THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board E. G. CONKLIN, Princeton University CARL R. MOORE, University of Chicago E. N. HARVEY, Princeton University GEORGE T. MOORE, Missouri Botanical Garden SELIG HECHT, Columbia University T H MORGAN, California Institute of Technology LEIGH HOADLEY, Harvard University G H pARKER Harvard University L. IRVING, Swarthmore College M. H. JACOBS, University of Pennsylvania A' C' AFIELD, Harvard University H. S. JENNINGS, Johns Hopkins University F. SCHRADER, Columbia University FRANK R. LrLLIE, University of Chicago DOUGLAS WHITAKER, Stanford University H. B. STEINBACH, Washington University Managing Editor VOLUME 87 AUGUST TO DECEMBER, 1944 Printed and Issued by LANCASTER PRESS, Inc. PRINCE &. LEMON STS. LANCASTER, PA. 11 THE BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Penn- sylvania. Subscriptions and similar matter should be addressed to The Biological Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain: Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, $1.75. Subscription per volume (three issues), $4.50. Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Massachusetts, between July 1 and October 1, and to the Depart- ment of Zoology, Washington University, St. Louis, Missouri, during the remainder of the year. Entered as second-class matter May 17, 1930, at the post office at Lancaster, Pa., under the Act of August 24, 1912. LANCASTER PRESS, INC., LANCASTER, PA. CONTENTS No. 1. AUGUST, 1944 PAGE ANNUAL REPORT OF THE MARINE BIOLOGICAL LABORATORY 1 SCRIMSHAW, NEVIN S. Embryonic Growth in the Viviparous Poeciliid, Heterandria formosa .... 37 SCHARRER, ERNST The Capillary Bed of the Central Nervous System of Certain Inverte- brates 52 COE, WESLEY R., AND DENIS L. Fox Biology of the California Sea-Mussel (Mytilus californianus). III. En- vironmental Conditions and Rate of Growth 59 DAVIS, JAMES O. Photochemical Spectral Analysis of Neural Tube Formation 73 WILHELMI, RAYMOND W. Serological Relationships between the Mollusca and Other Invertebrates. 96 No. 2. OCTOBER, 1944 DEWEY, VIRGINIA C. Biochemical Factors in the Maximal Growth of Tetrahymena 107 KIDDER, GEORGE W., AND VIRGINIA C. DEWEY Thiamine and Tetrahymena 121 CLARK, LEONARD B., AND GARDINER BUMP X-Rays and the Reproductive Cycle in Ring-Necked Pheasants 134 PACE, D. M., AND W. H. BELDA The Effects of Potassium Cyanide, Potassium Arsenite, and Ethyl Ure- thane on Respiration in Pelomyxa carolinensis 138 HOPKINS, SEWELL H. The External Morphology of the Third and Fourth Zoeal Stages of the Blue Crab, Callinectes sapidus Rathbun 145 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS PRESENTED AT TIII; MARINE BIOLOGICAL LABORATORY, SUMMER OF 1944 153 No. 3. DECEMBER, 1944 HUGHES-SCHRADER, SALLY A Primitive Coccid Chromosome Cycle in Puto sp 167 KENK, ROMAN Ecological Observations on Two Puerto-Rican Echinoderms, Mellita lata and Astropecten marginatus 177 MAST, S. O., AND W. J. BOWEN The Food Vacuole in Peritricha, with Special Reference to the Hydro- gen Ion Concentration of its Content and of the Cytoplasm 188 iii iv CONTENTS I'AGE ROMANOFF, ALEXIS L. Hydrogen Ion Concentration of Albumen and Yolk in the Developing Avian Egg 223 SPIEGELMAX, S., AND FLORENCE MOOG On the Interpretation of Rates of Regeneration in Tubularia and the Significance of the Independence of Mass and Time 227 SCIIARRKK, BKKTA AND ERNST SCHARRER \Yurosecretion VI. A Comparison between the Intercerebralis- Cardiacum-Allatum System of the Insects and the Hypothalamo- Hypophyscal System of the Vertebrates 242 MARSLAND, DOUGLAS Mechanism of Pigment Displacement in Unicellular Chromatophores. . 252 Vol. 87, No. 1 August, 1944 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY THE MARINE BIOLOGICAL LABORATORY FORTY-SIXTH REPORT. FOR THE YEAR 1943 — FIFTY-SIXTH YEAR I. TRUSTEES AND EXECUTIVE COMMITTEE (AS OF AUGUST 10, 1943) .... 1 STANDING COMMITTEES 2 II. ACT OF INCORPORATION 3 III. BY-LAWS OF THE CORPORATION 4 IV. REPORT OF THE TREASURER 5 V. REPORT OF THE LIBRARIAN 10 VI. REPORT OF THE DIRECTOR 11 Statement 11 Addenda : 1. Appreciation of the Services of Professor Frank R. Lillie to The Marine Biological^ Laboratory 14 2. Memorials to Deceased Trustees 14 3. The Staff. 1943 17 4. Investigators and Students, 1943 19 5. Tabular View of Attendance 24 6. Subscribing and Co-operating Institutions. 1943 24 7. Evening Lectures, 1943 25 8. Members of the Corporation 25 I. TRUSTEES EX OFFICIO FRANK R. LILLIE, President Emeritus of the Corporation, The University of Chicago LAWRASON RIGGS, President of the Corporation, 120 Broadway, New York City E. NEWTON HARVEY, Vice President of the Corporation, Princeton University CHARLES PACKARD, Director, Marine Biological Laboratory OTTO C. GLASER, Clerk of the Corporation, Amherst College DONALD M. BRODIE, Treasurer, 522 Fifth Avenue, New York City EMERITUS E. G. CONKLIN, Princeton University B. M. DUGGAR, University of Wisconsin W. E. CARREY, Vanderbilt University CASWELL GRAVE. \Yashington University R. A. HARPER, Columbia University Ross G. HARRISON, Yale University H. S. JENNINGS, University of California MA KIM- BIOLOGICAL LABORATORY C. E. McCLUNG, University of Pennsylvania S. O. MAST, Johns Hopkins University A. P. MATIIF.WS. University of Cincinnati T. H. MORCAX, California Institute of Technology \V. J. V. OSTERIIOUT. Rockefeller Institute .4o3.00 429.53 152.22 86.84 17.10 $ 29.310.84 7,300.14 5,085.00 8,662.50 6,859.33 41.652.10 19,093.67 15,439.65 18.796.00 500.00 3,600.0(1 360.00 300.00 999.96 245.00 446.14 613.55 33.06 $ 1,513.09 25.00 7.165.81 19,529.79 4,683.32 1,588.86 6,463.00 429.53 152.22 86.84 17.10 Income $ 29.310.S4 7,300.14 4.174.20 1,567.31 14.436.21 1,856.50 24,866.25 18,796.00 500.00 2,882.23 330.83 32.53 999.96 245.00 446.14 613.55 33.06 REPORT OF THE TREASURER General Expenses : Administration Expense 14,386.77 14,386.77 Endowment Fund Trustee and Safe-keep- ing 1,014.45 1,014.45 Bad Debts 720.29 720.29 Special Repairs, Supply Dept. Stone Build- ing 1,963.56 1,963.56 Reserve for Repairs and Replacements, Buildings occupied by Navy 2,358.34 2,358.34 Reserve for Depreciation 26,969.1 1 26,969.1 1 $139,973.27 $159,296.94 $ 89.067.08 $108,390.75 Excess of Income over Expense carried to Current Surplus $ 19,323.67 $ 19,323.67 $159,296.94 $108,390.75 EXHIBIT C MARINE BIOLOGICAL LABORATORY, CURRENT SURPLUS ACCOUNT, YEAR ENDED DECEMBER 31, 1943 Balance, January 1, 1943 $159,672.77 Add : Excess of Income over Expense $19,323.67 Gain on Gansett Lots Sold 279.41 Bad Debt Recovered .90 Transfer of Biological Bulletins, held at Lancaster Press for Un- completed Serial Exchanges, 1940, 1941, and 1942 from Plant Account, Library to Biological Bulletin Inventory . .' 2,552.50 Reserve for Depreciation charged to Plant Funds 26,969.11 49,125.59 $208,798.36 Deduct : Payments from Current Funds during Year for Plant Assets : Buildings $ 2,460.20 Equipment 984.64 Librarv 4.583.95 $ 8,028.79 Less Received for Plant Assets Sold 140.00 $ 7,888.79 Pensions Paid 3,460.00 Loss on Retirement Fund Securities 181.48 $ 3,641.48 Less Retirement Fund Income 3.408.76 Transfers to Reserve Fund, Portion of Dividends of Prior Years from General Biological Supply House, Inc. and Crane Company 8,500.00 Profit on Gansett Lots for 1941 and 1942 498.71 8.998.71 20,296.26 Balance, December 31, 1943 ... $188,502.10 Respectfully submitted, DONALD M. BRODIE, Treasurer 10 MAKIX1-: BIOLOGICAL LABORATORY V. KKI'ORT OF THE LIBRARIAN The librarian's re-port for 1942 gave a detailed account of the readjustments that the library lias gradually made to changes of the past few years in the budget, in the receipt of European and Asiatic current journals and in the available market - for purchases of "back sets." It is not necessary this year to comment on these adjustments except to state that definite effort has been made to renew exchange relationship with Russia. China and India and with the limited success that wa- anticipated. There should be some results to report for this in 1944-45, depending, of course, on the circumstances of the war. The budget of $12.200 for 1943 was expended as follows: books, $31 4.3.} : serials. $1871.51; binding, $645.41; express. $64.29; supplies. $202.91; salaries. $7200.00; back sets, $815.77; sundries. $3.07; and insurance. $45.00; total. $11.- 162.29. The sale of duplicates brought in this year $214.03; and the income from the microfilm service amounted to $221.33, 99 orders having been filled. From the "Carnegie Fund" $1745.92 was expended on two completed and 14 partially completed "back sets" and three books that are designated as "classics." The Woods Hole Oceanographic Institution appropriated $1850.00 for 1943 and a balance of $70.11 remained from the 1942 budget. An expended sum of $1657.03 has been reported to the Director. A balance of $263.08 was carried on to the year 1944. During 1943 the library received 645 current journals: 232 (10 new) by sub- scriptions to the Marine Biological Laboratory; 15 (none new) to the Woods Hole Oceanographic Institution; exchanges, 191 (two new) with the "Biological Bul- letin" and 22 (six new) with the Woods Hole Oceanographic Institution publica- tions; 181 as gifts to the former and four to the latter. The Marine Biological Laboratory acquired 113 books: 47 by purchase of the Marine Biological Laboratory (three "classics") ; 11 by purchase of the Woods Role Oceanographic Institution; 12 gifts from the authors, 19 from the publishers, 10 from Mr. John Crane, three standard medical books from Lt. F. G. Hirsch which came from his father's library, and 11 as miscellaneous gifts. There were 20 back sets of serial publications com- pleted: five, purchased by the Marine Biological Laboratory (two with "Carnegie Fund") ; three, by the Woods Hole Oceanographic Institution; one secured by ex- change with the "Biological Bulletin" ; five by exchange with the Woods Hole Oceanographic publications; one as a gift from Lt. F. G. Hirsch; and five secured by duplicate material exchange and by gift. Partially completed sets were 50: purchased by the Marine Biological Laboratory, 23 (14 with "Carnegie Fund); three by the Woods Hole Oceanographic Institution ; by exchange of the "Biological Bulletin," none; by exchange of the Woods Hole Oceanographic Institution pub- lications, three; by gift and by exchange of duplicate material, 21. The reprint additions to the library number 7927: current of 1941, 933; current of 1942, 1471 ; current of 1943, 381 ;'and of previous dates, 5142. A total of 981 reprints, 296 not duplicates of our holdings, were presented to the library: 396 by Dr. I I. K. Crampton ; 63 by Dr. H. S. Hopkins; and 522 by Dr. Libbie H. Hyman. At the end of the year 1943 the library contained 51.945 bound volumes and 130.650 reprints. REPORT OF THE DIRECTOR 11 VI. THE REPORT OF THE DIRECTOR To THE TRUSTEES OF THE MARINE BIOLOGICAL LABORATORY : Gentlemen : I beg to submit herewith a report of the fifty-sixth session of the Marine Bio- logical Laboratory for the year 1943. 1. ResearcJi During the past year research and instruction have gone on as usual, but the exigencies of war teaching and war research, and the draft have cut deeply into our attendance. There were 108 investigators as compared with an average of 366 in the five years preceding the war. The number of Library readers was somewhat larger than before. In most cases, the length of stay at the Laboratory was shorter than in normal years. The Friday Evening Lectures were continued as usual, but no seminars were held, nor was there a general scientific meeting at the end of the season. Research directly connected with the war was directed by Drs. D. E. S. Brown, Clowes, Heilbrunn, and Jacobs. Other projects were spon- sored by the Oceanographic Institution, the Massachusetts General Hospital, and by the Bell Telephone Laboratories. During the Spring of 1944, a number of investigators from the Oceanographic Institution, engaged in war research, have occupied some of our Laboratory rooms. 2. Instruction The Laboratory has been able to continue its courses of instruction with able teaching staffs and with a reasonable number of students. Of a total of 68 stu- dents in all of the courses, 11 were men and 57 were women. In peace times the total has been about 125, somewhat less than half being men. The Invertebrate course, under Dr. Buck and an almost entirely new corps of instructors, was suc- cessfully given to a class of almost maximum size. Dr. Hamburger and Dr. Costello, with the occasional help of Drs. Barth, Metz, and Rose, carried on the Embryology course with excellent results. Pressure of war work prevented most of the instructors in Physiology from remaining in residence during the entire period of the course, but in spite of this, Dr. Parpart and his associates were able to conduct the course satisfactorily. It was a great disappointment that the course in Algae had to be omitted because of a lack of students. This is the first time since the course first began in 1889 that instruction in Botany has not been given at this Laboratory. 3. Buildings and Grounds With the abatement of danger from enemy attacks in this region, the Navy closed its Base at Woods Hole and gave back to us, on January 1, 1944, our build- ings which they had occupied since May 1942. During their tenancy they made a number of alterations, many of which are of permanent value. The Lecture Hall and the Botany Building were given new roofs and shingles and solid shorings to support the floors. In the Homestead new sills and other timbers, and floors, all greatly needed, were installed. The Mess was improved by a new roof and in- sulated walls, a completely rebuilt ice box and store room, and by the addition of two new insulated store rooms. At the Penzance Garage a new concrete floor was laid where the old one was imperfect. In addition, the water front which had been 12 MARINE BIOLOGICAL LABORATORY weakened by the hurricane and only partially repaired, was thoroughly rebuilt. These improvements, which we should have had to make sooner or later, are con- servatively valued at $12.000. On the other hand, the Xavy made other alterations which rendered the build- ings unsuitable for our use. In the lease, the Navy agreed to restore the property to its original condition. But it was obvious that the necessary repairs and replace- ments could be made more satisfactorily by our own staff than by an outside contractor. Consequently, after a period of bargaining, it was agreed that the Navy should leave in place some valuable equipment, useful to the Laboratory, in lieu of actual payment for the cost of the repairs. Among the items transferred to us may be mentioned a series of gas cooking ranges equipped with a ventilating hood, and an oven. These take the place of our old coal range which was outworn. Also a water cooler, a refrigerating unit in the icebox, a hot water heating system, extensive plumbing installations in the Homestead, and 12 tons of hard coal in the Apartment House. This equipment is valued at more than $6000, an amount well above the cost of repairs and replacements made by our staff. The Lecture Hall and the Botany Building have not been completely restored since they will not be needed in 1944. The Apartment House, which was returned to use in fairly good condition, has been redecorated for the first time since it was built in 1927. The Penzance Garage is now leased to the Oceanographic Institution for the duration. Although a number of our buildings are now in better condition than heretofore, some need further attention. Specifically, the walls of the Apartment House should be waterproofed, at least on those sides exposed to storms. Unless this is done soon much of the new decoration will be marred. The Brick Building also should be similarly treated. 4. Financial The Treasurer's report shows that our finances are in some respects in excellent condition. The cash balance is larger than usual, due in part to the large dividend of the General Biological Supply House and increased sales by our own Supply Department, but especially to the rental paid by the Navy. At the same time our expenditures have been low for the same reasons that were noted last year, namely, that we have been unable to purchase new equipment and to pay for current foreign journals. Because of this favorable cash balance we have felt justified in adding a substantial amount to the Reserve Fund, and in making the extensive improvements in the Apartment House. On the other hand, our income for 1944 will be reduced by reason of the with- drawal of the Navy and the relatively small return from fees of investigators and students. We cannot expect any considerable increase in our regular sources of income. Our present prosperity is therefore transient. During the next few years we must provide funds, in addition to the regular budget, to pay for foreign jour- nals held in store for the duration, for the development of the Apparatus Depart- ment, including the services of a full-time mechanic, and especially for the build- ing up of the Retirement Fund which is now far too small. The Executive Committee is considering the last item. The Committee on Additional Funds find> that the present time is not favorable for obtaining money, but it has laid plans for the future. REPORT OF THE DIRECTOR 13 5. Losses by Death In the death of Dr. Caswell Grave the Laboratory loses one of its most loyal members, whose important services, willingly given during a period of more than 30 years, are gratefully remembered. Prof. A. D. Morrill. Emeritus Professor of Biology at Hamilton College, active in the early days of the Laboratory, died June 8, 1943. Pof. John M. McFarlane, Emeritus Professor of Botany at the University of Pennsylvania, Trustee of the Corporation from 1897 to 1902, died September 16, 1943. 6. Gifts The Laboratory is indebted to Mr. John Crane, for a letter written by his father, Charles R. Crane, to Mr. John D. Rockefeller, Jr., on the occasion of the setting up of the Endowment Fund to which both men had generously contributed. In the letter he refers in the following words to the unusual character of the Labora- tory when he first become acquainted with it. "These scientists were struggling and accomplishing marvelous things with most meagre equipment, making many sacrifices. It seemed to me that the precious thing to preserve was the spirit of the organization, a spirit everywhere recognized although hard to seize or to imitate. So we have been most careful not in any way to jeopardize this spirit : and processes of organization and management have continued as I found them. It is a valuable expression of a democracy of educated, high minded men." 7. Election of Trustees At the meeting of the Corporation held August 10. 1943, the following Trustees were elected Trustees Emeritus : B. M. Duggar, The University of Wisconsin W. E. Garrey, Vanderbilt University Medical School The new Trustees elected at that meeting are : Paul S. Galtsoff, U. S. Fish and Wild Life Service E. W. Sinnott, Professor of Botany, Yale University 8. There are appended as parts of this report : 1. Appreciation of the Services of Professor Frank R. Lillie. 2. Memorials to deceased Trustees. 3. The Staff. 4. Investigators and Students, 1943. 5. A Tabular View of Attendance, 1939-1943. 6. Subscribing and Co-operating Institutions. 7. Evening Lectures. 