three inch and a half long flat streight bones or Radij; they are white, serrated with Teeth on both Sides like a saw, made so as an arrow that’s barbed, to enter the flesh easily but not to come out without tearing it, they lie one on another on the upper part of the tail, where there is a hollow or cavity made to receive them like a sheath, that they may swim with less Impediment and only use them on Occasion. It is probable that Patrick Browne, in his ‘Civil and Natural History of Jamaica” (1756), refers to our ray when, on page 459, he describes a whip-ray thus: ‘Middle parts bluish mixed, tongue long, with a barbed spine on the finned tail.’”’ Ray (1713), who it will be remembered edited Willughby’s manuscript, in his ‘Synopsis Methodica'Piscium,” quotes Sloane and Marcgrave, but adds nothing new concerning our ray. I find no other reference to Aétobatus narinari until 1790. Excepting Abbeville, Marcgrave, and Sloane, and possibly Piso, none of the older writers above quoted seems to have ever seen the fish; they were all com- The Spotted Eagle Ray. 249 pilers. However, in the above-named year, a Swede, Bengt Anders Eu- phrasen, described and figured a specimen of this ray which had been taken at the island of St. Bartholomew, one of the West Indies at that time be- longing to Sweden. He calls this fish Raja narinari, thus fixing the specific name. He refers to Willughby, whose description and figure he criticizes, seemingly in ignorance of Marcgrave, from whom Willughby copied both figure and description. Euphrasen’s figure, which was drawn in Sweden from a preserved specimen, is herewith reproduced (fig. 5, plate m1), but modified slightly in one unessential detail. In the original figure, the long tail hangs down straight save for a fish-hook-like curve at the tip; in the reproduction, for the sake of economy in space, it is curled to the left as shown; the length, however, is correct. Euphrasen describes his specimen as follows: Head prominent, compressed, in form like a toad’s, on top 2 pits placed transversely. Mouth ventral as in its kind, transverse, teeth few and closely crowded. Eyes lateral, of medium size when compared with the head. Ears (aures) above the neck, or at the base of the head, with apertures round, and a little larger than the eyes. Spiracles 5, under the throat on both sides as in its kind. Body flat, very broad, ending laterally in acute angles, the margin feathered (pinnis). Above brown! (blue) with round white spots, in diameter about the size of the thumb, scattered about; beneath white. Anal fins one on each side at the base of the long tail, reaching to the base of the spine on the anterior part of the tail. Dorsal fin small, sub- triangular, situated on the base of the tail. Spines 2, oblique, dorsally placed behind the fin, compressed, retrorsally barbed, the posterior double the length of the anterior. Tail whiplash-like, compressed, attenuate from base to apex, without fin, 3 times longer than body. In Walbaum’s edition of Artedi’s “‘ Bibliotheca Ichthyologica,” part 111 (1792), the former quotes the name Raja narinari from Euphrasen’s paper above named, refers to Willughby and Marcgrave, and gives this brief description: ‘‘ Body smooth, steel colored above, with numerous white spots.”’ About this time, 7. e., the close of the seventeenth and the beginning of the eighteenth century, a number of ichthyologists include narinari in their lists of rays without adding anything to our knowledge of the fish and some of them fall into the grievous error of applying the name to rays which are not even eagle rays much less Aétobatids. Bloch in volume 111 of his “‘Ichthyologie” (1786) appends the name arinari to an ordinary sting ray with a long tail, and which he expressly states has many rows of teeth in its jaws. Other cases might be cited, but to no purpose. In the closing years of his life, Bloch had outlined and made some progress on a great ichthyological synopsis after the plan of Linnaeus’s “Systema Nature.’’ After his death, this important work was brought to completion by J. G. Schneider (Bloch and Schneider, 1801). Beginning on page 361 we find the description of Raja narinari. This I believe to be the work“of Schneider for two reasons; first because in Bloch’s other works there is nowhere found any description of A. narinari, and in the second place because the account under discussion is written in the first person. 1 Euphrasen uses the word chalybeum from chalybs, meaning steel or iron, but as his figure is a sepia- brown it may be that the color of his fish was brown. 