8. Members of the Corporation. Respectfully submitted, CHARLES PACKARD, Director 14 MARINE BIOLOGICAL LABORATORY 1. APPRECIATION OF THE SERVICES OF PROFESSOR FRANK R. LILLIE TO THE MARINE BIOLOGICAL LABORATORY READ AT THE MEETING OF THE CORPORATION, AUGUST 10, 1943 In the. history of the Marine Biological Laboratory the names of two men are pre-eminent: Dr. Whitman, who with prophetic insight, envisioned this institution as a national center of research in every department of Biology ; and Dr. Lillie, who transformed that vision into reality. Coming to Woods Hole first in 1891 as an investigator receiving instruction. Dr. Lillie, with Dr. Whitman, organized the course in Embryology in 1893. He was appointed Assistant Director in 1900 at a time when the fortunes of the Laboratory were at a low ebb; Director in 1908; and President of the Corporation in 1926, after the successful conclusion of the cam- paign to obtain new buildings and an endowment. During the period from 1900 to 1942, when he resigned from the Presidency, the Marine Biological Laboratory developed from a struggling organization to its present position as the leading co- operative laboratory of the world. It is of course true that only by the devoted work of the members of the Cor- poration, and the active interest of its many friends, could such an end be reached ; but it is equally true that without wise guidance this effort would have failed. From the beginning, when Whitman, against every force and discouragement, fought for the principles of co-operation and independence, this Laboratory has pursued its steady course, adapting itself wisely to new conditions as they arose, but always holding to those basic ideals. During his fruitful years as Director, Dr. Lillie frequently stressed these principles. "Our purpose," he wrote, "is essentially ideal, and its pursuit demands our best efforts and our loyalty." And again, "We have laid the principle of co-operation at our foundation, and we have attempted to build it into every one of our activities." In this course he has always quietly led. There has never been any though of division since he has been in charge. Here lies his strength, and here lies the secret of the continued success of the Laboratory. In accepting his resignation from the Presidency, the Corporation and the Trus- tees are rejoiced that he will continue his connection with the Laboratory as Presi- dent Emeritus. We extend to him and to Mrs. Lillie, who has so ably assisted him in the development of the Marine Biological Laboratory, our grateful thanks, and we pledge to him our best efforts to continue the work which he has so long and so wisely guided. C. E. MCCLUXG E. G. CONKLIN CHARLES PACKARD 2. MEMORIALS TO DECEASED TRUSTEES I. MEMORIAL TO DR. HEKMON CAREY BUMPUS Read by Dr. .1. /'. Hcrnion Carey Bumpus, Trustee Emeritus of the Marine Biological Laboratory, died June 21, 1943, at the age of eighty-one years. The Laboratory thus loses a REPORT OF THE DIRECTOR 15 member who played an active part in its development for more than forty years. Coming first to Woods Hole in 1889 when a graduate student of Whitman at Clark University, he worked here on his thesis, "The Embryology of the American Lobster." In 1890 he returned to Brown University, his Alma Mater, where he taught Comparative Anatomy for eleven years. It was during this period that he served at this Laboratory as head of the Invertebrate Zoology course ; as Assistant Director from 1893 to 1895; and as Clerk of the Corporation from 1897 to 1899. He was a Trustee from 1897 to 1932 when he became Trustee Emeritus. From 1898 to 1901 lie was the Director of the Laboratory of the Fish Com- mission at Woods Hole, during which time he made a careful survey of the fauna of his region. At this time also he made one of the first studies of variation and its bearing on Natural Selection. Many of his students from Brown came to us with him, among whom were George M. Gray, A. D. Mead, H. W. Walter, and F. P. Gorham. Dr. Bumpus was remarkably efficient in the work of organization in all the positions which he held. At this Laboratory he greatly improved the equipment for work and for collecting living material ; and due to his efforts the number of students at the Laboratory greatly increased, 85 being registered one year in the Invertebrate Course. To bring Biology to the people was his chief interest. Accordingly he left Brown in 1901 to become Director of the American Museum of Natural History in New York, where he remained ten years and carried out his long cherished plans for taking the resources of the museum to the school children in New York City, an educational project which has since grown to great proportions throughout the country. Subsequently, while in the National Park Service, he developed many museums in the State and National Parks. In recognition of this work he re- ceived the Pugsley Medal for his service to education. For three years he was Business Manager of the University of Wisconsin, and for five years, President of Tufts College. He was a Trustee also of several charitable institutions. These are only a few of the many accomplishments of this tireless worker. He Avas, in the words of his student and life-long friend. Dr. H. E. Walter, "A natural teacher, and enthusiastic scientist, a remarkable executive, and a genial gentleman." II. MEMORIAL TO DR. G. N. CALKINS By Dr. L. L. Woodruff The distinguished incumbent of the first Professorship of Protozoology in America, Gary Nathan Calkins, died at his home in Scarsdale, New York, on January 4, 1943, after a considerable period of ill health which was endured with characteristic cheerfulness and fortitude. Calkins was born at Valparaiso, Indiana, on January 18, 1869, but spent nearly all of his life on the Eastern seaboard. His scientific training began at the Massa- chusetts Institute of Technology where, under the influence of Professor William T. Sedgwick, an interest was aroused in biology as a profession. After receiving the B.S. degree in 1890 he served until 1893 as lecturer at the Institute and also as Assistant Biologist to the Massachusetts State Board of Health. Then he trans- 16 MARINE BIOLOGICAL LABORATORY t'erred to Columbia University to study under Professor Edmund D. Wilson and received the Ph.D. degree in 1898. While a graduate student he was appointed to the teaching staff and thus began the life-long membership in the Department of Zoology at Columbia, which in 1907 culminated as Professor of Protozoology. Cal- kins was for some years the Executive Officer of the Department, and retired in 1940 as Professor Emeritus in residence. Columbia University conferred on him the honorary degree of Sc.D. in 1929. Calkins' devotion of his life to the study of the Protozoa was inspired both by an inherent interest in the "little animals," and the well-founded belief that they afford highly favorable material for the approach to many general biological prob- lems. This is best exemplified by his most important treatise, The Biology of the Protozoa (1926, 2nd edition 1933), and his long-continued studies on the physi- ology and cytology of free-living Ciliates, with particular reference to the signifi- cance of fertilization and other factors influencing longevity. In this classic work- he devised more exact methods of pedigreed culture, involving daily isolation of the animals, that laid the foundations for present-day technique in the field, and he also developed what may be referred to as his philosophy of the Protozoan individual. Both phases stimulated many investigators to enter similar fields. The extensive series of important studies from Calkins' laboratory is but <>ne of his many contributions to science. A brilliant lecturer and teacher at Columbia and at The Marine Biological Laboratory, his courses revealed a comprehensive grasp of protozoology from both its theoretical and practical aspects that inspired many students; and his versatility was shown by numerous other activities. Thus, for example, he was Consulting Biologist to the New York State Cancer Labora- tory at Buffalo from 1902 to 1908. Lecturer before the Lowell Institute in 1907, President of the Association for Cancer Research in 1916, President of the Society for Experimental Biology and Medicine from 1919 to 1921. and Director of the University Union in Paris in 1926 and 1927. He was elected in 1919 to the Na- tional Academy of Sciences. Calkins' association with The Marine Biological Laboratory began just over a half century ago. and for about forty years lie was in regular attendance as an investigator. He was a member of the Corporation for 39 years and its Clerk for 17 years, member of the Board of Trustees for 30 years and its Secretary for 12 years, member of the Research Staff for 31 years, and head of the Protozoology course, which he founded, for 22 years. Zoology in general and Columbia University and The Marine Biological Labora- tory in particular owe to Calkins more than can be readily expressed for his scien- tific contributions, teaching, and administrative service. All this, as well as his personal charm, unfailing enthusiasm, and hearty good fellowship, was attested by his former students and associates who presented to him after retirement a volume of nearly two hundred letters of esteem and appreciation inscribed: (iiirv Xatluni Calkins J'hilosofher in Little Thinys and Friend. REPORT OF THE DIRECTOR 17 3. THE STAFF, 1943 CHARLES PACKARD, Director, Marine Biological Laboratory, Woods Hole, Massachusetts. SENIOR STAFF OF INVESTIGATION E. G. CONKLIN, Professor of Zoology, Emeritus, Princeton University. CAS WELL GRAVE, Professor of Zoology, Emeritus, Washington University. FRANK R. LILLIE, Professor of Embryology, Emeritus, The University of Chicago. RALPH S. LILLIE, Professor of General Physiology, Emeritus, The University of Chicago. C. E. McCLUNG, Professor of Zoology, Emeritus, University of Pennsylvania. S. O. MAST, Professor of Zoology, Emeritus, Johns Hopkins University. A. P. MATHEWS, Professor of Biochemistry, Emeritus, University of Cincinnati. T. H. MORGAN, Director of the Biological Laboratory, California Institute of Technology. G. H. PARKER, Professor of Zoology, Emeritus, Harvard University. ZOOLOGY I. CONSULTANTS T. H. BISSONNETTE, Professor of Biology, Trinity College. L. L. WOODRUFF, Professor of Protozoology, Yale University. II. INSTRUCTORS J. B. BUCK, Assistant Professor of Zoology, University of Rochester, in charge of course. W. D. BURBANCK, Associate Professor of Biology, Drury College. M. D. BURKENROAD, Assistant Curator, Bingham Oceanographic Foundation, Yale Uni- versity. C. G. GOODCHILD, Associate Professor of Biology, Southwest Missouri State Teachers College. RONALD GRANT, Lecturer in Physiology, McGill University. JOHN H. LOCHHEAD, Instructor in Zoology, University of Vermont. MADELENE E. PIERCE, Assistant Professor of Zoology, Vassar College. MARY D. ROGICK, Professor of Biology, College of New Rochelle. III. LABORATORY ASSISTANT MARGARET L. KEISTER, Instructor in Biology, Wheaton College. EMBRYOLOGY I. CONSULTANTS L. G. EARTH, Assistant Professor of Zoology, Columbia University. H. B. GOODRICH, Professor of Biology, Wesleyan University. II. INSTRUCTORS VIKTOR HAMBURGER, Professor of Zoology, Washington University, in charge of course. DONALD P. COSTELLO, Assistant Professor of Zoology, University of North Carolina. CHARLES B. METZ, Instructor, Wesleyan University. PHYSIOLOGY I. CONSULTANTS WILLIAM R. AMBERSON, Professor of Physiology, University of Maryland, School of Medicine. REPOR'OE THE DIRECTOR GENEI L MAINTENANCE T. E. I-iKix, Superintendent \V. ( . 1 1 KM! XWAV T. H. TAV. K. \\'. K \III.F.R J. \\"v- \. J. PIERCE THE GEORE M. GRAY MUSEUM GKtant Profit-, v ; Columbia University. hs A.. Director, I'.i.ii.^''' Laboratory, Fordliam University. HA M., Research Associi.- in Biology, University of California. » F. S.. Proft*vr ..i ^y. New York University. College of Dentisti !'.., Assistant Profe-»or of oology, University of Rochester. liKRT A., Professor of Ztlogy, Emeritus, Oberlin College. >s \\'.. Professor of I'- ;ind Head of Division of the Sciences, Dillard \\IU.IAM D., Associate Prcssor and Chairman of Department of Biology, D MI. MARTIN D.. Assistant C-ator, Bingham Oceanographic Foundation, Vale i r. Research Profes: i of Biology, Washington Square College. New 1 Cir I... Voluntary 1 <>r. University of Pennsylvania, School of Medicii and Head ofl Department of Anatomy, University of Pennsylv; S \ssociate Prof-sor of Biology, College of Charleston. i , 1 1 A . r of Researcl Eli Lilly and Company. :cssor of Biolcy. Emeritus, Princeton University, -sociate Profesir of Zoology. University of North Carolina. cssor of tint Physiology, University of Wisconsin. I \ .nt Professor i> Radiology. College of Physicians and Surgeons. lumbia Un: 1-M Kadiology. Cccge of Physicians and Surgeons. Columbia Univer MKK. KKNNUI .nt Profes>r of Physiological Zoology. University of Torontt [EL. Moui ;»artment of .'ology, Columbia University. :ii..r Biologist. I S. Fish and Wildlife Service. \\ ! I Physiolop. Vanderbilt University. School of Medicine. r of Biolog> Amherst College. H ii ii. i.. . Associate Pifessor of Biology, State Teachers College, Spring! Missouri. ,ate. Wa.sngton Square College. New "S'ork University. • NHK. SAM, A — '.-tant. Rockefeller Lititute. () MARINE lUni.OI.K'AI. I.AI'.OKATORY GRAVE, PROF. B. H., DePauw University, Greencastle. Indiana. GRAVK. PROF. CASWELL, \Yashington University, St. Louis, Missouri. GRAY, PROF. IRVING E.. Duke University, Durham, North Carolina. GREGORY, DR. LOUISE H.. Barnard College, Columbia University, New York City, New York. GUDERNATSCII, J. FREDRICK. New York University, 100 Washington Square, New York City, New York. GUTHRIE, DR. MARY J., University of Missouri, Columbia, Missouri. GUYER, PROF. M. F., University of Wisconsin, Madison, Wisconsin. HAGUE, DR. FLORENCE, Sweet Briar College, Sweet Briar, Virginia. HALL. PROF. FRANK G., Duke University, Durham, North Carolina. HAMBURGER, DR. VIKTOR, Department of Zoology, Washington University, St. Louis, Missouri. HANCE, DR. ROBERT T.. Department of Biology, Duquesne University, Pittsburgh, Pennsylvania. HARGITT, PROF. GEORGE T., Department of Zoology, Duke University, Durham, North Carolina. HARMAN, DR. MARY T., Kansas State Agricultural College, Manhattan, Kansas. HARNLY, DR. MORRIS H., Washington Square College, New York University, New York City, New York. HARPER, PROF. R. A., R. No. 5, Bedford, Virginia. HARRISON, PROF. Ross G., Yale University, New Haven, Connecticut. HARTLINE, DR. H. KEFFER, University of Pennsylvania, Philadelphia, Pennsylvania. HARTMAN, DR. FRANK A., Hamilton Hall, Ohio State University, Columbus, Ohio. HARVEY, DR. E. NEWTON, Guyot Hall, Princeton University, Princeton, New Jer- sey. HARVEY, DR. ETHEL BROWNE, 48 Cleveland Lane, Princeton, New Jersey. HAYDEN, DR. MARGARET A., Wellesley College, Wellesley, Massachusetts. HAYES, DR. FREDERICK R., Zoological Laboratory, Dalhousie University, Halifax, Nova Scotia. HAYWOOD, DR. CHARLOTTE, Mount Holyoke College, South Hadley, Massachusetts. HECHT, DR. SELIG, Columbia University, New York City, New York. HEILBRUNN, DR. L. V., Department of Zoology, University of Pennsylvania, Phila- delphia, Pennsylvania. HENDEE, DR. ESTHER CRISSEY, Russell Sage College, Troy, New York. HENSHAW, DR. PAUL S., National Cancer Institute, Bethesda, Maryland. 1 IKSS, PROF. WALTER N., Hamilton College, Clinton, New York. IIiATT. DR. E. P.. Xew York University, 100 Washington Square, New York City, Xew York. HIBBARD, DR. HOPE, Department of Zoology, Oberlin College, Oberlin, Ohio. I In. i., DR. SAMUKL E.. Department of Biology, Russell Sage College, Troy, New York. Mi. \RICHS, DR. MARIE. Department of Physiology and Health Education, South Illinois Normal University, Carbondale, Illinois. IfrsAW, DR. F. L., Harvard University, Cambridge, Massachusetts. 1 IOADLEY, DR. LEIGH. Harvard University, Cambridge, Massachusetts. HOBER, DR. RUDOLF, University of Pennsylvania, Philadelphia. Pennsylvania. REPORT OF THE DIRECTOR 31 HODGE, DR. CHARLES. IV, Temple University, Department of Zoology, Philadelphia, Pennsylvania. HOGUE, DR. MARY J.. University of Pennsylvania Medical School, Philadelphia, Pennsylvania. HOLLAENDER, DR. ALEXANDER, c/o National Institute of Health, Laboratory of In- dustrial Hygiene, Bethesda, Maryland. HOPKINS, DR. DWIGHT L., Mundelein College, 6363 Sheridan Road, Chicago, Illi- nois. HOPKINS, DR. HOYT S.. New York University, College of Dentistry, New York City, Newr York. HOWLAND. DR. RUTH B., Washington Square College, New York University, Washington Square East, New York City, New York. HOYT, DR. WILLIAM D., Washington and Lee University. Lexington, Virginia. HYMAN, DR. LIBBIE H., American Museum of Natural History, New York City, New York. IRVING, PROF. LAURENCE, Swarthmore College, Swarthmore, Pennsylvania. ISELIN, MR. COLUMBUS O'D., Woods Hole, Massachusetts. JACOBS, PROF. MERKEL H., School of Medicine, University of Pennsylvania, Phila- delphia, Pennsylvania. JENKINS, DR. GEORGE B., 30 Gallatin Street, N.W., Washington, D. C. JENNINGS, PROF. H. S., Department of Zoology, University of California, Los An- geles, California. JOHLIN, DR. J. M., Vanderbilt University Medical School, Nashville, Tennessee. JONES, DR. E. RUFFIN, JR., College of William and Mary, Williamsburg, Virginia. KAUFMANN, PROF. B. P., Carnegie Institution, Cold Spring Harbor, Long Island, New York. KEMPTON, PROF. RUDOLF T., Vassar College, Poughkeepsie, New York. KIDDER, DR. GEORGE W., Brown University, Providence, Rhode Island. KILLE, DR. FRANK R., Swarthmore College, Swarthmore, Pennsylvania. KINDRED, DR. J. E., University of Virginia, Charlottesville. Virginia. KING, DR. HELEN D., Wistar Institute of Anatomy and Biology, 36th Street and Woodland Avenue. Philadelphia, Pennsylvania. KING, DR. ROBERT L., State University of Iowa, Iowa City, Iowa. KNOWLTON, PROF. F. P.. Syracuse University, Syracuse, New York. KOPAC, DR. M. J., Washington Square College, New York University, New York City, New York. KORR, DR. I. M., Department of Physiology, New York University, College of Medi- cine, 477 First Avenue, New York City, New York. KRAHL, DR. M. E., Lilly Research Laboratories, Indianapolis, Indiana. KRIEG, DR. WENDELL J. S., New York University, College of Medicine, 477 First Avenue, New York City, New York. LANCEFIELD, DR. D. E., Queens College, Flushing, New York. LANCEFIELD, DR. REBECCA C.. Rockefeller Institute, 66th Street and York Avenue, New York City, New York. LANGE, DR. MATHILDE M., Wheaton College, Norton, Massachusetts. LEWIS, PROF. I. F., University of Virginia, Charlottesville. Virginia. LILLIE, PROF. FRANK R., The University of Chicago, Chicago, Illinois. MAK1XK BIOLOGICAL LABORATORY LILLIE, PROF. RALPH S., The University of Chicago, Chicago, Illinois. LITTLE, DR. E. P., Phillip Exeter Academy. Exeter, New Hampshire. I .OKI;. PROF. LEO, 40 Crestwood Drive, St. Louis, Missouri. LOEWI. PROF. OTTO, 155 East 93d Street, New York City, New York. LOWTHER, MRS. FLORENCE DEL., Barnard College, Columbia University, New York City. NYw York. LTCAS, DR. ALFRED M., Zoological Laboratory, Iowa State College, Ames, Iowa. LUCAS, DR. MIRIAM SCOTT, Department of Zoology, Iowa State College, Ames, Iowa. LUCKE, PROF. BALDUIN, University of Pennsylvania, Philadelphia, Pennsylvania. LYNCH, DR. CLARA J., Rockefeller Institute, 66th Street and York Avenue, New York City, New York. LYNCH, DR. RUTH STOCKING, Maryland State Teachers College, Towson, Mary- land. LYNN, DR. WILLIAM G., Department of Biology, The Catholic University of Amer- ica, Washington, D. C. MACDOUGALL, DR. MARY S., Agnes Scott College, Decatur, Georgia. MACLENNAN, DR. RONALD F., 174 Forest Street, Oberlin, Ohio. MACNAUGHT, MR. FRANK M., Marine Biological Laboratory, Woods Hole, Massa- chusetts. McCLUNG, PROF. C. E., 417 Harvard Avenue, Swarthmore, Pennsylvania. McCoucH, DR. MARGARET SUMWALT, University of Pennsylvania Medical School, Philadelphia, Pa. MCGREGOR, DR. J. H., Columbia University, New York City, New York. MACKLIN, DR. CHARLES C., School of Medicine, University of Western Ontario, London, Canada. MAGRUDER, DR. SAMUEL R., Department of Anatomy, Tufts Medical School, Bos- ton, Massachusetts. MALONE, PROF. E. F., College of Medicine, University of Cincinnati, Department of Anatomy, Cincinnati, Ohio. MAXWELL, DR. REGINALD D., Syracuse University, Syracuse, New York. MARSLAND, DR. DOUGLAS A., Washington Square College, New York University, Xew York City, New York. MARTIN, PROF. E. A., Department of Biology, Brooklyn College, Bedford Avenue and Avenue H, Brooklyn, New York. MAST, PROF. S. O., Johns Hopkins University, Baltimore, Maryland. 1 \TIIK\\S, PROF. A. P., Woods Hole, Massachusetts. MATTHEWS, DR. SAMUEL A., Thompson Biological Laboratory, Williams College, Williamstown, Massachusetts. MAVOR. I'ROF. J A. Mi'S W., Union College, Schenectady, New York. M A/.IA. DR. DANIEL. Department of Zoology, University of Missouri, Columbia, Missouri. MI:DI;S, DR. GRACE, Lankenau Research Institute, Philadelphia, Pennsylvania. MEIGS, MRS. E. I',., 1736 M Street, N.W., Washington, D. C. MKNKIN, DR. VALY, l-Yaring Krscarch Laboratory, Free Hospital for Women. 245 Pond Avenue, P.rooklinc, Massachusetts. MI.TX. PROF. CH \RI.I.S \\'., I'niviTsity of Pennsylvania, Philadelphia, Pennsylvania. REPORT OF THE DIRECTOR MICHAELIS, DR. LEONOR, Rockefeller Institute, 66th Street and York Avenue, New York City, New York. MILLER, DR. J. A., Division of Anatomy, College of Medicine, University of Ten- nessee, Memphis, Tennessee. MINNICII. PROF. D. E., Department of Zoology, University of Minnesota, Minne- apolis, Minnesota. MITCHELL, DR. PHILIP H., Brown University, Providence, Rhode Island. MOORE, DR. CARL R., The University of Chicago, Chicago, Illinois. MORGAN, DR. ISABEL M., Rockefeller Institute, York Avenue at 66th Street, New York City, New York. MORGULIS, DR. SERGIUS, University of Nebraska, Omaha, Nebraska. MORRILL, PROF. C. V., Cornell University Medical College, 1300 York Avenue, New York City, New York. MOSER. DR. FLOYD, Department of Biology, University of Alabama, University, Alabama. MULLER, PROF. H. J., Amherst College, Amherst, Massachusetts. NACHMANSOHN, DR. D.. College of Physicians and Surgeons, 630 W. 168th Street, New York City, New York. NAVEZ, DR. ALBERT E., Department of Biology, Milton Academy, Milton, Massa- chusetts. NEWMAN, PROF. H. H., 173 Devon Drive, Clearwater, Florida. NICHOLS, DR. M. LOUISE, Rosemont, Pennsylvania. NONIDEZ, DR. JOSE F., Cornell University Medical College, 1300 York Avenue, New York City, New York. NORTHROP, DR. JOHN H., The Rockefeller Institute, Princeton, New Jersey. OPPENHEIMER, DR. JANE M., Department of Biology, Bryn Mawr College, Bryn Mawr, Pennsylvania. OSBURN, PROF. R. C., Ohio State University, Columbus, Ohio. OSTERHOUT, PROF. W. J. V., Rockefeller Institute, 66th Street and York Avenue, New York City, New York. OSTERHOUT, MRS. MARIAN IRWIN, Rockefeller Institute, 66th Street and York Avenue. New York City, New York. PACKARD, DR. CHARLES, Marine Biological Laboratory, Woods Hole, Massachu- setts. PAGE, DR. IRVINE H., Lilly Laboratory Clinical Research, Indianapolis City Hos- pital, Indianapolis, Indiana. PAPPENHEIMER, DR. A. M., Columbia University, New York City, New York. PARKER, PROF. G. H., Harvard University, Cambridge, Massachusetts. PARMENTER, DR. C. L., Department of Zoology, University of Pennsylvania, Phila- delphia, Pennsylvania. PARPART, DR. ARTHUR K., Princeton University, Princeton, New Jersey. PATTEX, DR. BRADLEY M., University of Michigan Medical School, Ann Arbor, Michigan. PAYNE, PROF. F., University of Indiana, Bloomington, Indiana. PEEBLES, PROF. FLORENCE, Lewis and Clark College, Portland, Oregon. PINNEY, DR. MARY E., Milwaukee-Downer College, Milwaukee, Wisconsin. PLOUGH, PROF. HAROLD H.. Amherst College, Amherst, Massachusetts. POLLISTER, DR. A. W., Columbia University-. New York City, New York. : - s - I i • : • . • - . 3 s " - -- _ - • i - - - ' - ' ' • - rv,- " . " . J - . _~ ! _ 3. _— ~ - U "^-g _»— =• • — — we — i- ^L _1- -r -4L- — . il H H_ _. - :.: : r a. Zt - '. Z_ 36 MARINE BIOLOGICAL LABORATORY TE\YIXKKL, DR. L. E.. Department of Zoology, Smith College, Northampton, Massachusetts. TURNER. DR. ABBY H.. Wilson College, Chambersburg, Pennsylvania. TURNER, PROF. C. L., Northwestern University, Evanston, Illinois. TYLER, DR. ALBERT, California Institute of Technology, Pasadena, California. UHLENHUTH, DR. EDUARD, University of Maryland, School of Medicine. Balti- more. Maryland. UNGER, DR. W. BYERS, Dartmouth College. Hanover, New Hampshire. VISSCHER, DR. J. PAUL, Western Reserve University, Cleveland, Ohio. WALD. DR. GEORGE, Biological Laboratories, Harvard University, Cambridge, Massachusetts. WARD, PROF. HENRY B., 1201 W. Nevada, Urbana, Illinois. WARREN, DR. HERBERT S., 1405 Greywall Lane, Overbrook Hills, Pennsylvania. WATERMAN, DR. ALLYN J., Department of Biology, Williams College, Williams- town, Massachusetts. \Yi-:iss. DR. PAUL A., Department of Zoology, The University of Chicago, Chicago, Illinois. WENRICH, DR. D. H., University of Pennsylvania, Philadelphia, Pennsylvania. WHEDON, DR. A. D., North Dakota Agricultural College, Fargo, North Dakota. WHITAKER, DR. DOUGLAS M., P. O. Box 2514. Stanford University, California. WHITE, DR. E. GRACE, Wilson College, Chambersburg, Pennsylvania. WHITING, DR. PHINEAS W., Zoological Laboratory, University of Pennsylvania, Philadelphia, Pennsylvania. WHITNEY, DR. DAVID D., University of Nebraska, Lincoln, Nebraska. WICHTERMAN, DR. RALPH, Biology Department, Temple University, Philadelphia, Pennsylvania. WIEMAN, PROF. H. L., University of Cincinnati, Cincinnati, Ohio. WILLIER, DR. B. H., Department of Biology, Johns Hopkins University, Baltimore, Maryland. WILSON, DR. J. W., Brown University, Providence, Rhode Island. WITSCHI, PROF. EMIL, Department of Zoology, State University of Iowa, Iowa City, Iowa. WOLF, DR. ERNST, Biological Laboratories, Harvard University, Cambridge, Massachusetts. WOODRUFF, PROF. L. L., Yale University, New Haven, Connecticut. WOODWARD, DR. ALVALYN E., Zoology Department, University of Michigan, Ann Arbor, Michigan. WRINCH, DR. DOROTHY, Smith College, Northampton, Massachusetts. YNTEMA, DR. C. L., Department of Anatomy, Cornell University Medical College, 1300 York Avenue, New York City, New York. YOUNG, DR. B. P., Cornell University, Ithaca, New York. YOUNG, DR. D. B., 7128 Hampden Lane, Bethesda, Maryland. EMBRYONIC GROWTH IN THE VIVIPAROUS POECILIID, HETERANDRIA FORMOSA1 NEVIN S. SCRIMSHAW (From the Biological Laboratories, Harvard University, Cambridge) In Heterandria jormosa the fertilized eggs are minute and the embryos secure their nourishment for development through a placental type of association with the mother. This paper presents an analysis of embryonic growth in this viviparous species and a comparison of this growth with that in oviparous fishes. In the dis- cussion an attempt is made to evaluate the factors restricting growth in Heterandria. A similar study of embryonic growth in a number of ovoviviparous fishes is to follow. Bailey (1933) and Turner (1937) have directed attention to the development of embryos in poeciliid fishes in relation to the mother. Turner (1940a, b, c, d) surveyed the various types of association between mother and embryo for the four viviparous Cyprinodont families (Poeciliidae, Anablepidae, Goodeidae, and Jenynsiidae). For the oviparous fishes, the work of Gray ( 1926, 1928) on the development of the trout Salino jarlo stands almost alone. However, Kronfeld and Scheminzki ( 1926) have also contributed data on the trout. This work on a fish species totally dependent upon yolk for its embryonic nourishment provides the basis for the com- parison and interpretation of many of the observations described below. MATERIAL Heterandria fonnosa is remarkable in showing a high degree of superfetation. As many as eight broods of developing embryos occur within the ovary of a single female. Active sperm are retained in the ovary for many months following a single copulation. Thus it is impossible to determine the exact time of fertilization or the time it has taken any particular brood to reach the stage in which it is found. Its embryonic development resembles in its general aspects that of Fundulus and other Cyprinodont fishes. In Heterandria, as in all poeciliid fishes, the em- bryos are retained until parturition within the follicles of the single median ovary. Following fertilization the ovum and the follicle become separated by the perivitel- line space and fluid. The follicular membrane is henceforth generally referred to as the ovisac. Its diameter increases with the extension of the fluid filled spaces associated with the embryo and with the growth of the embryo itself. Sections of the immature ova show little yolk, but numerous small oil globules are already present (Fig. 1). The latter increase in volume as the eggs grow, and their number is reduced by the coalescence of the small vacuoles (Fig. 2). The large oil globules fuse at the time of fertilization to form a single large spherical oil 1 The author expresses appreciation to Dr. Leigh Hoadley for suggestions and encourage- ment in this work. 37 38 NEVIN S. SCRIMSHAW mass which occupies from 50 to 75 per cent of the volume of the ripe ovum (Fig. 3). When the blastodisc appears, it 1 tears the same relation to the oil globule in Heter- anclria as it does to the yolk mass in the more typical teleost egg. Thus a central perihlast is observed beneath the blastodisc and a thin syncytial periblast eventually surrounds the oil globule in precisely the same manner as it surrounds the yolk in Fundulus. The chemical composition of the oil globule has not been determined, but it >tains deeply with the fat dyes Sharlach R and Sudan III. The great reduction in the amount of the other yolk components present would suggest that the food of the embryo during the first part of its development might be derived from the oil. Figure 4 shows the relation between the diameter of the ovisac and the volume of FIGURE 1 FlGURE 3 FIGURE 2 FIGURE 1. Karly ovum of Ilclcrandria fonnnsa. O, oil; C, cytoplasm; N, nucleus; F, follicular epithelium; T, theca. FIGURE 2. Late ovum of Heterandria formosa. FIGURE 3. Early blastodisc of Heterandria formosa. B, blastodisc; Y, yolky periblast; Ch, chorion. the oil globule. It indicates that the volume is decreased by less than 30 per cent up to the time that the pseudo-placenta is fully formed. Fraser and Renton (1940) and Turner (1940a) have described the fetal adap- tations associated with viviparity in Heterandria. These are unique and may be summarized as follows: The antero-ventral somatopleure of the embryo becomes enormously extended and encloses the head completely. This produces a large extraembryonic pericardial cavity extending into the head fold. The postero- ventral somatopleure is also somewhat expanded and contains the liver, gut and the much extended urinary bladder. An extensive capillary network develops on this expanded portion of the somatopleure so that its whole surface opposing the hillicular epithelium is highly vascularized. Since the latter has likewise developed an extensive' vascular system, an association between embryonic and maternal cir- culation comparable to a non-deciduate mammalian placenta is formed. EMBRYONIC GROWTH IN HETERANDRIA 39 In relatively advanced embryos the pericardia! hood ruptures over the forepart of the head, leaving a strap of somatopleure over it. This strap gradually decreases in size and eventually disappears. As a result the area of the vascular association between embryo and ovisnc has been considerably reduced by the time of parturi- tion. Special types of cells in connection with the maternal and embryonic com- ponents of the pseudo-placenta also appear. Turner ( 1940a) has stated that the outer layer of the extra-embryonic pericardia! cavity, the ectoderm, is equipped with "conical cells, which are granular at the base and vacuolated apically . . . and sometimes gathered into low mounds." Their role has not been determined but it .02 0 . -J o j.01 o O.008 3006 _J O oo^ o o CO OIL GLOBULE DISAPPEARANCE HETERANDRIA FORMOSA o o O O 00000 o o o °o 0° CIRCULATION ESTABLISHED o °00 o o O> o c o .6 8 i Q 2 6 DIAMETER OF OVISAC IN MM. 2.0 FIGURE 4. The volume in cubic millimeters of the oil globule in fertilized eggs and embryos of Heterandria is plotted as a function of the diameter of the ovisac in millimeters. Each point represents a single pair of readings. is assumed that they aid in some way the nourishment of the embryo through the pseudo-placenta. Another cell specialization has not been described previously but seems to be- come important about the time that the head breaks through the enveloping peri- cardia! sac. It probably represents a supplementary nutritive mechanism which helps to compensate for the reduction in association surlace. Sometime during the last quarter of development, the wall of the ovisac. which has appeared as a squa- mous epithelium since fertilization, hypertrophies. In a restricted region of the ovisac prominent secretory cells are developed. These seem to form an approxi- mately one-third of the surface of the ovisac and parallel more or less closely the embryonic vascular surface. In Figure 5 these cells are seen opposite the blood vessels in the wall of the pericardium, but they may also occur opposite nonvascu- larized body epithelium. Histologically they appear to be active secretory cells 40 Xi'YIX S. SCRIMSHAW o ' 1 J 1 n,i KK 5. Photomicrograph of a section through the ovisac and pericardium of a later o of Hcterandria fonnosii showing secretory cells of the ovisac epithelium. Ov. Ep., ovarian epithelium; Th., theca ; Fol., follicular epithelium; Sec. Ep., secretory portion of fol- licular epithelium; Bl. V., hlood vessel; Per. Cav., pericardial cavity; Ht., heart. containing many secretion granules. Coagulum which may be from this source ap- pears in sections through the cavity of the ovisac and also in the gut. AT FT HODS Both wet and dry weights were determined for members of each brood of em- bryos taken from 16 females with varied previous histories. One hundred and fifty-one embryos were weighed either individually or in groups belonging to the same brood. The embryos were dissected Irom the ovary with the ovisac intact. 'Ilie diameter- of the ovisac was measured with an ocular micrometer. The em- bryos were allowed to touch fine absorbent paper to remove surplus fluid and im- mediately placed on flamed platinum squares of known weight. These were weighed at once in a single rapid trial to determine the wet weight of the embryo in the intact ovisac. They were then dried over calcium chloride to constant weight. All weighings were made with a Sartorms balance. The initial weights of the platinum squares and the final weights of the squares plus dried embryos were each determined by 10 to 20 readings (with zero correction before and after each reading). The standard deviation1'1 for a given series of readings was rarely over .05 mg. In several trial cases two weeks in the dessicator intervened between the first and second group of ten readings. The standard deviation was still below .<>5 mg. The smaller embryos were weighed in groups. The average wet weight ot oO o! the earliest embrvos was 0.026 mg. After a consideration urces of error, is spherical at first hut hecoines increasingly ellipsoidal as development pn>- In tin- late stages tin1 average of several measurements made on tin- >amc ovisac proved to 1» nalily consistent measure of si/.e and \\as recorded as il it were a true diameter. EMBRYONIC GROWTH IN HETERANDRIA 41 this value was estimated to be within at least 0.008 mg. of the true wet weight at the time of fertilization. The dry weight value of 0.017 mgs. was similarly estimated to be within at least 0.002 mgs. of the correct value for this stage. In all of the weight measurements except those on the very early stages the experimental error was small compared to the normal variation of the material. Ash determinations are impractical for early stages because of the extremely small weights involved. Later stages were ashed in a small muffle furnace and an average ash content of 2 to 3 per cent of the dry weight found. WEIGHT CHANGES OF THE DEVELOPING EMBRYO Since the age of the embryos was not known, the diameter of the ovisac was se- lected as a convenient indication of the stage of development, and the weight deter- minations were correlated with this dimension. When this was done the great increase in weight of the embryo in the intact ovisac was apparent (Fig. 6). The relationship between dry weight and ovisac diameter is repeated on a larger scale for later comparisons (Fig. 7). Before any appreciable gain in dry weight can be detected the wet weight can be seen to have increased markedly. In fact, the water content of the whole egg 6.0 5.0 <4.0 _ _J 5 ao 2.0 i -. .a .6 .4 .2 EMBRYONIC GROWTH HETERANDRIA FORMOSA WET WT. DRY WT^ .? .4 .6 .8 . DIAMETER OF OVISAC IN MM. 3.0 FIGURE 6. The curves show the increase in both wet and dry weights of developing em- bryos of Hetcrandria formosa plotted as a function of the diameter of the ovisac, a convenient measure of development. plus ovisac increases from 34.6 per cent to 86.8 per cent in this period. There is considerable experimental error in the first figure because of the small weights in- volved, but a value of 35 per cent is consistent with the large amount of oil present. After the dry weight has begun to increase, the relation between wet and dry weight stays remarkably constant. When the period was divided into two sub- 42 NEVIN S. SCRIMSHAW periods, the first including embryos \vith ovisac diameters between 1.00 and 2.00 and the second including embryos with ovisac diameters greater than 2.00, the dif- ferences in per cent dry weight were not significant. During the first sub-period the dry weight average 12.84 per cent of the wet weight (o- 1.8$?) and during the second' it was 13.49 per cent (o- 1.9$;). The average was 13.2 per cent. After parturition the embryo breaks out of the fluid-filled ovisac and the extra liquid in the expanded pericardial cavity disappears. Accordingly the percentage dry weight increases abruptly to a stable value for the young fish of 20.6 per cent (o- !./$.)• This information can be diagrammed to show the changes in com- position from the egg at the time of fertilization to the larva at the time of parturi- tion (Fig. 8). 2-6 < o: o 3-5 I- • 1 u 3.3 EMBRYONIC GROWTH HETERANDRIA FORMOSA c o o o DRY WEIGHT D-O-O? .2 A .6 .8 |_Q 2 A .6 .6 2.0 2 .4 .6 DIAMETER OF OVISAC IN MM. .8 3.0 FIGURE 7. The curve shows the increase in the dry weight of embryos of Hctcrandria formosa plotted as a function of the diameter of the ovisac. This represents the same weight values shown in the dry weight curve of Figure 6, but the units of the weight ordinate are much expanded. These changes can also be represented in the form of a generalized equation which shows the relation of .1 mg. of ovum to the final weight of the larva. 0.1 mg. (wet weight of egg) 2.6 mg. (dry weight from mother) 15.8 mg. (wet weight from mother 18.5 mg. (wet weight of young fish) This equation expresses the overall changes. A similar expression has been de- veloped for the trout by Gray (1926). His formula holds from the time the em- bryo is 50 per cent developed to the time it is 80 per cent formed during which time it has increased about 400 per cent in weight. During this period the trout embryo converts 1.0 grams of yolk (wet weight) plus 0.7 grams of water into 1.56 grams of fish. EMBRYONIC GROWTH IN HETERANDRIA 43 Since the yolk makes up almost the entire egg, it can be estimated that one gram of trout egg makes only slightly more than 1.5 grams of trout embryo. In com- parison one gram of Heterandria egg would result in 185 grams of embryo. This strikingly illustrates the importance of the maternal contribution of nourishment in this species. .007 .006 Z 004 X O LJ .003 .002 .001 WATER IN OVISAC PERIC"ARDIAL CAVITY [LOST AT BIRTH] .00228 GRAMS WATER IN EMBRYO .00410 GRAMS \\\\\\\\\x COMPOSITION OF EGG AND EMBRYO HETERANDRIA FORMOSA EMBRYO AT TIME OF PARTURITION EGG FERTILIZATION WATER FROM EGG.00003 / .000009 GR. .NON-EMBRYO DRY WT. .00015 GR. EMBRYO WT. 00066 DRY WT. FROM E GG «— .OOOOI 7 GR O WATER .OOOOO9 GR. DRY WEIGHT YOLK, O I L 8. PROTOPLASM .000017 GR. FIGURE 8. The composition of a typical embryo ready for parturition and of a recently fertilized egg is diagrammed. Wide variations from the actual figures cited were found but the proportions remained generally the same in the various larvae studied. As discussed in the text, the weight figures and the relative amounts of wet and dry material in the fertilized egg have been determined only approximately. In the above figures for the trout the discrepancy of .14 grams between the 1.7 grams of yolk and water and the resulting 1.56 gram weight of the larva represents the dry weight used for maintenance metabolism. The dry weight used for main- tenance in Heterandria could not be measured directly. However, it can be as- 44 NEVIN S. SCRIMSHAW sumed that approximately one-third of the total food available is used for the maintenance of the embryo in Heterandria as well as in the trout.1 The main- tenance dry weight calculated in this way can be introduced into an equation ex- pressing the actual dry weight drain on the mother per embryo. water + 6.8 mg. + 2.2 mg. — 9.0 mg. -f- water (dry weight (maintenance (total dry of embryo) dry weight) weight from mother) GROWTH RATE OF THE DEVELOPING EMBRYO Thus far the magnitude of the embryonic weight increase has been described, but no indication of the rate has been obtained. Observations of living gravid females suggested a method whereby a time axis might be obtained. The young of females kept under constant conditions showed a tendency to be born at regular intervals.5 Accordingly, the weights of all the broods of a single female kept under relatively constant conditions were plotted as if the time intervals between them were the same. This treatment yielded consistent and apparently significant growth curves. The female whose embryos are presented in Figure 9 has the unusually large number of eight broods. The growth relationships believed to be general for Heterandria are therefore well shown. The initial portion of the curve is a straight line which appears to be parallel to the time axis. This line represents the early developmental period when no post-fertilization contribution of nourishment from the mother can be detected. The second part of the curve is a steeply ascending straight line. It extends throughout the greater part of the embryonic period and represents an approximately constant growth increment in dry weight contributed by the mother. Obviously, there must be a transitional period, which the data can- not show, during which the pseudo-placental associations are being established. This has been indicated by the dotted lines in Figure 9 and also in Figure 10 which represents the similar treatment of embryo weights obtained in the study of other females. All of the females which had been kept under relatively constant condi- tions showed this relationship. Unfortunately, the concept cannot be tested for females with fewer than five broods as can be seen from the number of points required to establish the two straight lines in Figures 9 and 10. Tn order to extend the study to more females, records of ovaries dissected over a three year period were re-examined. These records gave the number of broods per female together with their average ovisac diameters for fish kept under a variety ot experimental conditions. The average weight of the broods in these females 1 The ratio between the dry weight of the embryo and the dry weight of nutriment required to produce this quantity of embryo is the efficiency coefficient of development. Gray reports a value of .65 for the trout and points out that this figure has been found to be approximately the same for a wide variety of organisms (cf. Murray, 1926, chick; Hayes, 1930, Atlantic salmon: etc.). Unpublished data. It is difficult to obtain precise records