250 Papers from the Marine Biological Laboratory at Tortugas. The description of this ray is interesting not because it contains any- thing new or unusual or erroneous, but because it is a curious jumble of Marcgrave, Willughby, Euphrasen, J. R. Foster (ms. ‘‘Descriptiones Ani- malium”’), Prince Maurice of Nassau, and Hans Sloane (‘‘History of Jamaica”). However, Schneider examined several ‘‘Swedish specimens,”’ and among them probably Euphrasen’s, while he also had several sets of jaws and a dried specimen cut in half. Omitting those parts of Schneider’s description, which will be taken up in the special sections, his general account is now given. He says that the head is compressed and prominent; above the apex there may be noted holes (spiracles) placed transversely ; the mouth is inferior and is filled with a few teeth crowded together; the laterally placed eyes are of medium size in round orbits but little larger than themselves; the body is broad and flat, the lateral angles are acute, the emarginate fins are steel-colored above with round white spots the size of the thumb, and the tail is three times longer than the body. Its habitat is given as the American Ocean, especially around the Caribbean Islands. In 1803, the East India Company published in London, in two sumptuous volumes, ‘‘ Descriptions and Figures of Two Hundred Fishes Collected at Vizagatapam on the Coast of Coromandel”’ by Patrick Russell, a physician in its employ. On page 5, there is described a ray called by the natives eel tenkee and by Russell the ocellated raja. Figure 6, plate 11, is a repro- duction of plate vi1t of the above work, showing this fish. That this is Aétobatus narinari, inspection of the figure and study of the description make clear. The general outline of the fish, the pointed pectorals with fimbriated hinder edges, the white spots profusely covering the body back of the shoulder region, the projecting head with its pointed snout, the relative position of the eyes and spiracles, the rounded ventrals with the dorsal placed well between them on the root of the tail, all spell narinari. When one reads that the tail is long and whip-like, and that the jaws are dissimilar, “the lower arched, narrow, and projecting beyond the wider immovable upper jaw,”’ one is quite sure that it is narinari despite the absence of spots on the anterior dorsal region and the fact that both jaws are described as devoid of teeth. These points will be considered later. Blainville, in the Journal de Physique, tome 83, for 1816, has a systematic paper in which he establishes the genus Aétobatus, assigns the generic characters, and notes ten species, of which narinari is one. Later, Blain- ville (1828) changed the generic name to Aéfobatis, though a careful search through his paper fails to reveal any reason therefor. His characters for the genus are as follows: Body together with the pectoral fins in the shape of a bird with extended wings; head free and provided with a simple appendage in front; eyes lateral; teeth large, smooth, polygonal, united into two plates, one lingual, the other palatine; pectoral fins pointed, anterior border convex, posterior concave; pelvic fins as in the sting rays, only one fin above on the root of a tail often very long, flagellate, armed with one or two spines. — The Spotted Eagle Ray. 251 This, it should be noted, is a statement of the characters of the genus only; narinari is not named and no specific description is given save for Raja aquila of European waters. In the celebrated voyage around the world of the corvette Uranie, a spotted sting ray was taken at Guam. The zoology of this voyage was written up by Quoy and Gaimard (1824), the ship’s surgeons. They described their ray as having an elongated snout, while its general color was a light brown sprinkled with round spots of cerulean blue, and its tail was armed with five very long, barbed spines. Their description is very imperfect and their drawing ‘‘made on the spot’’ was lost. They were much im- pressed by the fact that this ray had five spines on the tail and refer to it as such an extraordinary matter as to make the tail worthy of deposit in the Museum (of Paris?), and from this phenomenon gave to the ray the name Raja quinqueaculata, the 5-spined ray. A photographic reproduction of their fine figure of the spines is given in figure 7, plate Iv, of this paper. Below in the same figure is a photograph of a portion of the tail with four stings of A. narinari from Beaufort. In the section dealing with the tail and its spines, a careful comparison of the two tails will be made. In 1835, Ruppell described an eagle ray from the Red Sea under the name Myliobatis eeltenkee. WHowever, he has doubts about the identity, for he says that he is not able to say whether it is different from Linne’s “form called Raja narinari after Marcgrave’s description.’! In his syn- onymy, Ruppell says his fish is identical with Russell’s eel tenkee which Gunther says, and which we have just seen, is Aéfobatis narinari. That it is A. narinari can not be doubted when one reads of the long, pointed pectorals with fine notchings on the posterior edges (found on the ventrals also) the white spots covering the dorsal surface; the elevated head with lateral eyes; the long, pointed snout, and the mouth with its teeth in a single row in each jaw, the lower being angled outwardly. Some doubtful points are seven teeth only in each jaw (these will be considered at length later), the dark olive color above, and the flabelliform tail. Despite the fact that Swainson (1838-1839) spent some months on the northern coast of Brazil in 1816-1817, there is no evidence that he ever saw our ray. He quotes Russell, and also Miiller and Henle, whose great work was then appearing in parts, but adds nothing to our knowledge. In 1841 the celebrated Johannes Miiller, with the cooperation of Jacob Henle, published at Berlin in folio their ‘‘Systematische Beschreibung der Plagiostomen.”’ In it they definitely established the form A étobatis nari- nart based upon the examination of twelve specimens from Brazil, the Indies, and the Red Sea, six being in alcohol and six dry. The congeneric Raja flagellum of Bloch and Schneider, which it must be remembered is declared to be identical with A. narinari by Giinther and by Jordan and Evermann, is based on the examination of 19 specimens and one head. : 1In none of the several editions of Linne’s “‘Systema Nature,” which I have examined, is there any record of narinari. 252 Papers from the Marine Biological Laboratory at Tortugas. They describe the upper surfaces of their specimens of A. narinari as brown with round, white, regularly scattered spots which are lacking only on the head, and add that the number of the spots is often small. Both my photographs and Duméril’s figure (reproduced as fig. 8, plate 1v, of this paper and referred to elsewhere), show that the head does not lack for spots and that the number on the body is great. Curiously enough, however, Jordan and Evermann’s figure (text-fig. 3) shows very few spots in the cephalic region. The fish from which it was drawn, however, like the twelve examined by Miiller and Henle, was a preserved specimen. In the years 1772-1774, Johann Rheinold Forster collected fishes in the southern seas and later described his collection. His manuscript, however, remained unpublished, though noted by Bloch and Schneider (1801), until it was edited by Lichtenstein and published at Berlin in 1844. He collected both in Brazilian waters and in the South Pacific, and described and figured a fish which he called Raja edentula, but which is easily recognizable as Aétobatus narinart. According to his description, this ray is lead or steel colored above with many round white spots, white below, and everywhere smooth and bare; the head, slenderer than the body, terminates in a tri- angular blunt snout, cartilaginous in structure, and having the edges rolled up. The mouth is typically that of narinari; the crescent-shaped lower teeth form a spatula projecting beyond the shorter and broader upper jaw. The eyes are lateral, prominent, and rather far back and have vertical pupils. Behind the eyes are the spiracles, connected with each other and with the mouth. The tail is three times as long as the body, slender, cylin- drical, pinnate (dorsal?), black, and armed with two serrate spines. Pectoral and ventral fins are crenate behind. This is an admirable description of A. narinari, but unfortunately accompanied by no figure. Lichtenstein does not state why the figures, made to accompany the text, were not published. Cantor, in his ‘Catalogue of Malayan Fishes’”’ (1849), describes an East Indian Ray, Stoasodon narinari, which seems to be identical with our ray. His description need not detain us here longer than to note that his fish is greenish-olive or greenish-gray above, while the greenish-white spots edged with black are found on all parts of the dorsum save the head and anterior margin of the pectorals. These points will be taken up later. Bleeker (1852) declares that A. narinari is common throughout the East Indies. His description of this fish does not differ from that given by other authors save as to the color of the body and the structure of the jaws, consideration of which points will be entered into later. Although he gives no figure of his fish it is undoubtedly A. narinari. The most elegant figure of an eagle ray that has ever been published is that of August Duméril found in tome Io of the “Archives du Museum d’Histoire Naturelle” for 1858-61. Figure 8, plate Iv, is a photographic reproduction of plate 20 from the above volume. A is A. latirostris from the region of the Gaboon (river), B (lower left corner) is a head of A. nari- 1 This is the earliest