because the young arc fre- quently eaten by the parent and arc so small as to be easily missed. Rather constant optimal conditions and mature females in good health are required to demonstrate this. Small females will not show it dearly. On the whole it is remarkable that this tendency toward even spacing of the broods is as frequent and demonstrable as the data suggest it to be. EMBRYONIC GROWTH IN HETERANDRIA 45 3 O .3 EMBRYONIC GROWTH HETERANDRIA FORMOSA PRE-PARTURIT1ON EMBRYO (ASSUMING BROODS EQUALLY SPACED IN TIME] FEMALE 1-22-40 C FERTILIZED EGG 6 ROOD G E D C RELATIVE TIME FIGURE 9. The average dry weight of the embryos in each brood of a female with eight fertilized broods is plotted as a function of time. Since the actual age of the embryos cannot be determined, no specific time units can be employed. However, in any large, mature female of Heterandria kept under reasonably constant conditions the broods of embryos seem to be born at approximately equal intervals. Therefore, the broods of embryos still in the ovary can be equally spaced on a time axis without the necessity of specifying the actual number of days or hours between them. Since the time relations between the various broods contained in the ovary are thus represented without using known time units, a relative time scale is actually employed. This concept of relative time is also used in Figures 10-12 and throughout the text. The female has been kept under constant illumination in a room with only slight variations in temperature. < 5 o o u $ EMBRYONIC GROWTH HETERANDRIA FORMOSA DRY WEIGHT, TIME o (ASSUMING BROODS EQUALLY SPACED IN TIME] / BROOD E' C' B' E A' D REL ATIVE TIME FIGURE 10. The average dry weight of the embryos in each brood of two different females, one with five and the other with nine broods, is plotted as a function of relative time. The treatment is the same as in Figure 9. NEVIN S. SCRIMSHAW could be estimated liy comparison of the ovisac diameters with the average weight of the embryos as .shown in Figure 7. The results were plotted exactly as in the above cases where the weights were obtained directly. Despite the wide variety of ages and conditions represented, the majority showed the generally linear nature of growth in the two periods. Data from three such females are presented in Figure 11. This even spacing can also be demonstrated by the superposition of growth curves from different females. Numerous females were examined by this method and their intermediate broods were found to fall on the growth curves of other females. For example, Figure 9 shows the growth curve for the embryos of a female with eight broods. Broods B and E of this female (second and fifth broods respectively) were found to correspond in weight to the first and third broods of EMBRYONIC GROWTH HETERANDRIA FORMOSA DRY WEIGHT/ TIME 6 J a f .5 [ASSUMING BROODS EQUALLY SPACED IN TIME] BROOD C' D' E C' 0 RELATIVE TIME FIGURE 11. The average dry weight of the embryos in each brood of t\vo different females, "in- with seven and the other with five broods, is plotted as a function of relative time. The dry weight values have been obtained by use of the ovisac diameter and the dry weight curve • if Fi.gure 7. The broods are assumed to be evenly spaced in time. a four-brooded female. When the second brood of the four-brooded female was arbitrarily placed midway between broods B and E on the time axis, it was found to have a weight value which placed it on the growth curve representing the em- bryos of the eight-brooded fish. The diameter of the ovisac for each brood of an eight-brooded female can itself be plotted as a function of relative time (Fig. 12). The resulting curve is smooth and sigmoid, showing a slower diameter increase during the early and late periods. In the female represented in Figure 12 the earliest brood was observed to represent recently fertilized ova, and the latest brood was found to represent embryos ready tor parturition. Hence a point midway on the relative time axis should indicate the ovisac diameter of an embryo roughly half way through development. It is possible in this fashion to determine the percentage of development represented by other ovisac diameters. EMBRYONIC GROWTH IN HETERANDRIA 47 In the course of these studies on Heterandria the specific stage of development corresponding to various ovisac diameters has been noted. It is now possible to assign the percentages of development determined for certain ovisac diameters to developmental stages in Heterandria. These can then be compared with similar stages described in closely related fish in which a more direct measure of time can be obtained. Bailey (1933) faced with similar difficulties in determining the exact time of development in the poeciliid Xiphophorus licUcri, staged 50 ova and em- bryos and selected the tenth, twentieth, thirtieth, etc. as representing corresponding percentages of development. Fitiuiulits heteroclitus was studied at the Marine Biological Laboratory, Woods Hole, in July 1939 and together with Xiphophorus was compared with Heterandria. The relative times between common stages were 2.8 2.4 2.0 z o: h u 1.6 1.2 < .8 Q .4 OVISAC DIAMETER/" /^DEVELOPMENT TIME [ASSUMING BROODS EVENLY SPACED IN TIME] -O PARTUR I TION ) CIRCULATION ^FERTILIZATION ^LIBRARY 25% PERCE N TAGE OF 75°7o DEVELOPMENT 100% FIGURE 12. The average ovisac diameter of each brood in a female with eight broods is plotted as a function of percentage of development. This is the same female shown in Figure 6 and the time units are obtained in the same manner. However, since the entire range from the fertilized egg to the embryo ready for parturition is represented, these extremes are indicated as 0 per cent and 100 per cent of development and the relative time axis is subdivided accordingly. clearly similar in the three species. Although the comparison could be only ap- proximate,6 it reduced the likelihood of a serious error in the time relations assigned to Heterandria. If these time relationships are used to determine the increments of growth per unit of time during the later embryonic period, it will be seen that the increments remain constant. When the percentage of maximum increment of dry weight is plotted as function of time in Heterandria (Fig. 13), the result is a straight line parallel to the time axis throughout most of the embryonic period. Gray, treating 6 In addition to the difficulties of comparing stages described for different species, the rela- tive time between embryonic stages may be changed by exposure to different temperatures. At the same temperature the relative time between similar stages in two different species may be different (Moore, 1939; Worley, 1933). 48 NT.VIX S. SCRIMSHAW his growth rate data for the trout in this manner, demonstrated an asymmetrical rise and fall (Fig. 13). This represented a deviation from Robertson's formula (1923) in which the percentage of maximum increment plotted as a function of the size of the embryo shows a symmetrical rise and fall. This difference Gray was . 0 75 o X o 5 50 D 5 X. _ * z UJ O a: u a ?S 50 75 PERCENT Or LARVAL PERIOD 100 FIGURE 13. The percentage of maximum increment of embryonic growth is plotted as a function of the per cent of the larval period for both Hctcrandria formosa and Salmo fario. No measurements are available for the early stages of Heterandria and the growth increments are necessarily inferred. The curve for Salmo also represents the product of the dry weight of the embryo times the dry weight of the remaining yolk plotted as a function of the per cent of larval development. able to explain by demonstrating that the growth rate is also a function of the yolk remaining in the yolk sac. DISCUSSION The fact that Heterandria formosa represents the development of a true vivi- parity in which nearly all of the nourishment for embryonic development comes from the mother is of interest in itself. When the assumption of equal spacing of broods is made and the data treated accordingly, it also appears that the nourish- ment for the growth process is being used by the embryo at a constant rate. If this is true, it suggests that some specific extrinsic factor or factors is limiting embryonic growth in this species. Restricted food supply and limited oxygen availability were considered likely to affect the growth rate in this manner. The discussion which follows is an attempt to evaluate these two factors. It has been shown that morphogenesis in the oviparous fish adjusts itself to the amount of nourishment available, i.e. that the size of the larva is dependent on the amount of yolk available (Morgan, 1896) and not on the total amount of cyto- plasm (Sunnier. 1900; Hoadley, 1928). That the rate of growth of the oviparous lish embryo is dependent not only on the mass of the embryo but also on the actual amount of yolk remaining has been shown by Gray (1926; 1928a, b) (cf. Fig. 12). EMBRYONIC GROWTH IN HETERANDRIA 49 It is conceivable that the pseudo-placental barrier itself may increase in effec- tiveness only enough to allow for the steadily increasing demands of the main- tenance metabolism. Such a relationship seems rather remarkable under the cir- cumstances and might he expected to break down with the complications of retraction of the pericardial sac and development of special secretory cells in the follicle wall. No alteration in growth rate can be detected when these changes in the pseudo-placental barrier occur. Instead of the limitation lying in this barrier, it may be that certain of the raw materials for growth are present in the maternal blood stream in limited amounts. How such a limitation could affect all of the broods in a similar manner is not clear. There is, however, some indirect evidence that the growth of the individual embryo is responsive to changes in the total maternal supply of nourishment available to all the embryos. This evidence involves the young of fish not heavily burdened with embryos. Since the straight line nature of the growth curve cannot be tested unless more than four broods are present, all of the data used to develop the idea of a constant in- crement of growth have been obtained from females with many broods. In these the food requirements of the embryo must constitute a great drain on the mother. The physiological drain would not be as great in females which are recovering from unfavorable conditions, because they contain fewer broods than they are capable of supporting. In these females more nourishment per embryo should be available than in females which have been kept under relatively constant conditions. If more nourishment is available, the embryos should be larger at the end of the larval period (cf. Gray, 1928a). The actual results in Heterandria are that the first young born of a female recovering from unfavorable conditions are large.7 This constitutes the best available evidence that the food supply of the embryos in a many brooded female is restricted in some manner, and supports the hypothesis that during the main growth period food may be the determining factor in the develop- ment of Heterandria at normal temperatures. There is considerable evidence that at higher temperatures the oxygen supply to the embryo may be a factor limiting growth. According to Gray, all of the oxygen used by the trout embryo is for maintenance metabolism, the amount used for growth being almost negligible. He also found that a large drop in the growth rate of the trout caused no corresponding drop in oxygen consumption. Nevertheless, Jacques Loeb showed (1894) in Fundulus that development is directly retarded by lack of oxygen. The Heterandria females may under certain conditions at 25° temperature give birth to young regularly every four days. A rise in temperature of two or three degrees during the daytime for even two or three days may delay the next brood.8 It seems likely that the decreased oxygen supply may account f or 7 The following figures will not be discussed in detail but are presented in support of the statements made above. A female recovering from unfavorable conditions was observed to give birth to young on Dec. 13 and 14 after a lapse of several months. These young were very much larger than those born four and ten days later. The actual dry weights were found to be : Dec. 13 and 14, 1940 Dec. 18 and 26, 1940 .88 mg. .63 mg. .95 mg. .64 mg. .89 mg. .61 mg. 8 Scrimshaw — unpublished data. 50 NEVIN S. SCRIMSHAW this. At the higher temperatures both the mother and the embryos require more oxygen per unit time, but less oxygen is dissolved in the water. It also seems likely (ci. Irving, 1941) that the oxygen dissociation curve of the hemoglobin in the blood would be shifted to the right and flattened, and as a result the oxygen carrying power of the blood would be reduced by the increased temperature. It can also be observed that a female kept at a constant temperature of 25° C. will have a number of its embryos dying within several hours when the temperature is raised to 28° C.:) Such a temperature is not in itself supra-maximal, for the temperature tolerance of the mother runs well above 34° C. The young after birth growr well at this high temperature. The embryos at all stages of development seem to tolerate this temperature satisfactorily when isolated from the mother if the water is well aerated. For example, their heart rate shows no irregularity in IJL value on an Arrhenius plot until the temperature reaches 34.6° C. The death of the embryos in the above case can be explained on the basis of limited oxygen supply to them. Oxygen supply is not likely to be the limiting factor at ordinary temperatures. \Yhen exposed only to natural daylight, a female kept at a constant temperature and under approximately uniform feeding conditions will contain a certain number of embryos and these will show the constant growth increment described. When such a female, other conditions remaining the same, is exposed to continuous arti- ficial light for about a month, the number of embryos markedly increases.10 There is no reason to believe that the total availability of oxygen has significantly in- creased. It seems reasonable to believe that the oxygen is the principal limiting factor at higher temperatures, and food supply at the moderate ones. This would mean an intermediate range in which the two factors are complementary in a regular fashion. The data do not serve to distinguish between different degrees of limitation. Fur- thermore, Heterandria kept at temperatures high enough to limit the oxygen supply do not have enough broods to enable the growth rate to be determined. The pos- sibility of other factors such as endocrine balance influencing the growth rate has not been excluded. SUMMARY The fertilized egg of the viviparous poeciliid Heterandria jonnosa is minute and is made up almost entirely of a single large oil globule. At least 70 per cent of the original volume of the oil globule is still present when maternal contribution of nourishment begins. Secretory cells develop in the ovisac wall late in the em- bryonic period. These together with adaptations previously described permit the mother to contribute nearly all of the raw materials for growth and development of the embryo after the egg has been fertilized. The increase in wet and dry weight of the embryos at the expense of the mother has been determined by obtaining both wet and dry weights at various stages of development. The dry weight of the embryo increases from 0.017 milligrams at the time of fertilization to (>.X milligrams at the time of parturition. The percent- imshaw— unpublished data. 1IJ Scrimshaw — unpublished data. EMBRYONIC GROWTH IN HETERANDRIA 51 age dry weight remains constant at 13.2 per cent after the pseudo-placental associa- tion is established. Observations of living females suggested that tinder constant optimal conditions the broods of a single female tend to be evenly spaced in time. Upon this assump- tion the dry weights of the embryos in each brood of suitable females were plotted against relative age. The resulting curves suggested that the rate of growth after the maternal contribution of nourishment can be detected is approximately constant. Food supply and oxygen supply are discussed as factors which might limit this growth. Embryonic growth in Heterandria is compared with that in oviparous fishes. LITERATURE CITED BAILEY, R. J., 1933. The ovarian cycle in the viviparous teleost, Xiphorphorus helleri. Biol. Bull., 64: 206-225. ERASER, E. A., AND R. M. RENTON, 1940. Observations on the breeding and development of the viviparous fish, Heterandria formosa. Quart. Jour. Micr. Sci., 81 : 479-520. GRAY, JAMES, 1926. The growth of fish: I. The relationship between embryo and yolk in Salmo fario. Brit. Jour. E.vp. Biol., 4: 215-225. GRAY, JAMES, 1928a. The growth of fish: II. The growth rate of the embryo of Salmo fario. " Brit. Jour. £.r/>. Biol., 6: 110-124. GRAY, JAMES, 1928b. The growth of fish : III. The effect of temperature on the development of the eggs of Salmo fario. Brit. Jour. E.vp. Biol., 6: 125-130. HAYES, F. R., 1930. The metabolism of developing Salmon eggs. I. The significance of hatch- ing and the role of water in development. Biochcm. Jour., 24: 723-734. HOADLEY, LEIGH, 1928. On the localization of developmental potencies in the embryo of Fundulus heteroclitus. Jour. Exp. Zool., 52 : 7-44. IRVING, L., E. C. BLACK, AND V. SAFFORD, 1941. The influence of temperature upon the com- bination of oxygen with the blood of the trout. Biol. Bull., 80: 1-17. KRONFELD, P., AND F. SCHEMINZKI, 1926. Beitrage zur Physikalischchemischen Biologic der Forellenentwicklung. 2. Mitteilung : Wachstum, Dotterresorption und Wasserhaushalt. Arch. f. Entw. Mcch. 107 : 129-153. LOEB, JACQUES, 1894. Uber die relative Empfindlichkeit von Fischembryonen gegen Sauer- stoffmangel und Wasserentziehung in verschiedenen Entwicklungstadien. Arch. ges. Physiol, 55: 530-541. MOORE, J. A., 1939. Temperature tolerance and rates of development in the eggs of amphibia. Ecology, 20 : 459-478. MORGAN, T. H., 1896. The formation of the fish embryo. Jour. Morph., 10 : 419-472. MURRAY, H. A., 1926. Physiological ontogeny. A. Chicken embryos. VIII. The concen- tration of the organic constituents and the calorific value as functions of age. Jour. Gen. Physiol., 9 : 405-432. ROBERTSON, T. B., 1923. The Chemical Basis of Growth and Senescence. Philadelphia. SUMNER, F. B., 1900. A study of early fish development. Arch. f. Entu: Mech., 17 : 92-149. TURNER, C. L., 1937. Reproductive cycles and superfetation in poeciliid fishes. Biol. Bull., 72 : 145-164. TURNER, C. L., 1940a. Pseudoamnion, pseudochorion and follicular pseudoplacenta in poeciliid fishes. Jour. Morph., 67 : 58-89. TURNER, C. L., 1940b. Follicular pseudoplacenta and gut modifications in anablepid fishes. Jour. Morph., 67 : 91-105. TURNER, C. L., 1940c. Pericardial sac, trophotaeniae, and alimentary tract in the embryos of goodeid fishes. Jour. Morph., 67 : 271-289. TURNER, C. L., 1940d. Adaptations for viviparity in Jenynsiid fishes. Jour. Morph., 67 : 291- 297. WORLEY, L. G., 1933. Development of the eggs of the mackerel at different constant tempera- tures. Jour. Gen. Physio!., 16 : 841-857. THE CAPILLARY BED OF THE CENTRAL NERVOUS SYSTEM OF CERTAIN INVERTEBRATES ERNST SCHARRER (From the Department of Anatomy, Western Reserve University School of Medicine, Cleveland, and the Marine Biological Laboratory. \Voods Hole) * The vascular pattern of the vertebrate brain may be either one of two types : the one consists of single vessels that anastomose to form a continuous capillary network; the other consists of paired vessels that end in capillary loops. These two types are, as a rule, mutually exclusive, except in the lungfish, Epiceratodus (Craigie, 1943), and in the salamander, Ambystoma (Craigie, 1938a), where both networks and loops occur. The network pattern is the more common type. It is found in monotremes (Sunderland, 1941) and in all placental mammals, in reptiles with the exception of the lizards, in anuran amphibians, and in the fishes including the hagfish, Myxine. The paired vessels ending in capillary loops are character- istic of the marsupials (Wislocki and Campell, 1937; Craigie, 1938b; Sunderland, 1941), the lizards (Schobl, 1878; Sterzi, 1904) including Sphenodon (Craigie, 1941a), and the tailed amphibians (Schobl, 1882; Sterzi, 1904; Craigie, 1938a; 1939; 1940a) including the Gymnophiona (Craigie, 1940b ; 1941b). The brain of the lamprey, Petromyzon, is also supplied by loops (Craigie, 1938a). The study of patterns of cerebral vascularization has been extended here to in- clude invertebrates. In most invertebrates blood vessels do not enter the nervous tissue. There are, however, exceptions. Havet (1916), for instance, in his in- vestigation of the glia cells of the invertebrates mentions the existence of blood vessels within the central nervous system of the earthworm. Another reference may be found in Cajal's (1929) paper on the origin of unipolar neurons in inverte- brates according to which the cerebral ganglia of the squid are vascularized by intraganglionic blood vessels.- Both of these animals, the earthworm and the squid, have been studied, therefore, and the blood vessels supplying their ganglia have been compared with those of vertebrates. MATRRIAL AND METHODS Large earthworms (Lnmbrlcus lerrestris*) were collected on lawns in the Cleveland area during rainy nights and were fixed with Zenker-formol. The ring consisting of cerebral and subesophageal ganglia and their connectives was em- bedded in paraffin and cut 7 micra thick. The sections were stained with Masson's 1 This research was aided by a grant made to Western Reserve University by the Rocke- feller Foundation. 2 The papers of Williams (1902) and Grimpe (1913) give excellent accounts of the vascular system of cephalopoda, but do not include descriptions of the vascularization of the cerebral ganglia. " Lunibncus terrestris was introduced from Europe and has become widely distributed in Ohio in the past 25 years (Eaton, 1942). 52 CAPILLARIES OF INVERTEBRATE GANGLIA 53 •j I FIGURE 1. A pair of branch ing blood vessels in the central nervous system of the earth- worm. Zenker-formol, paraffin, 7 micra, Masson's trichrome stain. Photomicrograph, X 350. FIGURE 2. A pair of blood vessels in the brain of the opossum branching in the same man- ner as those of the earthworm shown in Figure 1. Injection with India ink-gelatin, formalin, nitrocellulose, 100 micra. Photomicrograph, X 350. FIGURE 3. Terminal loops in the central nervous system of the earthworm. Technique and magnification as in Figure 1. FIGURE 4. Terminal loops in the brain of the opossum. Technique and magnification as in Figure 2. 54 I RNST Si II AKK1-K trichromc stain. I'robably because of the strong contraction of the animals when the fixing fluid is injected into the body cavity, the central ^an^lia sometimes be- come very hyperemic. In such animals the blood vessels ot the central nervous system are tilled with the- blood fluid which stains well with the red component of the Masson stain i see H.US. 1 and o). 5 1 n, i KI 5. Capillary network in tin- cerebral ganglion of the squid. Injection with India ink-gelatin, formalin, nitrocellulose, Kid micra. Photomicrograph, '.-• 220. FIGI RI 6. ( apillary network in the brain of the rat. Technique and magnification as in e 5. Squids I l./>(/li /v5 compared with those in various vertebrates, including monkey, cat, guinea pig, rat, opossum, alligator, and several species of teleosts. These were all injected with carmin- or India ink-gelatin, were embedded in nitrocellulose, and were sectioned 100 or 200 micra thick. OBSERVATIONS A comparison of the illustrations (Figs. 1 to 4) shows that the blood vessels of the central nervous system of the earthworm are of the same type as those of the opossum brain. In the earthworm blood vessels enter the tissue of the central nervous system in pairs. They divide together and their branches form corre- sponding pairs (Fig. 1). Finally the two limbs of each pair join and thus end in hairpin-like loops (Fig. 3). This is essentially the same arrangement which Wis- locki and Campbell (1937) described in the opossum where arteries and veins stay together in pairs after they have entered the brain tissue. Whenever an artery divides, the accompanying vein divides the same way (Fig. 2), and all blood vessels within the opossum brain end finally in non-anastomosing loops (Fig. 4). The vascular pattern of the central nervous system of the squid is entirely dif- ferent from that of the earthworm. In the squid arteries and veins enter the cerebral ganglia singly. Their branches form a network of anastomosing capil- laries (Fig. 5), just as in the brain of placental mammals (Fig. 6). DISCUSSION "Since the discovery by Schobl (1878) that there exist in reptiles two radically different types of cerebral vascular bed, one reticular and the other composed of independent, non-anastomosing capillary loops, the relationship between these two types has remained obscure and attempts to reconstruct the phylogenetic his- tory of this mechanism have been complicated rather than simplified by increasing knowledge of the occurrence of the loop arrangement in various vertebrate classes" (Craigie, 1941a, p. 263). The difficulties inherent in the application of the phylo- genetic concept to the cerebral vascular patterns are well illustrated by the fact that among the cyclostomes, a group of primitive vertebrates, one (Petromyzon) shows loops, another (Myxine) a network (Craigie, 1938a). The description presented here of loops in the earthworm and of a network in the squid only serves to accentuate these difficulties. An attempt is made here, therefore, to illustrate a common origin of both sys- tems by dispensing with the phylogenetic aspect altogether. The loop system is not considered as a primitive forerunner of the network pattern, but is presented as the result of a parallel development capable of differentiation and functional effi- ciency corresponding to that of the reticular type. The origin of the cerebral vascular system may be compared with that of the endocellular blood vessels in the large extramedullary nerve cells of certain fishes such as the swellfish, Spheroidcs maculatus. In young specimens each one of these cells is surrounded by a network of blood vessels. As the cells become larger in older animals the distance between the center of the cell and the blood vessels ap- parently becomes too great and the blood vessels enter the cytoplasm. Similarly the whole nervous system while still small could be vascularized by a network of superficial blood vessels (Fig. 7 AB). Such a condition actually obtains in the 56 ERNST SCHARRER B, FIGURE 7. Diagram illustrating the derivation of loops and network patterns from a com- mon origin. AB, primitive central nervous system vascularized by superficial network. A, and B,, vessels of the superficial network come to lie within the nervous tissue. There are two possibilities of further development: the blood vessels approach each other in the direction of the arrows (A,) and thus become paired, or they send out branches which anastomose (B,). In the one cast- loops are formed (A2), in the other a network results (B,). Both these types are capable of further development to more complex systems ( A; and B,.). CAPILLARIES OF INVERTEBRATE GANGLIA 57 central nervous system of Amphioxus and in the spinal cord of Petromyzon which are vascularized by networks of superficial blood vessels. With the increase in size of the central nervous system segments of vessels forming part of the surface net- work come to lie within the nervous tissue (Figs. 7 A1 and B,). From this stage both the paired vessels ending in loops (Fig. 7 A2) and the network (Fig. 7 B2), may be derived as indicated. Both types occur in invertebrates and vertebrates, and both become eventually highly complex in mammals (Figs. 7 A., and B.;). In this scheme the position of the animal in the phylogenetic order is not con- sidered. This means that the step from Aj to AL, or from Bt to B2 can be taken anywhere within the vertebrates or the invertebrates. Thus the earthworm fol- lows Aj to Ao, the squid B! to BL>. Among the cyclostomes Petromyzon follows Aj to Ao, Myxine Bt to Bo. Epiceratodus and Ambystoma combine the two pat- terns, a situation which is not illustrated in Figure 7, but which can easily be visualized. The question still remains: Which factors cause the cerebral blood vessels <>i the earthworm, of Petromyzon, of the opossum, etc. to differentiate as loops, and those of the squid, of Myxine, and of most vertebrates as a network. An answer to this question is to be expected from the study of the early development of cere- bral blood vessels and from the application of the methods of experimental em- bryology. SUMMARY The blood vessels supplying the central nervous system of the earthworm are of the same type as those in the brains of tailed amphibians, lizards, and marsupials, i.e. the blood vessels are paired and end in loops. The blood vessels in the cere- bral ganglia of the squid form a network like that which occurs in the brains of fishes, anuran amphibians, reptiles (except lizards), birds, and placental mammals. The origin of both systems is discussed. LITERATURE CITED CRAIGIE, E. H., 1938a. The blood vessels of the brain substance in sonic amphibians. Proc. Amer. Philos. Soc., 78: 615-649. CRAIGIE, E. H., 1938b. The blood vessels in the central nervous system of the kangaroo. Science, 88 : 359-360. CRAIGIE, E. H., 1939. Vascularity in the brains of tailed amphibians. I. Ambystoma tigrmum (Green). Proc. Amcr. Philos. Soc., 81 : 21-27. CRAIGIE, E. H., 1940a. Vascularity in the brains of tailed amphibians. II. Necturus maculosus Rafinesque. Proc. Amcr. Philos. Soc., 82 : 395-410. CRAIGIE, E. H., 1940b. The capillary bed of the central nervous system of Dermophis (Amphi- bia, Gymnophiona). Jour. Morph., 67: 477-487. CRAIGIE, E. H.^ 1941a. Vascularization in the brains of reptiles. II. The cerebral capillary bed in Sphenodon punctatum. Jour. Morph., 69: 263-277. CRAIGIE, E. H., 1941b. The capillary bed of the central nervous system in a member of a second genus of Gymnophiona — Siphonops. Jour. Anat., 76: 56-64. CRAIGIE, E. H., 1943. The architecture of the cerebral capillary bed in lungfishes. Jour. Comp. Neur.,79: 19-31. EATON, T. H., 1942. Earthworms of the Northeastern United States: A key, with distribution records. Journal of the Washington Academy of Sciences, 32: 242-249. GRIMPE, G., 1913. Das Blutgefassystem der dibranchiaten Cephalopoden. Teil I. Octopoda. ' Zcitschr. n'iss. Zool, 104: 531-621. 58 ERNST SCHARRKR UAVET, J., 1916. Contribution a 1'etude de la nevroglie des invertebres. Trab. Lab. Im<. Blol. Madrid, 14: 35-85. RAMON Y CAJAL, S., 1929. Signification probable de la morphologic des neurones des in- vertebres. Trab. Lab. Inv. Biol. Madrid, 26: 131-153. SCHOBL, J., 1878. Ueber eine eigenthiimliche Schleifenbildung der Blutgefasse im Gehirn und Riickenmark der Saurier. Arch. f. mikr. Anat., 15 : 60-64. SCHOBL, J., 1882. Ueber die Blutgefasse des cerebrospinalen Nervensystems der Urodelen. Arch. f. mikr. Anat., 20: 87-92. STERZI, G., 1904. Die Blutgefasse des Riickenmarks. Untersuchungen iiber ihre vergleichende Anatomic und Entwickelungsgeschichte. Anat. Hcfte, 1. Abt., 24 : 1-364. SUNDERLAND, S., 1941. The vascular pattern in the central nervous system of the monotremes and Australian marsupials. Jour. Conip. Ncur., 75 : 123-129. WILLIAMS, L. W., 1902. The vascular .system of the common squid, Loligo pealii. Amcr. Nat., 36 : 787-794. \VISLOCKI, G. B., AND A. C. P. CAMPBELL, 1937. The unusual manner of vascularization of the brain of the opossum (Didelphys virginiana). Anat. Rec., 67: 177-191. BIOLOGY OF THE CALIFORNIA SEA-MUSSEL (MYTILUS CALIFORNIANUS). III. ENVIRONMENTAL CONDI- TIONS AND RATE OF GROWTH l WESLEY R. COE AND DENIS L. FOX (Scripps Institution of Oceanography, University of California, La Jolla} In the cultivation of oysters, clams and other mollusks, as well as in studies on their rates of growth under natural conditions, it has been frequently observed that certain years are more favorable than others for rapid increment in size. But there has been no satisfactory analysis of the environmental conditions responsible for the observed differences in growth rates. For this reason an experimental study, extending over four years, has been made of the growth of the California sea-mussel at the pier of the Scripps Institution of Oceanography. At this station daily records are made of the temperature of the water and of the numbers of dinoflagellates, diatoms and bacteria present and potentially contributory to the mussels' ultimate food supply. For the temperature records the writers are in- debted to Capt. S. W. Chambers; for the data on phytoplankton to Prof. W. E. Allen, and for those on bacteria to Prof. C. E. ZoBell. They also appreciate the technical assistance of Miss Harriet Dunn and Mr. Carl Johnson. In two previous papers (Coe and Fox, 1942; Fox and Coe, 1943) the writers have presented evidence relative to the normal rates of growth in this species at different seasons and at different ages and sizes, the different rates in the two sexes, the nature of the food materials and the influence of environmental conditions on the rates of growth. It was concluded from these observations that there is a gen- erally positive correlation of the growth rates both with temperature changes and with the abundance of dinoflagellates present in the water. No similar correlation was found relative to the numbers of diatoms or bacteria. It was emphasized how- ever that the correspondence between the size of the dinoflagellate populations and the growth rates of the mussels was not to be interpreted as the direct effect of dinoflagellates as potential food material, since a large proportion of the living dino- flagellates ingested usually pass apparently unchanged through the mussel's diges- tive tract. Furthermore the total supply of living phytoplankton which the mussel could possibly obtain is estimated to be so small in amount that even if all the con- stituents could be fully utilized they would furnish less than one-fifth of the mus- sels' nutritive requirements. The principal portion of the food was shown to con- sist of finely divided organic detritus, derived from the disintegration of many kinds of marine animals and plants, including both unicellular and multicellular forms (Fox and Coe, 1943). Continuation of those observations during three additional years has shown that there are wide variations in the mussels' growth rates, not only from month to 1 Contributions from the Scripps Institution of Oceanography of the University of Cali- fornia, New Series No. 233. 59 60 \\ . K. (JOE AXU D. L. FOX month but also from year to year. It is the object of this paper to record these variations and particularly to present such evidence as has now been obtained as to their causes. For this purpose the changes in the environmental conditions from month to month have been analyzed as fully as possible. Since there are no rivers in the vicinity and the annual rainfall is small, there is but little variation in salinity. The principal effective variables are the temperature and the food supply. The experimental mussels were kept in wire-screened boxes immersed in the >ea below the lowr-tide level. Since the increment in size was found to vary with the age of the individual under identical environmental conditions (Coe and Fox, 1()42), it was necessary to have the same ages represented at all times. This re- quired the addition of young individuals from month to month and the removal of the oldest. The experiment was continued from January 1940, to January 1944, with the exception of the first five months of 1942. At nearly all times the experimental boxes contained from 100 to 400 or more individuals, each age group being in a separate compartment. The average age remained nearly constant and all were sexually immature. When the individuals of a group were separately numbered, it was found that some grew rapidly for a period and were then overtaken by others ; some became leaders for several months, while others remained dwarfs. For statistical purposes it was therefore desirable to follow the growth of 20 or more individuals of each age group. The mean monthly increments in size for all groups are shown in Figures 1 to 4. In any consideration of the environmental conditions, it must be kept in mind that these conditions are constantly changing, due to the water currents that con- tinually sweep past the pier at rates averaging from four to five miles per day. Consequently these conditions may vary considerably from day to day and even from hour to hour. The monthly means, however, will give a reasonably close approximation to the prevailing environments. A comparison of the graphs in Figures 1 to 4 shows that the growth rates of the mussels have varied considerably, not only from year to year but also for the corresponding months of the years. The mean monthly increment in length for all groups was 3.43 mm. in the year 1940. 3.96 mm. in 1941, 5.43 mm. during the last seven months of 1942 and 5.11 mm. in 1943. It is evident from these figures that 1942 and 1943 were more favorable for rapid growth than either of the other two years and that 1940 was the least favorable. The lowest rate for any month of the four years was in August 1940. During that month the mean increment in length was only 1.6 mm., which was less than one-third as great as in the corre- sponding month of each of the three other years. The maximum rate occurred during April in 1940, during June in 1941, during July in 1942 and during May and July in 1943 (Figs. 1-4). An examination of the environmental conditions, particularly as concerns tem- perature, storms and abundance of the phytoplankton during these years, will give .Mime indication of the influence of each on the observed growth rates of the mussels. First Year, 19-10 The monthly growth rates of the mussels during this year were exceptional in that they showed fewer positive correlations with the temperature and with the BIOLOGY OF THE CALIFORNIA SEA-MfSSKL 61 abundance of dinoflagellates than in any of the other years. Following a decrease in the rate during February there \vas a rapid increase to a maximum in May, fol- 40000 30000 20000 200000 IOOOOO 50000 20° 18" MYTILUS 1940 MONTHLY MEAN INCREASE IN LENGTH DINOFLAGELLATES MONTHLY MEAN—, CELLS PER LITER DIATOMS MONTHLY MEAN CELLS PER LITER BACTERIA MONTHLY WEAN i " CELLS PER LITEFJ MEAN SURFACE TEMPERATURE 1940 000000 750000 500000 250000 J FM AM J Jv, A SOND FIGURE 1. Graphs showing the average monthly growth rate of 453 mussels, divided into 11 groups according to age, and the abundance of dinoflagellates, diatoms and bacteria, as well as the average monthly temperature of the water during the year 1940. The depression of the growth rate in February was mainly due to a reduction of the feeding period to 22 days because of accidents caused by storms; the dotted line indicates the estimated increment if the accidents had not occurred. It was necessary also to estimate the growth in December because of an accident due to storm. The numbers of dinoflagellates shown in the graph differ in several cases from those indicated in Figure 4 of our previous paper (1942) because of erroneous data supplied to us at that time. With some exceptions the growth rates were highest during those months having large dinoflagellate populations and in which the temperature exceeded 16° C. 62 \V. R. COE AND D. L. FOX lowed by a continuous decline to the lowest rate for any month of the four years in August (Fig. 1). The only explanation that can now be given for this excep- tionally low rate in .August is that for some unknown reason the organic detritus which furnishes the greater part of the mussel's nutrition was not present in suffi- cient quantity. The sharp drop in February was in part due to storms which necessitated removing the mussels from the sea and keeping them in the aquarium for six days. Since no increase in size occurs in the aquarium except when addi- tional food is supplied (Coe and Fox, 1942), there was a possible feeding period of only 22 days during that month. Computed on the basis of the growth during that period, the estimated increase per day in February would be but slightly less than during the preceding month, as indicated by the dotted line in the graph (Fig. 1). A sharp rise in the growth rate in September and an additional in- crease in October was followed by the usual decline during the last two months of the year. As a general rule, but with some conspicuous exceptions, the most rapid incre- ment in size occurred during those months in which a large population of dino- flagellates was present and in which the temperature exceeded 16° C. Neither the diatoms nor the bacteria showed definite correlations with the growth rates of the mussels (Fig. 1). Second Year, 1941 During the second year the growth rate was somewhat higher than in the pre- ceding year, although the average number of dinoflagellates was smaller and the diatoms were less than half as abundant as in 1940. With the exception of Feb- ruary there was a continuous rise in the growth rate to a maximum in June, with a steady decrease thereafter (Fig. 2). The dinoflagellate population correspond- ingly reached a maximum in July, followed by a continuous decline to a minimum in December. Neither the diatoms nor the bacteria showed similar trends. The rate indicated for December is lower than it would have been except for a severe storm which allowed a feeding period of only 27 days. Third Year, 1942 The experiment was interrupted for the first five months of 1942, but the last seven months of the year showed a greater increment in growth than in the corre- sponding months of any of the other years. The maximum rate occurred in July, followed by a continuous decrease during the rest of the year, with the exception of a slight rise in November, followed by a small decrease in December (Fig. 3). The water during those months contained an average of more than five times as many dinoflagellates as in the last seven months of the preceding year and the aver- age monthly increment in the lengths of the mussels was 5.4 mm. as compared with 4.3 mm. in the corresponding period of 1941. The average number of diatoms was smaller than in any of the other three years. Fourth Year, 1<>43 During the year 1943 the average monthly increment in size was considerably greater than in 1940 or 1941 but somewhat less than in 1942 (Fig. 4). By com- BIOLOGY OF THE CALIFORNIA SEA-MUSSEL 63 paring the rate for December 1942 (Fig. 3), with that of January 1943, it will be seen that a sharp drop occurred during the latter month. The cause of this de- crease in growth rate may have been due to a severe storm which necessitated transferring the experimental box from the sea to the aquarium, where it remained for four days. 40000 50000 20000 10000 0 200000 ISOOOO 100000 50000 0 20' IS" 16 14' MYTILUS MONTHLY MEAN INCREASE IN LENGTH 1941 DINOFLAGELLATES MONTHLY MEAN CELLS PER LITER J DIATOMS MONTHLY MEAN CELLS PER LITER BACTERIA. r MONTHLY MEAN CELLS PER LITER MEAN SURFACE TEMPERATURE 1941 I 1000000 750000 500000 250000 JFMAMJJuASOND FIGURE 2. dinoflagellates, the year 1941. the dotted line The most dinoflagellates, Graphs showing average monthly growth rate of mussels and abundance of diatoms and bacteria, as well as average monthly temperature of the water during The growth indicated for December represents a feeding period of only 27 days ; indicates the computed increase for a month of 31 days. rapid increment in size occurred in those months having large populations of accompanied, presumably, by an abundance of organic detritus. 