note made of this peculiarity. The Spotted Eagle Ray. 253 nari, C (lower right corner) is A. fouet, (A. flagellum?). One observes at once the larger spots of A. latirostris, and the fact that they are few in the cephalic region and rather widely scattered over the other parts of the body, while in A. narinari they are much smaller, thickly scattered, so much so on the head as to appear concentrated. Worthy of note is the fact that the congeneric form A. fouet is entirely devoid of spots. These differences will be discussed further in the section dealing with the species. In 1865, Day in his ‘‘ Fishes of Malabar” and Duméril in his “‘ Histoire Naturelle des Poissons’’ both described Aétobatis narinari, but as the dis- tinctive points in their description have to do with the color and with the structure of the jaws, both of which subjects will be discussed fully later, and as neither published plates, further attention will not be given them here. In the eighth volume of his ‘‘ Catalogue of Fishes in the British Museum,” Giinther (1870) gives a description of Aétobatis narinari, which, since it is based on a careful examination of no less than ten specimens (mainly from oriental waters), is possibly the most authoritative we have had since Miiller and Henle. In general he agrees with the writers cited above, but gives some interesting variations in color and teeth which will be considered at some length later. Klunzinger (1871) gives a very careful and exact description of A étobatis narinart from the Red Sea. This does not differ materially from Ruppell’s earlier account (1835), but goes into much greater detail. It is interesting to note that both writers record the fact that the spiracle is broader than long; Ruppell says that it is pear-shaped. The points of chief interest in Klunzinger’s description have to do with teeth, tail and spine, and color, all of which will be dealt with later in their appropriate sections. Day (1878), in his ‘‘Fishes of India,’’ gives a description of Aétobatis narinart based evidently upon a number of specimens from Indian waters. These, however, were probably pre- served specimens, since he speaks of the tail as triangular as far as the spine and compressed beyond that, structures much plainer in preserved than in fresh specimens. His figure was certainly made from a preserved fish, since it is much shrunken in the head region. The spine is disproportionally large and the dorsum is entirely devoid of spots despite the fact that in his description the body is noted as covered behind the head with dirty white or bluish spots. This fig- ure is hardly to be recognized as A. narinari and it has not seemed neces- sary to reproduce it here. TEXT-FIG. 2.—A. narinari, after Gunther, 1880. 254 Papers from the Marine Biological Laboratory at Tortugas. In 1880, Gunther published his ‘Introduction to the Study of Fishes,”’ but therein adds little to our knowledge of this ray. His figure (text-fig. 2) is a mere outline drawing, but is one of the best for general appearance which we have. Judged by my specimens, the ventrals are too long and too rounded, the spine too large, the depression over the brain too accentuated, while the absence of spots from the anterior part of the body is especially noticeable. The snout is good and the eyes, it should be noted, are almost invisible, as in the living fish. The same figure is given in this author’s (1886) article on Rays in the ninth edition of the Encyclopedia Britannica. After giving the general characters of the family Myliobatide and of the genus Aétobatus, Jordan and Evermann, in their great work ‘‘The Fishes of North and Middle America”’ (1896), thus describe A étobatus narinari: Disk twice as broad as long, its anterior borders a little convex, posterior concave, outer angles pointed. Cephalic fin about one-third broader than long. Teeth of lower jaw straight or more or less angularly bent. Tail three or four times length of disk. Brown with small round pale spots (Duméril). Tropical seas, north on the Atlantic coast to Virginia: not very common on our shores. Narinari, a Brazilian name. Their figures, both dorsal and ventral, are herein reproduced as text- figures 3 and 4. They were made from a preserved (alcoholic?) specimen taken near Cedar Keys, Florida, and deposited in the United States National Museum. Beyond calling the reader’s attention to the parallel dark markings on the dorsum, which are not mentioned in the text, criticism is deferred until later. The writer has elsewhere (Gudger, 1910) published measurements and brief descriptions of two spotted sting rays taken at Beaufort in 1909, and of another taken at an earlier date. The largest of the three (sex unknown) was taken by some fishermen in the deeper part of the outer harbor in September 1901. Its body was 2 feet 2 inches long and 4 feet wide, while the tail was 4 feet 8 inches long. The second, a female, was taken by the writer in a seine on June 12, 1909, at the Narrows of Newport River, a wide-mouthed tidal estuary whose lower reaches form part of Beaufort Harbor. When just out of the water, its measurements were as follows: length, snout to the tip of the ventrals, 18 inches; width over pectorals 26 inches; length of tail 401% inches. The third specimen, also a female, was, on July 3, 1909, brought to the Fisheries Laboratory by Mr. Russell J. Coles, of Danville, Virginia. A few days previously he had caught it on a hook in the Bight of Cape Lookout, and rightly thinking it to be an unusual form, had preserved it in a tank of formalin loaned him by the laboratory and had brought it in for identifica- tion. This fish was 2314 inches wide and 16 inches long, with a tail meas- uring 3514 inches. A fourth specimen, of the same sex as the preceding, also taken at Cape Lookout (in 1906), is recorded in the card catalogue of fishes at the Laboratory, but no measurements are given. In 1910, I had the good fortune to get three fine specimens of this fish, at the capture of two of which I was present, while the third was brought to The Spotted Eagle Ray. 255 me some 30 minutes after death. These rays have been previously described (Gudger, 1912A), but their measurements will be given here and their mark- ings discussed in the section dealing with color and spots. No. 1, a female, was taken about half-way between the laboratory and the Narrows of Newport River. It was 2814 inches wide, 19 inches long, with a 33 inch tail. The total length was 49 inches and its weight 1114 pounds. It had two spines. No. 2 was a male taken in the channel connecting the inner and outer harbors. Its width was 271% inches, length 18%, tail 3914, total length 541% inches. It had two spines. No. 3, also a male, was taken within 200 yards of the hauling ground where No. 2 was seined. Its width was 37 inches; length, snout to ventrals, 261% inches, tail only 2734, over all 4934 inches. The width between its eyes was 5 inches, the longest diameter of the spiracle was 134 inches, and its weight was over 25 pounds—the limit of my spring balance. This ray had three spines, the middle one of which was torn loose, as was the anterior spine of No. 2 above. Its tail was very short, far too short in proportion to the other dimensions if the fish be compared to a normal ray. Whether or not this abbreviation was the result of an accident, can not be said, since there was nothing to indicate the cause. Interesting points concerning the color, number, size, arrangement, and position of spots of this ray will be taken up in the section dealing with the color. This specimen was the largest and finest of the five which I have had from Beaufort waters. Coles (1910), who (as noted above) had furnished me with a fine specimen early in July 1909, had the good fortune to take more than fifty specimens during that season. During July 1911, he saw nearly as many more but only took eight. The largest of these, a female without eggs or embryos, was 534 feet wide, 3 feet long, tail 534 feet long, total length of body 834 feet, weight 132 pounds. In the summer of 1912, Mr. Coles captured a number of spotted rays at Cape Lookout. The measurements of the three largest are as follows: first specimen, a female, 914 feet long from tip of snout to end of tail, and 7 feet 2 inches wide; the second, also a female, was 12 feet long over all, and 7 feet 7 inches wide, and 20 inches thick measured on a lance thrust through the body; the third, a male, was 10 feet long, and 6 feet 11 inches wide. It is to be regretted that measurements of the body proper and of tails alone were not made. These giant rays would probably have weighed 400 to 500 pounds each. Further consideration of these specimens will be taken up in the section on markings. The specimens taken by Coles and myself, together with certain obser- vations of his noted elsewhere (Coles, 1910), indicate that these rays are far more abundant in the Beaufort region than I had previously thought (Gudger 1910 and 1912A). Their relative abundance along our coasts at certain seasons of the year is also testified to by various fishermen of Beau- fort and Morehead City. 256 Papers from the Marine Biological Laboratory at Tortugas. My Florida specimens agree in general characters with those from North Carolina. The largest (and the largest I have ever seen) measured as follows: width 5 feet 2 inches; length to tip of ventrals 3 feet 8 inches; length of tail only 6 feet 10 inches; length all over, 9 feet 9 inches; width between eyes 8% inches, between spiracles 514 inches; weight 120 pounds. The others need not