64 \\ . K. C< )!•: AX'D D. I.. FOX From this depression in January, the growth rate increased continuously until May, when the average increment was 6.4 mm. (Fig. 4). An unaccountable drop in the °ro\vth rate during June was followed by an average increment of 6.6 mm. s IOOOOO 4 00000 5 80000 2 70OOO 60000 o 50000 40000 20000 10000 200 OOO 150000 IOOOOO 50000 MYTILUS MONTHLY MEAN INCREASE IN LENGTH o 8 1942 DINOFLAGELLATES MONTHLY MEAN CELLS PER LITER o o o o o o o o DIATOMS MONTHLY MEAN CELLS PER LITER BACTERIA MONTHLY" MEAN CELLS PER LITER MEAN SURFACE TEMPERATURE 1942 4- 1000000 710000 500000 250000 M M u FK.TKK .1. (iraphs showing the mean monthly growth of mussels, the mean monthly tem- perature of tlu- water and the mean monthly abundance of dinorlagellates, diatoms and bacteria during W2. in July, which was the highest rate for the year. Following the usual decrease in August, the rate- continued high and steady during the two succeeding months; then, instead of the usual decline in November, there was a rise to an average of (\ Him., as compared with about 3 mm. in tin- corresponding month of 1940, 2 mm. BIOLOGY OF THE CALIFORNIA SEA-MUSSEL 65 in 1941 and 4.3 mm. in 1942. In December the rate was but little more than half as great as during the preceding month (Fig. 4). 100000 4 90000 3 800OO z 70000 I eoooo o soooo 40000 30000 20000 10000 * 0 200000 1 10000 100000 60000 20 18 16 14 MYTILUS MONTHLY MEAN INCREASE IN LENGTH 1943 DINOFLAGEL . MONTHLY MEAN Co o 8 o o o _ATES CELLS PER LITER DIATOMS - MONTHLY MEAN CELLS PER LITER MEAN SURFACE TEMPERATURE 1943 L 1 I i I I I 1 L MAM O N FKU-RK 4. Graphs showing the nu-an monthly growth of mussels, the mean monthly tem- perature of the water and the mean monthly abundance of dinoflagellates and diatoms during the year 1943. The average monthly increment in length was 5.1 mm. as compared with 3.43 mm. in 1940 and 3.96 mm. in 1941. This increased rate of growth was accom- panied by populations of dinoflagellates more than four times as great as in either 66 W. R. COE AND D. L. FOX 1940 or 1941. The diatoms were also four times as numerous as in 1942, but again it should be emphasized that neither the dinoflagellates nor the diatoms furnish more than a small proportion of the food supply of the mussel. ENVIRONMENTAL INFLUENCES The observations described on the foregoing pages show how great a variation was found in the growth rates of the mussels in different years and in different months of each year. The comparisons of these rates with the temperature and with the abundance of dinoflagellates. diatoms and bacteria are shown in Figures 1 to 4. There are obviously many other environmental conditions which are con- stantly exerting their influence on the growth of the mussels. Some of these may be of great importance but they are so sporadic in their action or so difficult to measure that no precise evaluation of their influence has as yet been possible. There is little variation in salinity throughout the year and there is often a cor- relation between the amount of oxygen and the relative abundance of phytoplankton. 1. Temperature As a general rule the rate of growth in mollusks increases with the temperature to a certain optimum and then rapidly decreases. Consequently the annual incre- ment in length is greater in southern than in more northern localities because of the longer season favorable for rapid growth. The observations of Weymouth, Mc- Millan and Rich (1931) on Siliqua, of Newcombe (1936) on Mya, of Chamberlain (1931) on Lampsilis, of Orton (1926-27) on Cardium and of'Coe (1938) on Ostrea support this conclusion. The size eventually reached by the individual how- ever is commonly much greater in the north because of the greater length of life. At the pier of the Scripps Institution of Oceanography, where the experiment was conducted, the variation in the mean monthly surface temperature of the water during the year seldom exceeds 8° C. Both the low point of about 14° in winter and the high of about 22° in summer are well within range of the normal activities of the mussel. The highest temperature recorded at any time during these four years was 22.9° C., in August 1943, and the lowest was 13.4° in January 1943. Consequently growth continues throughout the year in this locality, although the rate of increment in length is only about half as great in midwinter as it usually is in the early summer. This decreased rate in winter is presumably due both to a lower state of metabolic activity and a decreased supply of nutritive materials. In M. calijoriiianits, as in M. cdiills (Loosanoff, 1942), feeding continues at temperatures both lower and higher than the extremes mentioned in the preced- ing paragraph. Under experimental conditions the California mussel will secrete liyssus threads, ingest food and discharge feces at temperatures as high as 24 to 26° C. and to a less extent at 27 to 28°. The individuals subjected to the highest oi these temperatures however died within 5 to 7 days. The lowest temperature at which in-cstion and fecal discharge were found to occur was 7 to 8°. Since these musseK were subjected to the temperatures mentioned without any period of acclimatization from an aquarium temperature of 15°, it is considered probable that the figures given do not represent the extreme range of the mussels' potential metabolic activities BIOLOGY OF THE CALIFORNIA SEA-MUSSEL 67 Experiments previously made in this laboratory indicate that the maximum rate of filtration and maximum oxygen consumption take place at about 20° C, with distinctly lower rates below 15°. A decreased growth rate has occurred in Au- gust, the month of highest temperature, in each of the four years. But in 1940 the decrease began in June and reached its lowest rate in August, while in 1941 the highest growth rate occurred in June and in 1942 in July. In 1943 maxima oc- curred in May and July. The variability in the growth rates at corresponding temperatures in different years indicates that the food supply is more influential than small variations in temperature in determining the rates of growth. The prevalence of storms and high seas common in winter are doubtless detrimental to active growth. 2. Food The mussel is essentially a scavenger, utilizing as food not only small unicellular organisms and dissociated cells but also the participate disintegration products of any of the animals and plants which die in the vicinity or similar products which are brought from a distance by currents. Even the bacteria which cause the de- composition may themselves be utilized as an additional source of nourishment (Coe and Fox, 1942; Fox and Coe, 1943). The constituents of the ingested materials have been ascertained at frequent in- tervals by examination of the stomach and intestinal contents and of the feces. The substances most commonly present are finely divided organic detritus, dinoflagel- lates, diatoms, silicoflagellates and bacteria; also tintinnids, flagellates, ciliates and other protozoans, as well as algal cells and fragments, algal spores, spermatozoa and ova (including those of its own species), together with inorganic substances such as particles of sand and shells. At times the organic materials may be in- gested in amounts much greater than the mussels' capacity for assimilation. The excess, if not too great, may pass unchanged through the digestive system but in case of a very great surplus most of the material is rejected by the palps and is dis- charged from the mantle cavity as pseudofeces. No satisfactory evidence of selec- tion, except as to size, from among these small cells and particles has been obtained, although chemically injurious substances are rejected, together with the larger cells and other objects. Digestion in the mussel, with the exception of starch and glycogen, appears to be mainly or wholly intracellular. Many of the smallest objects and particles are phagocytized by the cells lining the digestive diverticula. Others are ingested by phagocytic cells which migrate into the lumens of the stomach and intestine and later return with their ingested materials through the epithelial lining of the di- gestive tract and thence to the connective tissues of the body, as Yonge (1926, 1931) has so fully described for the oyster. Most of the local dinoflagellates and many of the diatoms are far too large to be assimilated in this manner. Some of the starch and glycogen, on the contrary, undergoes extracellular digestion in the stomach through the action of enzymes in the style. No evidence of the digestion of cellulose, which forms the covering walls of most neritic dinoflagellates, nor of any cells with completely closed cellulose walls, has been obtained (Fox and Coe. 1943). Diatoms. These organisms, either living or dead, furnish a small portion of 68 \V. R. COE AND D. L. FOX the mussels' nutrition. Their disintegration products are also utilized. They are usually present in numbers ranging from 1000 to 200,000 per liter but the large and spiny ones are not ingested. Many of those that enter the digestive tract are seized and digested by the phagocytic cells mentioned in a preceding paragraph, while others pass through the tract without apparent change. The mean number of diatoms, as counted by the settling method, per liter of water for each month of the four years is shown in Figures 1 to 4 and the com- bined monthly averages for 1940. 1941 and 1943 in Figure 5. In none of these years has there been a direct correlation between the diatom populations and the mussels' growth rate, although positive correlations have been reported by New- combe (1935) for My a and by Nelson (1942) and others for oysters. The number of diatoms in the water about the mussel beds has varied greatly from year to year. The average in 1940 was 38,700 per liter, in 1941 16,600, in 1(>42 12,600 and in 1943 54,300 (Figs. 1-4). The average monthly increase in the lengths of the mussels for the same years was 3.43 mm., 3.96 mm., 5.43 mm. and 5.10 mm., respectively. It is obvious that there was in these four years no direct correlation in the two groups of data. In spite of a four-fold increase in the num- ber of diatoms in 1943 as compared with 1942 there was nevertheless a somewhat lower rate of growth in the mussels. This should not be surprising when it is realized that even if these organisms had been uniformly distributed in the water throughout the year, instead of occurring in dense swarms, the rate of filtration by the mussel is such that an adult animal could have secured no more than 200 mil- lion to 800 million per year. If all of these could have been fully utilized they would have furnished only a minute fraction of the material required for the up- building of the mussel's tissues and gametes. Such of these organisms as can. be ingested by the mussel are so minute that it would require some 600 million to supply one gram of organic matter, while the adult mussel is estimated to need about 40 grams annually (Fox and Coe, 1943). Bacteria. Bacteria are ingested in vast numbers (Figs. 1-3) but their total mass is so small that they have little quantitative influence on the mussels' nu- trition (Fox and Coe, 1943). Dino flagellates. It has been mentioned that in each of the four years there was generally, but with some conspicuous exceptions, a rather close correspondence be- tween the monthly and yearly growth rates of the mussels and the abundance or scarcity of dinoflagellates in the water. In 1940 the average daily number of these organisms per liter of water was 12,100, as compared with 9880 in 1941, 54,750 in 1{|42 and 49,500 in 1943. The corresponding average monthly growth rates of the mussels were 3.43 mm. in 1940, 3.96 mm. in 1941. 5.43 in 1942 and 5.10 in 1943. Krom thoe figures alone it may be concluded that the mussels grow most rapidly in those years in which the populations of dinoflagellates are the largest. More precise evidence as to this association however is furnished by an inspec- tion of the monthly data as shown in Figures 1-4. It has been emphasized in a foregoing paragraph however that a large proportion of the living dinoflagellates, which may be ingested in vast numbers, usually pass apparently unchanged through the intestinal tract and often constitute much of the fecal material. Their cellulose walls cannot be digested by the secretions in the stomach or intestine and there is no satisfactory evidence that they are phagocyti/.ed by the cells of the digestive diver- BIOLOGY OF THE CALIFORNIA SEA-MUSSEL 69 ticula in any considerable numbers. Following the death of these unicellular or- ganisms however, either before or after entering the mussel's digestive tract, they doubtless form participate disintegration products which are readily assimilated. It is well known that species of Gonyaulax and less frequently of some other genera may be ingested in such numbers and the contained toxic substance accumulated in such quantity as to cause sickness or even death when the mussels are eaten by man. It has been shown by Fox and Coe (1943) that the mussel filters the water at such a rate that the available supplies of dinoflagellates, even if they could be fully utilized, would furnish only a small fraction of the food which the mussels require for their growth and reproduction. Assuming a filtration rate of 2.5 liters per hour, or 22,000 liters per year, it would be necessary to have an average popula- tion of about 2000 of these cells per liter in order to supply one gram of organic matter in a year. This is only about two and one-half per cent of the amount which an adult mussel is estimated to require annually for the upbuilding of its tissues and gametes. During these four years the water has contained averages of 10,000 to 54,750 of these cells per liter but a large proportion of these were present in such dense swarms that relatively few of those that were drawn into the mantle cavities of the mussels could have been actually ingested ; the others were presumably dis- charged as pseudofeces. Furthermore, as has been stated, many of those that are ingested usually pass through the digestive tract without visible change, while the large and spiny forms are seldom ingested. Therefore any correlation between the abundance of dinoflagellates and the growth rates of the mussels must be merely indicative of other, associated sources of nutrition. The principal source is organic detritus. Detritus. The organic detritus ingested by the mussel consists of various frag- ments of cells or of entire cells of minute size, as well as suspended proteins, lipids and polysaccharides. It may be recalled that the mussel obtains its food by secret- ing over its gills a thin sheet of mucus to which the particles are adsorbed. The mucus sheet with its attached particles is then drawn into the mouth. There is no evidence that the mussel is capable of securing substances in true solution until these have first been changed to participate form through the agency of various unicellular organisms (Fox and Coe, 1943). Consequently it may be concluded that, with the exception of refractory humus materials, cellulose, chitin and other indigestible substances, the total organic con- stituents of all marine organisms, from the smallest to the largest, are potential sources of nutrient for the mussel. After the death and disintegration of the animal or plant, the residual organic matter, or detritus, may remain suspended in the water for an indefinite period before it chances to enter the digestive system of the mussel. The amount available is obviously subject to great variation locally and it is to this variation that many of the differences in growth rates are ascribed. Inspection of Figure 5, which indicates the combined average monthly growth rates for three years, will show that the rate increases from a low in January or February to a maximum in May, June and July. This period corresponds with the increasing reproduction of many of the invertebrates in the vicinity and elsewhere along the coast. The striped barnacle (Balanus tintinnabulum) , for example, -has minimum and maximum periods of reproduction coinciding almost precisely with the low and high growth rates of the mussels (Fig. 5). These reproductive peri- 70 W. R. COE AND D. L. FOX odicities are undoubtedly associated with similar variations in the detritus which the barnacles yield, since a large proportion of the free-swimming larvae die with- out finding a place of attachment and presumably less than one per cent of those which succeed in transforming to the adult stage survive to reach sexual maturity 6h MYTILUJL MEAN INCREASE IN LENGTH '. 