detain us here. Their measurements are given in the table on page 261. Details of color and number and size of teeth, etc., will be taken up in the sections dealing with these structures. It may be re- marked in passing that these rays seem to be abundant about Key West. Dr. Mayer, in one afternoon in May, off Slaughter-house Point, secured three from fishermen using the grains. At the same place a month later I took the large one above referred to. Dr. Mayer’s specimens were pre- served in a rather unique way. They were small rays and this made it possible to suspend them in a can at the ice factory in Key West where they were frozen solidly in a 300-pound block of ice to await my coming some weeks later. Figure 1, plate 1, which serves as a frontispiece to this paper is repro- duced from a photograph of my specimen No. 3 of I910. Figure 2, plate 1, is a ventral view of specimen No. 2 of 1910, while figure 13, plate vI, is a lateral view of Coles’s 1909 specimen. My specimens were photographed while perfectly fresh, in fact within an hour after being taken. The speci- men from which the lateral view was made had been in formalin for a few days, but so far as could be told was quite normal in all respects. The photographs reproduced in figures 1 and 2, plate 1, were made as follows; to a post standing just before the largest window in the laboratory a cross-piece was nailed; over this was hung a white sheet to serve as a background; in front of this the fish was suspended from the cross-piece by a cord run through the spiracles and their inter-communicating passage. The ray was first suspended by fish-hooks caught in the upper gill-slits, but as the very considerable weight of the fish caused these to tear the flesh, as shown in the ventral view, they were discarded. The method used in getting the lateral view is clear from the figure. The fin is turned up to show the gill-slits. If figure 1, plate 1, be compared with the various illustrations reproduced in this article, some interesting comparisons may be drawn. In Marc- grave’s figure (text-fig. 1), it will be seen that the general outlines are good, even though the drawing is crude. His figure is drawn from a ‘‘quartering”’ view, 7. e., from a point above, but about 45 degrees to the left. He says that the whole upper part is of a steel color and that scattered over the whole of this are white spots. The anterior borders of the pectorals are too convex, the posterior edges too concave, the angles tooacute. The snout is too blunt, the eyes and spiracles too high onthe head. The fore and aft striations were not found in any of my specimens but are shown in Duméril’s figure (fig. 8, plate rv). But with all its defects, even to a novice, text-figure I is plainly a drawing of an eagle ray; and when the description of the teeth is The Spotted Eagle Ray. 257 taken into account, it is unmistakably Aétobatus narinari. But if compar- ison is made with Marcgrave’s original figure, the water-color drawing shown in figure 3, plate 11, the excellence of the latter will be greatly emphasized. As indicated above, the older writers merely copy Marcgrave, and those who do not do worse, for quite a number of them, notably Bloch, apply the name narinari to rays which are not only not Aétobatids but in some cases not even Myliobatids. It will be remembered that Euphrasen declared his figure of the spotted sting ray to be much superior to Willughby’s, which is merely Marcgrave’s figure redrawn and somewhat touched up. But if the figure in question (fig. 5, plate m1) be compared with my photograph or Dumé€ril’s drawing of A étobatus narinart or Marcgrave’s water-color drawing, some extremely grave faults will be revealed. In figure 5, plate 111, body, head, and pectorals seem to be on a level. Both the anterior and posterior edges of the wings are too flat in outline, while the ventral fins are shown as bifurcated backward prolongations of the body. Inspection of figure 1, plate 1, shows them to be short and clearly marked off from the body. The dorsal fin is too long and is wrongly inserted. Figure I, plate 1, does not show it clearly, but text-figure 3 (from Jordan and Evermann) shows both size and position properly. Where Euphrasen got the fringed edges of the pectoral and ventral fins is inexplicable. It is, however, probable that his figure was drawn from a much shrunken, dried or preserved specimen in which the fin rays were very prominent; but even that would not account for their extending beyond the substance of the body proper. Inspection of figure 1, plate 1, shows that the fins in question possess slightly notched edges but nothing more. Attention is also called to the exag- gerated size of the spots, to their relatively small number, and to the fact that there are practically none on the head; also to the position of the spir- acles and eyes, which are dorsal instead of lateral, the eyes in the fresh specimen not being visible from above. Further the snout is too short, too narrow, and too much on a level with the main part of the head. Russell’s (1803) Eel Tenkee (fig. 6, plate 11) is a good drawing with some rather apparent defects from the standpoint of my specimens. The body is too flat, and the head especially so; the anterior parts lack spots, and the eyes are too prominent; posteriorly the line of demarcation between pectorals and ventrals is continued too far forward. However, the crenate posterior edges of the fins are correctly shown. This drawing was made by a native artist, but Russell especially vouches for the accuracy of the details. Duméril’s elegant figures (fig. 8, plate Iv) have already been referred to on page 252. For the reasons noted elsewhere it seems to me that A, Aétobatus latirostris is not (as Gunther asserts) A. narinari, but that B (lower left head) is narinari. The drawing was evidently made from a preserved specimen, in which by drying or hardening the flesh had sunk over the anterior fontanelle of the skull. The position of the spiracles is for the same reason somewhat distorted. 258 Papers from the Marine Biological Laboratory at Tortugas. Gunther's excellent little outline drawing (text-fig. 2) has been previously referred to and need detain us but a moment. One’s attention is forcibly called to the absence of spots on the anterior parts, to the elongated and very rounded ventrals, and to the exaggerated spine. The dorsal fin and head are well drawn, the eyes being almost invisible, but the spiracles are placed rather too near the mid-dorsal line. TEXT-FIG. 3.—A. narinari, dorsal view, after Jordan and Evermann (1900), Florida. The only other figures of this fish (and the only drawing of the ventral surface known prior to 1913) are those previously referred to from Jordan and Evermann’s ‘‘Fishes of North and Middle America,” and widely repro- duced in various papers by one or the other of these ichthyologists, or by other American writers, some of whose papers will be referred to later. If text-figure 3 (their drawing) be compared with figure 1, plate 1 (my photo- graph), the following differences are observable: the pectorals are more sharply pointed, the ventrals more widely spread, longer, and more rounded than in my specimens. The spots are larger and fewer in number (especially on the head and the edges of the pectorals) than in Beaufort and Key West specimens. The most striking difference, however, is to be found in the absence from the latter fish of the dark, roughly parallel lines, which extend from the dorsum of the former out over the pectorals. I have elsewhere recorded (Gudger, 1910) the difficulties I had in identifying my first specimen as A. narinari, but it may not be out of place to repeat them here. The lines above referred to are sometimes present in the living or at any rate in the freshly killed fish, but are so indistinct as The Spotted Eagle Ray. 259 to be found only after close search. So nearly invisible are they that my first specimen, examined while alive and in brilliant sunshine, showed no traces of them. Some hours later, when this fish was brought to the labor- atory and compared with Jordan and Evermann’s figure, despite its agree- ment with the characters given on pp. 87 and 88 of vol. 1 of their ‘‘ Fishes of North and Middle America,’’ I thought it to be a new species; and it was not until Mr. Henry D. Aller, director of the Laboratory, pointed out the very faint striations, that I was convinced to the contrary. It is significant that Jordan and Evermann make no reference whatever in their text to such lines. The photograph (fig. 1, plate 1), made from my fresh fish, shows that they are too indistinct to affect a sensitized plate, even when exposed 60 seconds. Attention may be called here to the sharpness of focus in this photograph, as even the fin rays are shown with marked clearness. Although the question of the presence or absence of lines on the dorsum of this fish will be taken up at length later, it does not seem out of place here to refer to Abbeville’s (1614) Narinnary with its back “‘all striped of black and white.’’ Save for De Laet’s (1633) quotation of Abbeville, no writer has referred to these stripes until Smith (1907) wrote: ‘‘Color above brown, with numerous small, round, pale spots, and transverse dark lines.” With reference to the fins, another difference between Jordan and Ever- mann’s figure and my photograph should be pointed out. In the former the rays at the angles of the pectorals and on their hinder edges are very prominent. Since the drawing is from an alcoholic specimen, this must be due to the macerating action of the alcohol. In the photograph, which was made from a fresh specimen (dead an hour), such rays are visible only in the ventrals, the posterior edges of the pectorals are finely scalloped, as first noted by Russell (1803), but first shown in the painting in the Royal Library of Berlin and here reproduced as figure 3, plate 11. Since particular attention will be given to the head and snout in the section bearing that title, nothing more will be done here than to call attention to the fact that the head is too light in color and too prominent. The spiracles are too marked, the eyes are too prominent and placed too high on the sides of the head, while the snout is too short and too blunt. Figure 2, plate I, is a photograph of the ventral surface of the fish pre- viously described, and, with the exception of Jordan and Evermann’s and Coles’s (1913) figures herein reproduced, is the only one known to the writer. Leaving out of further account the various general features already spoken of, in comparing the two figures, no essential differences are noticeable save in the head. The snout in the photograph shows to better advantage than in the dorsal view, since it is in the ventral plane of the body. That in Jordan and Evermann’s figure is plainly too short and rounded. The mouth is too squarely cut, the lower lip especially so. In the photograph the nasal flaps are longer and fit more closely in the corners of the mouth. Still more marked is the sharper angle of the point of the lower jaw and its 260 Papers from the Marine Biological Laboratory at Tortugas. projection beyond the mouth and indeed over the lower lip. These struc- tures were found in every fish examined by the writer. The first and only description of our spotted sting ray from Brazilian waters since Marcgrave (1648) is that by Miranda Ribeiro (1907), 259 years later. His very clear diagnosis of the characters, both generic and specific, agrees Closely with Jordan and Evermann. These Brazilian rays, however, are dark olive above with blue-white spots, which are round, more or less equidistant from each other, and in size about equal to that of the eyes. TEXT-FIG. 4.—Ventral view of A. marinari, after Jordan and Evermann, 1900, Florida. As Gunther’s (1870) characterization of A. narinari from specimens in the British Museum is the most authoritative since Miiller and Henle (1841), so his description of the single specimen in the Museum Godeffroy at Hamburg (1910) is the most satisfactory since Jordan and Evermann’s (1896). His specimen from Samoa is, he says, not to be distinguished from Atlantic specimens. However, he describes its nose as blunt, whereas my figures and observations plainly show it to be sharp-pointed in all, although of varying widths towards the base. The whole upper surface of the body in his young specimen was covered with numerous bluish-white spots. In connection with notes on the embryos of certain rays (A. narinari especially) taken at Cape Lookout in 1912, Coles has (1913) incidentally published some elegant figures of our ray, which, through his kindness and that of the American Museum of Natural History, are reproduced herein as figures 9, 10, and 11, plate v. One text-figure of his paper will be omitted, since it shows no points not given in figure 9, plate v. This figure is a lateral The Spotted Eagle Ray. 261 view of a female 7 feet 2 inches wide. Figures 10 and 11 are head-on and ventral views of the same fish. Figure 10 is the first head-on view ever published, and the only ventral representation except the Jordan and Evermann drawing, (text-fig. 4) isfigure11. Unfortunately this last is taken at too flat an angle to give a clear representation of the whole ventral surface. Further attention will be called to these figures in other sections.! The following table gives in comparative form the measurements (in inches), absolute and relative, of the various specimens of the spotted eagle ray described by the authors previously quoted: Author. ‘Length. Breadth.| Tail. | B=Lx | T=Lx Abbeville. &. sebicevers sureieyeiate asus "2 72 72 n= r= NEAT GBTAVE Rt nit ysterite cl c.cvecs erenersue s 18 46 51 2.6 2.8 sph rasentsc/o che charostercesterssnetes WED Croc aceoncalivis bp OGAAoratiae 3 BlochySee Schneider eccusc ctere otavenel|| cotevot scone !lvavakerey ckeretlithecponerarailittee Groce 3 RN DElliyvassicretsiebagrens <-s.0:0 dieieters the agek seerat ol NG: daawete syeulleebovard ae 2 4 SCISSEN asi wuss okevareve.e:0 ae) 34 62 3-4 6.2 Miiller & Henle.... 12 28 52 2.3 4.3 OLStEN coats cates oc II 20.5 35 I.9 ce Gantionariyersccca sis iites sieiiscevovsisiepere 725 ELS | Baers 1.6 4-7 BIGEKEI ET ayercu calor ah alan ehencratarore'al|lcvere: cravere'| tate eterersvedl notes 2- 3+ Dame 5 ays. ai paiatsatoia.c Se