4 3 . I 0 50000 DINOfLAGELATES 25000 0 20° 18° 16" 14" 1940, 1941 AND 1943 BARNACLES ------- MEAN NUMBER PER LITER 150000 DIATOMS MEAN NUMBER 100000 PER LITER 50000 MEAN SURFACE TEMPERATURE NO. 300 200 100 M M J1 O N K 5. Correlations between the mean monthly increment in length of the experimental of mussels for the years 1940, 1941 and 1943, the mean numbers of dinoflagellates and diatoms per liter and the mean temperature of the water. The year 1942 is not included because of lack of complete data for the first five months of that year. Except for the decrease both in growth rate of the mussels and numbers of dinoflagellates during August, when the tempera- tures were highest, the general correspondence of three of these groups of data is evident. But this correspondence does not hold for the diatoms. As indicative of the relative amount of organic detritus presumably available in each month, the estimated average numbers of barnacles which became attached to each square inch of surface ni submerged plates in previous years have been included. These numbers indicate a close correlation with the growth rates of the mussels. (Coe, 1932). The disintegrated bodies of those that perish doubtless supply more nutritive material to the mussel than can be obtained from the living phytoplankton. In some years the growth rate has been more or less distinctly bimodal, rising to a maximum in late spring or early summer, followed by an invariable decrease in BIOLOGY OF THE CALIFORNIA SEA-MUSSEL 71 August and in three of the four years with another rise preceding the decrease at the end of the year. In three of the four years the dinoflagellates showed some- what similar himodal periodicities, with distinct spring and autumn maxima, and this has heen shown to be the average condition of these organisms for the twenty preceding years (Allen, 1941). The diatom populations, on the contrary, reached a conspicuous maximum in August in 1940 and in 1943, although in many other years that has heen a month of extremely low production. The bimodal periodicities in the growth rates of the mussel are closely parallel with similar periodicities in the reproduction of many of the associated invertebrates, including the mussels themselves. The growth rates of the entire mussel popula- tion would obviously be complicated by the reproductive processes but these com- plications were avoided in this experiment by using sexually immature individuals. It is unfortunate that no precise measurements are available relative to the local variations in the amount of organic detritus, but the seasonal differences in the rates of growth of the mussels and other detritus feeders presumably afford a fairly reliable criterion. SUMMARY This study offers additional evidence as to the complexity of the environmental conditions found along the shores of the ocean and which affect so profoundly the lives of the organisms residing there. Variations in the growth rates will obviously depend upon the interaction of several of these conditions, not the least important of which are the temperature and the character and abundance of the food supply. Furthermore these environmental conditions are constantly changing, due in part to the continual motion of the water. At the locality where the foregoing ob- servations were made, there are not only the variable currents caused by wind and tidal changes, but there is also a drift along the coast at a rate averaging four to five miles per day. Consequently the water in which the mussels are living and the conditions associated therewith may differ not only from month to month but also from day to day and even from hour to hour. In one week there may be ten to fifty times as much phytoplankton in the water as in the following week. The yearly averages are more stable but these may vary by more than five fold. Monthly correlations, extending over four years, between the growth rates of the mussels and the prevalent environmental conditions offer conclusive evidence that the most rapid increase in size takes place at temperatures from 17 to 20° C, although growth continues less rapidly at 14° or lower. Feeding continues at a temperature as low as 7 to 8° and as high as 27 to 28°. The average number of diatoms per liter was 38,700 in 1940, 16,600 in 1941, 12,600 in 1942 and 54,300 in 1943. The average number of dinoflagellates for the same years was 12,100, 9880, 54,750 and 49,500, respectively. The average monthly increment in the lengths of the mussels was 3.43 mm. in 1940, 3.96 in 1941, 5.43 mm. in 1942 and 5.11 mm. in 1943, indicating a yearly variation of more than 50 per cent. It is obvious that the two years with the largest dinoflagellate populations have been conducive to the most rapid growth of the mussels, but an increase of four fold in abundance has been associated with an increased growth rate of only 42 per cent. The correlation noted is evidently not direct, since the living dinoflagellates 72 \V. K. COE AND D. L. FOX can supply only a small fraction of the mussels' nutritive requirements. Both mus- sels and dinorlagellates appear to thrive under the same environmental conditions. That the living diatoms and hacteria are of even less importance in the life of the mussel is indicated not only by the small amount of organic matter that they con- tain but also by the fact that the mussels grew most rapidly in the year with the smallest number of diatoms and least rapidly in the year when the number was three times as great. More direct correlations with the growth rates of the mussels are found in the reproductive periodicities of various invertebrate populations which contribute so largely to the available organic detritus and thereby to the mussels' nutrition. The principal food supply of this species of mussel consists of minute particles of organic detritus derived from the disintegration of the cells of all kinds of marine organisms, both animals and plants, supplemented by living and dead unicellular organisms of minute size as well as living and dead gametes. There is no evidence that organic matter in true solution can be utilized until after it has been changed into participate form by the action of unicellular organisms. LITERATURE CITED AIJ.EX, WINFRED EMORY, 1941. Twenty years statistical studies on marine plankton dinoflagel- lates of Southern California. Amcr. Midland Not., 26: 603-635. CHAMBERLAIN, THOMAS K., 1931. Annual growth of fresh-water mussels. Bull. U. S. Bur. I'lsh., 46: 713-739. COE, WESLEY R., 1932. Season of attachment and rate of growth of sedentary marine organ- isms at the pier of the Scripps Institution of Oceanography, La .folia, California. Bull. Scrips Just. Ocean., Univ. Calif., 3: 37-86. COE, WESLEY R., 1938. Primary sexual phases in the oviparous oyster (Ostrea virginica). Biol. Bull., 74 : 64-75. COE, WESLEY R., AND DENIS L. Fox, 1942. Biology of the California sea-mussel (Mytilus californianus). I. Influence of temperature, food supply, sex and age on the rate of growth. Jour. £.r/>. Zool, 90: 1-30. Fox, DENIS L., AND WESLEY R. COE, 1943. Biology of the California sea-mussel (Mytilus californianus). II. Nutrition, metabolism, growth and calcium deposition. Jour. £.r/'. Zool., 93 : 205-249. LOOSANOFF, VICTOR C., 1942. Shell movements of the edible mussel, Mytilus ednlis (L.) in relation to temperature. Ecology, 23 : 231-234. NELSON, T. C., 1942. On the role of diatoms in the fattening of oysters. Bull. Oyster lust. N. America, 8: 5-8. NEWCOMBE, CURTIS L., 1935. Growth of Mya arenaria in the Bay of Fundy region. Canadian Jour. Res., 13 : 97-137. NEWCOMBE, CURTIS L., 1936. A comparative study of the abundance and the rate of growth of Mya arenaria L. in the Gulf of St. Lawrence and Bay of Fundy regions. Ecolof/v, 17 : 418-428. ORTOX, J. H., 1927. On the rate of growth of Cardium edule. Jour. Mar. Biol. .Issoc., 14: 239-279. \VEYMOUTH, F. W., H. C. McMiLLAN ANO WILLIS H. RICH, 1931. Latitude and relative growth in the razor clam, Siliqua patula. Jour. E.vp. Biol.. 28 : 228-249. YoNGE, C. M., 1(J_'(>. Structure and physiology of the organs of feeding and digestion in Ostrea edulis. Jour. Mar. Biol. Assn., 14: 295-386. VoNGE, ( . M., 1931. Digestive processes in marine invertebrates and fishes. Jour. Cons. Jut. Explor. dc l» Mer., 6: 175-212. PHOTOCHEMICAL SPECTRAL ANALYSIS OF NEURAL TUBE FORMATION x JAMES O. DAVIS (Prom flic Dcfuirtnii'iit of Zooloiiy, University of Missouri, Colniii/>iti, Missouri) INTRODUCTION In recent years the analysis of morphogenesis has been concerned with the study of the chemical nature of substances which play a decisive role in developmental phenomena. In the field of neural induction the usual procedure used is that of extraction of the active material with solvents specific for a particular group of compounds ; the degree of substitution is tested by implantation of the extracted material. This approach has failed to yield conclusive results because large quan- tities of tissue known to contain the normal inductor are not available. Investiga- tors have been unable to determine whether induction resulting from implanted substances is the result of the direct action of the substance on the tissue or of a substance released in the reacting tissue. This technique is also subject to the criti- cism that two chemical substances may not necessarily be identical because they produce the same histological or morphological changes. It is a well-known fact that histological changes produced in the vagina and uterus by a number of arti- ficial estrogens are identical with those produced by the natural estrogens (Mo Kenzie, 1941). The investigation of active substances need not be restricted to attempts to iso- late them, although isolation and synthesis is the ultimate goal. If the action of a developmental substance is inhibited by a specific agent, a preliminary identification will have been made and it will be certain that the substance inactivated is opera- tive in the organism. The use of chemical poisons has demonstrated the impor- tance of this technique in the field of cellular oxidation. The technique of the photochemical inactivation of substances involved in the developmental processes has enabled the investigator to study the chemical nature of this material during its action in normal development. The classical experiments of Warburg (1927) in the identification of the re- spiratory enzyme by absorption spectrophotometry show the importance of the tech- nique of photochemical inactivation. This method involves the irradiation of a biological system with monochromatic radiation and the consequent inactivation of a chemical substance in the system. Absorption is measured indirectly in terms of a physiological or morphological change produced in the biological system. War- burg measured absorption by determining the change in oxygen consumption of yeast cells in the presence of carbon monoxide following irradiation with mono- chromatic light. In the present investigation absorption is measured in terms of the amount of energy required to inhibit the folding process in neural tube forma- 1 This paper is part of a dissertation presented to the faculty of the Graduate School of the University of Missouri in fulfillment of the requirement for the degree of Doctor of Philosophy. 73 74 JAMES O. DAVIS tion. By this method an indirect photochemical absorption spectrum of the ma- terial involved in a biological activity can be obtained. By comparing this spectrum with the absorption spectra of chemical compounds, information can be ascertained concerning the chemical nature of the material. A preliminary exploration showed that ultraviolet radiation would not inhibit the transformation of gastrula ectoderm into neural plate unless extremely large doses were used and the cells severely altered or killed. However, the folding proc- ess of neural tube formation was inhibited by very weak doses of ultraviolet light with little effect on the embryo in other ways. The neural plate merely continued to develop as a plate. The effect was uniform enough to compare quantitatively the effects of different wave-lengths. Consequently, an attempt was made to identify by its. absorption spectrum a substance which is apparently of decisive importance in the process of neural tube formation. Since most of the work concerned with neural tube formation has been done in Amphibia, it might be expected that this material would offer more advantages than any other. This is not the case. Amphibian embryos possess yolk granules and pigment which absorb and scatter incident radiation. Consequently, the photo- chemical efficiency curve obtained for inhibition of the folding process in Amphibia would not give a true measure of absorption. For this reason and because an abundant source of avian material was available, it was decided to use chick embryos in this investigation. The author wishes to thank Dr. Daniel Mazia under whose direction the study was made. Grateful acknowledgment is made to Dr. F. M. Uber who permitted the author to make the thermopile measurements in his laboratory and to Dr. L. J. Stadler for use of the monochromator. MATERIALS AND METHODS The material for this study consisted of the eggs of two breeds of the domestic fowl, the White Leghorn and the New Hampshire Red. In order to secure uni- form results, all the eggs were obtained from two pens of hens, one of New Hamp- shire Reds and the other of White Leghorns. The University of Missouri poultry farm was the source of this material. Each egg was incubated and the position of the blastoderm determined by can- dling. The egg was then placed in a Syracuse watch glass which contained model- ing clay to hold the egg in place. Only sterile equipment was used in these experiments. The surface of the egg was sterilized with a piece of cotton which previously had been soaked in 70 per cent alcohol. An opening of 7 to 9 sq. mm. was cut in the egg shell by means of a small saw. This revealed the blastoderm through the shell membrane. After removal of the shell membrane with forceps, a sterile .9 per cent salt solution was used to float the embryo to the level of the surface of the egg shell. The age in terms of somites and general condition of each embryo was determined with a dissecting binocular ; all embryos in which the neural tube was closed in any region and all abnormal embryos were discarded. The egg shell was marked with a pencil to indicate the position of the embryonic axis so that it could be placed parallel to the slit on the monochromator. The egg, which was tightly fixed in the watch glass by modeling clay, was placed upon a stand which had been attached to a rack and pinion. The egg was elevated until its sur- ANALYSIS OF NEURAL TUBE FORMATION 75 face was beneath a quartz prism which was situated at the slit on the monochromator. This "mechanical jack" enabled the investigator to place each embryo the same dis- tance from the source of light. The experimental embryos were irradiated for varying lengths of time with monochromatic ultraviolet light. The control embryos were treated in exactly the same manner except that a glass microscopic slide wa> placed in the path of the light beam so that no ultraviolet light struck the embryo. Following irradiation, the opening in the egg shell was covered with a piece of a glass cover slip and sealed with a mixture of beeswax and paraffin. The eggs were then incubated for 30 hours or longer; the glass window was always placed down in order to prevent the blastoderm from adhering to the cover slip. After incubation, the blastoderm was removed from the yolk by cutting around the periphery of it with iridectomy scissors and lifting it off with a metal spatula. The blastoderm was washed in saline and fixed with picro-sulfuric acid. Observa- tions were made upon embryos in alcohol, from whole mounts, and from sectioned material. The whole mounts were stained with borax carmine and the sectioned material was stained with Delafield's hematoxylin or borax carmine. In the majority of the experiments in this investigation, monochromatic radia- tion was obtained by means of a large crystal monochromator, described in detail by Uber and Jacobsohn (1938). The monochromator was operated in a hori- zontal position. Since it was necessary to obtain a vertical beam of light in order to irradiate the embryo, a small quartz prism was placed at the slit on the mono- chromator. The source of light for the monochromator was a vertical mercury arc which operated at 4 amperes on a 110 volt direct current. In order to determine the incident dose in ergs mm.- on the embryos, the mono- chromatic source was calibrated with a surface-type vacuum thermopile. The thermopile had been calibrated previously with a standard carbon-filament lamp (C-241) obtained from the United States Bureau of Standards. The dosage in ergs/mm.- emitted by the monochromatic source was determined by comparing the deflection which it produced with that produced by the standard lamp. In the other experiments a mercury discharge tube served as a source of radia- tion. It was of the Hanovia Sc-2537 type operating at 120 milliamperes and 5000 volts. The transformer was a Jefferson luminous type. Since spectral studies show that such discharge tubes frequently have an additional line around 1800 A (Landen, 1940), a water filter was placed in the path of the beam in order to ab- sorb the radiation of the shorter wave length. A slit of approximately the same size as that of the monochromator used in this investigation was made on the bot- tom of the filter in order to approximate the experimental set-up with the mono- chromator. OBSERVATIONS Histological studies <>j Irradiated and control cinhr\os • Chick embryos ranging in age from the primitive streak to the 8-s< unite stage were irradiated with monochromatic ultraviolet light of wave lengths 2483, 2537, 2576, 2650, 2699, 2804, 2894, 2967. and 3130 A and subsequently incubated for a period of approximately 30 hours. The smallest doses produce no detectable changes; the first visible effects to ap- pear as the dose is increased are on the formation of the neural tube. The neural 76 JAMKS O. DAVIS folds fail in close- and instead form flat or halt-folded neural plates. In some em- bryos this occurs in the anterior part of the hody; however, in other cases, it is present only in the middle portion. Large doses result in destruction of cells and death of the embryo. In the consideration of the effect ol monochromatic radiation on a develop- mental process one of the first questions to arise is this: Is it possible to set up a quantitative standard of measurement for comparing the effectiveness of different wave lengths? Such a standard would be a morphological endpoint. The pro- cedure would be to compare doses required to attain such an endpoint. If the data arc significant, the results should be the same regardless of the particular endpoint chosen. In the present study two morphological endpoints are used: (1) failure of the neural tube to close for a distance of one-third its length in 50 per cent of the cases and ( 2 i failure of the neural tube to close for a distance of one-half its length in 50 per cent of the cases. This investigation is concerned in particular with the embryos irradiated with the amounts of energy necessary to produce these two morphological endpoints. The description is made from a study of embryos in 70 per cent alcohol, sectioned embryos, and whole mounts. The primarv effect of radiation is on the neural plate. Embryos irradiated with wave lengths 2483. 2537. 2570. 2050. 2. 2804, and 28' >4 A are very uniform in appearance. A broad flat plate is present in the anterior one-third to five-sixths of the embryo; the neural tube is nearly always closed in the posterior end. The an- terior end of the neural plate bends around the anterior tip of the tree head and extends to the ventral surface. The optic cups and infundibulum develop from the portion of the neural plate on the ventral surface of the free head. A lens forms in most cases. A typical case with a broad flat plate, optic cups, lenses, and infundi- bulum is shown in a section through the anterior end of an embryo irradiated with wave length 2804 A ( Plate 1, Figure 1). A section at a more posterior level is shown in Plate I, Figure 2; this embryo was irradiated with wave length 2537 A. ( )bservations show that the auditory pits are normal in appearance in every case. A group of cells which is probably the neural crest often lies adjacent to the lateral edges of the neural plate. In the region of the rhoinbencephalon the motor roots of the spinal nerves are present. In the posterior end ot the embryo a double neural tube occurs occasionally. The broad flat plate which is present in embryos of this group is nearly uni- form in thickness except in the region of the midline. In this region the plate is thinner ( Plate I. Figures 1 and 2). The volume of the broad plate is much larger than that of the neural plate of a 0-somite embryo. Mitoses are abundant on the upper surface of the plate and on the inside portion of the closed region of the neural tube. In a few emhrvos a small group ot cells is present on the surlace at the lateral edges of the neural plate. They are filled with granules and irregular In these cases the lateral ectoderm is similar in appearance to these cells; otherwise, the lateral ectoderm appears normal. In some cases it continues to grow and expand so that a projecting group ot cells lorms on the dorsal surlace ol the neural plate. Another group of embryos is characterized by the presence ol a flat or lolded neural plate in the middle portion of the bodv. A complete neural tube forms in the anterior part of the embrvo, but it is abnormal in shape and smaller than a ANALYSIS OF NEURAL TUBE FORMATION 77 PLATE I FIGURE 1. Anterior end of an embryo irradiated with approximately 71 ergs/mm.'-' at wave length 28(14 A and subsequently incubated for 3d hours, showing the broad flat neural plate, optic cups, lens, infundibulum, gut, dorsal aortae, small notochord, and disorganized mesenchyme beneath the neural plate. < 107. FIGURE 2. At the level of the heart of an embryo irradiated with approximately 95 ergs/mm.- at wave length 2537 A and subsequently incubated for 30 hours, showing the flat neural plate which possesses a well-defined floor plate, small notochord, dorsal and ventral aortae, heart, and gut. 107. JAMES O. DAVIS normal tube. The typical thin roof plate of the tnyelencephalon fails to develop. The auditory pits and optic cups are normal. The cellular appearance of the notochord in the irradiated embryos is the same as that of a normal embryo. Measurements of the cross-section area of the noto- chord show that it varies greatly at different anterior-posterior levels. The noto- churd is invariably separated from the neural plate; in the region of the neural tube the notochord lies in contact with the floor plate. In Plate I, Figures 1 and 2 illus- trate the small size and relative position of the notochord. As far as can be de- tected, the somites are normal. In many cases, the mesenchyme beneath the neural plate is abnormally vesiculated in places and considerably disorganized ( Plate I, Figure 1). Stained sections show that the mesenchyme beneath the neural plate ha- been injured; this is suggested by the dark appearance of its cells. The vascu- lar system is well developed ; however, the size of the vascular bed is smaller than that of control embryos. The embryos which were irradiated with wave length 2(H>7 A are different in appearance from those described previously. The explanation for this is that ex- tremely large doses had to be used in order to produce a detectable effect. In most cases the neural tube forms only in the most anterior part of the embryo. It is very small and abnormally shaped, being extremely flattened dorsoventrally. Occasionally the tube fails to develop and a neural plate is present in the anterior region. In the posterior three-fourths to four-fifths of the body the neural plate is either a disorganized mass of cells or completely absent. The superficial ectoderm appears normal and forms a continuous layer of flat epithelial cells dorsal to the neural plate. In the embryos irradiated with wave length 2967 A, the cellular structure of the notochord is normal in appearance. However, the notochord is not in contact with either the neural tube or plate in most regions. The somites are either highly dis- organized or absent. The vascular system is poorly developed. Observations of 111 control embryos which were made from embryos in 70 per cent alcohol, whole mounts, and sections show that they are normal in 100 per cent of the cases. It will be remembered that all embryos were examined immedi- ately before irradiation and the abnormal ones discarded. This explains the fact that all control embryos are normal. From this description, it is evident that all wave lengths except 2(>u7 A pro- duce a uniform effect on neural tube formation; consequently, quantitative studies of the relative efficiency of different wave lengths in preventing closure ol the neural tube can be made. The effect <>j radiation on mitosis and volume <>) the central nei'i'oiis system The purpose of this studv is to determine if radiation has a detectable influence- on mitosis and volume changes with the low doses used. Alitotic count.- and meas- urement.- of the area of cross sections of the central nervous system were made on the embrvo> which were described histologically in the previous section. An analy.-i- of the effect of radiation on mitosis was made by counting the number of mito.se.- in three -cctions of the neural plate or tube at each of three different levels. Volume was measured indirectly bv determining the cross-section area of one ol ANALYSIS OF NEURAL TUBE FORMATION 79 TABLE I Mitotic counts and cross-section measurements of irradiated and control embryos Group Number of mitoses Cross-section Mitoses per of per section area cross-section embryos per embryo per embryo area unit Level a Embryos irradiated at wave lengths from 2483 to 2804 A (10)* 45.90±14.7 1189.2±278 .038 Embryos irradiated at wave lengths 2894 and 2967 A (4) 20.75±9.0 719.5±95 .029 Control embryos (5) 44.80±15.0 1504.6±409 .030 Level b Embryos irradiated at wave lengths from 2483 to 2804 A (10) 7.80±2.6 251.3±34 .031 Embryos irradiated at wave lengths 2894 and 2967 A (4) 1.75±2.1 93.75±47 .018 Control embryos (5) 12.60±4.7 379±115 .033 Level c Embryos irradiated at wave lengths from 2483 to 2804 A (10) 6.90±2.1 191.5±51 .036 Embryos irradiated at wave lengths 2894 and 2967 A (4) 4.75±5.6 99.5±85 .048 Control embryos (5) 10.00±2.0 284.60±46 .035 Level a consists of three sections adjacent to the anterior end of the notochord ; level b is represented by the middle section of the central nervous system and this level is determined by counting the total number of sections of the central nervous system; and level c is ten sections posterior to the most posterior section in which the neural tube has failed to close in the ir- radiated embryos and a comparable section in control embryos. In the irradiated embryos in which the counts were made, level c is always in the posterior quarter of the nervous system. In determining the number of mitoses per section per embryo, counts were made on three sec- tions at each level of each embryo. The number of mitoses per section at each level was obtained by averaging these three figures ; the number of mitoses per section per embryo was determined by averaging the number of mitoses per section for the group of embryos. Volume was measured by determining the cross-section area of one of the sections at each level in which the mitoses were counted; cross-section area per embryo was obtained by averaging the cross- section area for the group of embryos. The mean deviation has been calculated for the average values of the number of mitoses and cross-section area. * The figures in parenthesis represent the total number of embryos studied in each group. 80 JAMES (>. DAVIS tin- sections at each level in which the mitoses were counted. This was accom- plished by tracing the outline of the neural plate or tube on millimeter paper by means of a camera lucida and then counting the number of square millimeters TABLE II Tin- effect af an increased incubation period on closure oj the neural lithe Irradiated with 143.6 ergs/mm. • Irradiated with 170.0 ergs/mm.2 30-hour 54-hour 30-hour 54-hour incubation incubation incubation incubation Egg number Age in Age in Age in Age in somites at time of irra- Distance open somites at time of irra- Distance open somites at time of irra- Distance open somites at time of irra- Distance open diation diation diation diation 1 6 ys 6 ':, 6 !., 6 0 2 6 '. 4 '3 -> ,1 ', 5 i : .> 3 6 ••', 6 0 2 '._, 6 0 4 6 0 5 0 6 1 _' o 6 ', 5 6 open 6 0 6 ^ 6 0 6 6 ':* 1 0 0 0 6 0 7 6 H 7 0 7 H T J Vt 8 5 0 4 0 7 3,4 4 H 9 6 open 4 0 6 ":; 7 H 10 3 open 6 0 6 ':; 6 open 11 2 :i, 6 0 8 ', 8 ]4 12 3 ':< 6 0 7 0 3 0 13 0* 0 6 0 3 :i.. 5 y2 11 3 0 4 ':; 2 '., 7 o 15 6 ', 6 ':; 2 :!, 6 >, 16 0 0 6 0 3 '., 7 0 17 6 ';i 5 ', 7 ':; 6 0 18 6 0 7 0 19 6 0 7 YB 20 6 y3 21 6 0 22 8 0 23 4 0 24 6 X 25 7 '6 ' , open ':t waj or more 47 21 70 26 Hmbryos designated as having no somites \\cre in the primitive streak s within the outline. Xo attempt was made to transfer the values obtained into abso- lute ones, since this investigation is concerned only with relative data. The results are shown in Table I. From the data presented, it is concluded that the results with wave lengths .MX.i to _'S()4 A are not decisive enouh to establish an influence of radiation on mitosis and volume. However, wave lenths _'X(>4 and .\ arc very effective in decrcasin<'- mitosis and volume. g ANALYSIS OF NEURAL TUBE FORMATION 81 The effect oj an increased incubation period on closure of the neural tube The purpose of this experiment is to determine whether or not the neural folds which have failed to close in irradiated embryos will close if the incubation period is increased. Embryos ranging in age from the primitive streak to the 8-somite stage were irradiated with the mercury discharge tube with doses of 143.6 and 170.0 ergs/mm.2. The embryos were then incubated for periods of 30 and 54 hours. After removal and fixation of the embryos, observations were made with a dissect- ing binocular. The results are shown in Table II. Columns 3, 5, 7, and 9 which are designated as "Distance open" refer to the distance for which the neural tube has failed to close. The per cent of embryos in which the neural tube is open for a distance of one-third its length or more decreases with an increase in the period of incubation. For the 143.6 ergs/mm.2 group the per cent decreases from 47 to 21 ; in the 170.0 ergs/mm.2 group the per cent decreases from 70 to 26. This ex- periment shows that an increased incubation period results in partial closure of the region of the central nervous system which had failed to close after 30 hours in- cubation. Method of calculation of the incident energy on the embryos In the histological study of the irradiated embryos described in the first section of the observations, all wave lengths except 2967 A are found to produce the same TABLE III Absolute intensity of monochromatic radiation Wave length Cm. deflection Ergs/mm.2/sec. producing 1 cm. deflection Ergs/mm.2/sec. 2483 1.5 .726 1.089 2537 5.2 .726 3.775 2576 .6 .726 .436 2650 3.8 .726 2.759 2699 .9 .726 .653 2804 2.1 .726 1.525 2894 1.1 .726 1.799 2967 2.9 .726 2.105 '130 12.5 .726 9.075 qualitative effects. Since the effect of radiation is to inhibit the folding process, a quantitative comparison of the relative photochemical efficiency of different wave lengths can be made. The amount of incident energy in ergs/mm.2 required to inhibit folding was determined for the wave lengths used in this investigation. To facilitate a comparison of the relative efficiency of these wave lengths, two morpho- logical endpoints, namely, the inhibition of closure of the neural tube for a distance of one-third and for a distance of one-half its length in 50 per cent of the embryos were chosen. Attention is called to the fact that these morphological endpoints have dimensions of energy and for tliis reason are an indirect measure of absorp- tion. In order to determine the incident energy on the eggs, the intensity of the 82 JAMES O. DAVIS radiation at the egg surface for each wave length \vas measured by means of a surface-type vacuum thermopile. The thermopile measurements are shown in Table III, where "cm. deflection" refers to the number of centimeters the galvanom- eter needle was deflected when the intensity of each wave length was measured. By multiplying this value by .726 ergs/mm. 2/sec., which is the amount of energy producing a deflection of 1 cm., the number of ergs mm. -/sec. emitted by each line was obtained. The total numbers of embryos irradiated at wave lengths 2483. 2537, 2576, 2650, 2699, 2S04. 2S<>4. _*>67. and 3130 A are 75, 93, 45, 100. 68, 79. 42. 54. and 16. rc- TABLE IV Results with experimental doses stronger and weaker than the endpoint dose Incident Open Open less Per cent Open Open less Per cent Wave dose on Js way than 4 open 1 '2 way than >j open length egg in ergs/mm.2 or more way or closed 1 .1 way or more or more way or closed 1 •_> way or more 2483 196.06 16 25 39 13 28 32 261.36 10 9 53 6 13 32 2537 113.25 4 14 22 169.89 19 9 68 11 17 39 226.50 9 6 60 2576 130.80 10 12 45 5 17 23 156.96 13 5 72 11 7 61 2650 207.00 14 16 47 6 24 20 248.40 14 9 60 12 11 52 2699 195.90 4 6 40 215.49 2 2 50 235.08 11 10 52 10 11 48 274.26 4 1 80 2804 137.25 17 17 50 15 19 44 183.00 4 1 80 2894 239.70 6 10 38 5 11 31 287.64 12 6 67 8 10 44 2967 631.50 0 13 ? 0 13 ? 757.80 0 17 ? 0 17 ? 3130 5445.00 0 7 ? 0 7 ? 9256.50 0 1 ? 0 1 ? The figures in columns 3, 4, 6, and 7 represent the number of embryos irradiated ; the fig- ures in columns 5 and 8 refer to the per cent of embryos irradiated. "Open one-third way" or "open one-half way" means that the neural tube is open at least one-third or one-half its length; "open" refers to cases in which the neural tube is open less than one-third or one-half way; and "closed" indicates that the neural tube is closed. spectively. The results obtained with incident doses on the egg stronger or weakei than the doses required to inhibit closure of the neural tube for distances of one- third and one-half its length in 50 per cent of the cases are summarized in Table IV. In order to determine the amount of energy incident on the embryo, it is neces- sary to measure the amount of light transmitted by the vitelline membrane. The ultraviolet transmission of the vitelline membrane was measured by I'ber, Hayashi, and Klls (1941). These measurements were made with a Spekker photometer and a Ililger medium quart/ spectrograph. Three vitelline membranes were studied and the results are shown in numerical form in Table V for each wave ANALYSIS OF NEURAL TUBE FORMATION 83 TABLE V Ultraviolet transmission of three vitelline membranes (Hilgcr spectrograph data) Per cent transmission Wave length Membrane 1 Membrane 2 Membrane 3 Average 2483 6.35 4.65 3.05 4.68 2537 8.00 6.20 4.20 6.13 2576 9.90 7.00 5.20 7.37 2650 10.00 7.80 6.80 8.20 2699 10.00 8.00 7.45 8.48 2804 11.00 8.10 8.00 9.03 2894 16.30 13.65 11.00 13.65 2967 21.80 20.70 15.85 19.45 3130 26.10 30.10 30.10 28.76 90 80 70 — 2 z < K I- 60 z UJ O K 111 °- 50 40 30 2400 2500 2600 2700 2800 2900 3000 WAVE LENGTH IN ANGSTROM UNITS 3100 FIGURE 1. Ultraviolet transmission by vitelline membrane (T. Hayashi, unpublished). 84 JAMES 0. DAVIS length used in this study. In addition to tin- spectrograph data, Air. Tern Hayashi (unpublished) measured the ultraviolet transmission of the vitelline membrane with a photocell and a quart/ microscope. Mis data are shown in Figure 1. The amount of energy incident on the embryo was then calculated for these two sets of data on transmission by the vitelline membrane by multiplying the per cent trans- mission by the vitelline membrane by the incident energy on the egg. Relative photochemical efficiency curves The results of the calculations of the incident energy on the embryos are pre- sented in Tables VI and VII in which correction for absorption by the vitelline membrane is made with the Hilger spectrograph ; in Tables VITT and IX. the TABLE VI Incident energy on embryos for inhibition of closure of Yz the length of the neural tube (Hilger spectrograph data) Wave length Incident dose on egg in ergs/mm.2 Per cent transmission by vitelline membrane Per cent open H way Incident dose on embryo in ergs/mm.2 Reciprocal of incident dose on embryo Calcu- lated endpoint dose Reciprocal of calculated endpoint dose 2483 196.06 4.68 39 9.18 .109 11.58 .086 261.36 53 12.23 .082 2537 113.25 6.13 22 6.84 .146 9.01 .110 169.89 68 10.41 .096 2576 130.80 7.37 45 9.64 .104 10.00 .100 156.96 72 11.57 .086 2650 207.00 8.20 47 16.97 .059 17.75 .056 248.40 60 20.37 .049 2699 195.90 8.48 40 16.61 215.49 50 18.27 .055 18.27 .055 235.08 52 19.93 2804 137.25 9.03 50 12.30 .080 12.39 .080 2894 239.70 13.65 38 32.62 .031 35.37 .028 287.64 67 39.26 .025 2967 631.50 19.45 ? 122.83 .0081 p ? 757.80 ? 147.39 .0068 ? ? 3130 5445.00 28.76 ? 1565.98 .0006 ? ? 9256.50 ? 2662.17 .0004 ? ? photocell and quartz microscope transmission measurements are used for this cor- rection. At most wave lengths the doses recorded in the tables are just stronger or weaker than the endpoint doses, namely, the amounts of energy necessary to prevent closure of the neural tube for distances of one-third and one-half its length in 50 per cent of the embryos. At wave length 2804 A, a dose was used which prevented neural tube formation in exactly 50 per cent of the cases when one-third the length of the tube was used as an endpoint. When the doses used did not pro- duce the endpoint. the endpoint dose was determined by interpolation. The validity of this interpolation is based upon the assumption that the effect produced is directly proportional to dose. This assumption was used during the course of this investigation to predict the dose which would produce the endpoint. These predictions were fairly accurate, particularly in view of the small number of em- ANALYSIS OF NEURAL TUBE FORMATION 85 TABLE VII Incident energy on embryos for inhibition of closure of % the length of the neural tube (Hilger spectrograph data) Wave length Incident dose on egg in ergs/mm.2 Per cent transmission by vitelline membrane Per cent open Yi way Incident dose on embryo in ergs/mm.2 Reciprocal of incident dose on embryo Calcu- lated endpoint dose Reciprocal of calculated endpoint dose 2483 196.06 4.68 32 9.18 .109 ? ? 261.36 32 12.23 .082 2537 169.89 6.13 39 10.41 .096 12.12 .082 226.50 60 13.88 .072 2576 130.80 7.37 23 9.64 .104 11.01 .090 156.96 61 11.57 .086 2650 207.00 8.20 20 16.97 .059 20.16 .050 248.40 52 20.37 .049 2699 235.08 8.48 48 19.93 .050 20.20 .050 274.26 80 23.25 .043 2804 137.25 9.03 44 12.39 .081 13.08 .076 183.00 80 16.52 .061 2894 239.70 13.65 31 32.62 .031 42.32 .023 287.64 44 39.26 .025 2967 631.50 19.45 ? 122.83 .0081 ? ? 757.80 ? 147.39 .0068 3130 5445.00 28.76 ? 1565.98 .0006 ? ? 9256.50 ? 2662.17 .0004 TABLE VIII Incident energy on embryos for inhibition of closure of }•& the length of the neural tube (Quartz microscope and photocell data) Wave length Incident dose on egg in ergs/mm.2 Per cent transmission by vitelline membrane Per cent open ^3 way Incident energy on embryo in ergs/mm.2 Reciprocal of incident dose on embryo Calcu- lated endpoint dose Reciprocal of calculated endpoint dose 2483 196.06 60 39 117.63 .0085 148.42 .0067 261.36 53 156.82 .0064 2537 113.25 56 22 63.42 .0158 82.75 .0121 169.89 68 95.14 .0105 2576 130.80 58.5 45 76.52 .0131 79.35 .0126 156.96 72 91.82 .0109 2650 207.00 64 47 132.48 .0075 137.78 .0073 248.40 60 158.98 .0063 2699 195.90 59 40 115.58 .0087 215.49 50 127.14 .0080 124.14 .0080 235.08 52 138.69 .0073 2804 137.25 52 50 71.37 .0140 71.37 .0140 2894 239.70 56 38 134.20 .0074 145.32 .0069 287.64 67 161.08 .0062 2967 631.50 74 ? 467.31 .0021 ? ? 757.80 560.77 .0018 3130 5445.00 84 ? 4573.80 .00022 ? ? 9256.50 ? 7775.46 .00013 86 I \\11-S o. DAVIS bryos obtained at stronger and weaker doses wln'eb served as a basis for tbe pre- dictions. In most east's tliree or tour experiments were run and comparable re- sults were obtained in each experiment. Furthermore, if the experimental data which most closely cm-respond to the endpoint are used instead of the interpolated data, the maxima and minima ot the curves are not significantly changed. Although the embryos at the time of irradiation varied in age from the primi- tive streak to the eight somite stage, the results were not altered. Davis (1942) recorded the age of each embryo at the time of irradiation and the results show that the effect of radiation is independent of age for the small age range used in this TABLE IX Incident energy on embryos for inhibition of closure of *.<> the length of the neural lube (Quartz microscope and photocell data) Wave length Incident dose on eKK in ergs/mm.s Per cent traiismis-.ii m by vitelline membrane Per cent open 1 _. way Incident energy on embryo in ergs/mm.2 Reciprocal dl incident dose on embryo Calcu- lated endpoint dose Reciprocal of calculated endpoint dose 2483 196.06 60 32 117.63 .0085 ? ? 261.36 32 156.82 .0064 2537 169.89 56 39 95.14 .0105 111.74 .0089 226.50 60 126.81 .0079 2576 130.80 58.5 23 76.52 .0131 87.39 .0114 156.96 61 91.82 .0109 2650 207.00 64 20 132.48 .0075 157.32 .0064 248.40 52 158.98 .0063 2699 235.08 59 48 138.69 .0073 153.22 .0065 274.26 80 161.81 .0062 2804 137.25 52 44 71.37 .0140 75.34 .0133 183.00 80 95.16 .0105 2804 239.70 56 31 134.2 .0074 172.71 .0058 287.64 14 161.08 .0062 2967 631.50 74 p 467.31 .002 1 ? ? 757.80 560.77 .0018 ? ? 3 1 30 5445.00 SI ? 4573.80 .00023 ? ? 0256.50 ? 7775.46 .00013 group of experiments. Furthermore, it was possible to predict the experimental results that would be obtained with slightly larger or smaller doses regardless o! the age of the embryos used. Interpolations cannot be made at wave lengths 2%7 and 3130 A. A large per- centage of the tubes are open in the embryos which were irradiated with wave length 2()(>7 A; however, none are open as much as one-third the length of the neural tube. Embryos irradiated with 495.20 ergs, "mm.- or less show a slight in- jury, while the stronger doses used produce considerable injurv. In view ot this and since 757. SO ergs mm.- result in failure of the neural tube to close in 35 per rent of tlu- cases, a dose of approximately 757.80 ergs 'mm.- is chosen as an end- point dose. Wave length 3130 A is found to be ineffective since a dose of 5445.00 ergs •mm.- fails to produce a detectable change' in the embryos. The reciprocal of the calculated endpoint dose in ergs/mm.- is plotted against wave length for each of the four sets of data and the photochemical efficiency curves are presented in Figures 2. 3, 4, and 5. All four curves show two well-defined ab- ANALYSIS OF NHl'KAL TU'.K FORMATION 87 sorption maxima at wave lengths 2576 and 2804 A and a minimum at wave length s 2650-2700 A. The validity of using a morphological endpoint as a measure of the effect produced is shown hy the fact that the curves for the two different morpho- logical endpoints are almost identical in shape. Furthermore, the curves are very similar in shape regardless of the data used in correction for ahsorption by the vitelline membrane. This shows that the differences in effectiveness of different wave lengths as shown in the photochemical efficiency curves are truly significant. 0 14 013 012 on 010 009 \ O 0006 i 0005 0004 0.003 0002 0.001 2400 2500 2600 2700 2800 2900 3000 WAVE LENGTH IN ANGSTROM UNITS 3100 FIGURE 4. Photochemical efficiency curve. Inhibition of closure of the neural tube for a distance of at least one-third its length in 50 per cent of the embryos. Correction for extinction by vitelline membrane based upon transmission measurements with the photocell and quartz microscope. (O) and (X) are symbols for experimentally determined doses stronger and weaker respectively than the calculated end-point doses (A). neural tube seems to be restricted to an effect upon the folding process because radiation does not appear to produce other effects on the neural plate cells. The results presented in Table II show that prolonged incubation leads to closure of the neural tube in regions which were open after 30 hours incubation. This prob- ably can best be interpreted as indicating that material and not the capacity to pro- duce or utilize material is altered. Microscopic observations fail to reveal any ab- normal changes in the cells of the neural tube. The volume of the neural plate or 90 JAMES < ». DAVIS tube has increased considerably. In certain regions, the motor roots ol the spinal nerves grow out of the neural plate which shows that histological differentiation continues even though folding of the plate as a whole has been inhibited. Mitoses arc' very abundant and normal in position. The quantitative deter- minations presented in Table I suggest that mitosis is slightly affected by radiation at wave lengths from J4S3 to 2804 A. However, enough data have not been 0.014 0013 0.01 0.0 II 0.010 0.009 : 0.008 i 0007 0.006 0.005 I 0.004 0.003 0.002 0.001 2400 2500 2600 2700 2800 2900 3000 WAVE LENGTH IN ANGSTROM UNITS 3100 FIGURE 5. Photochemical efficiency curve. Inhibition of closure of the neural tube for a distance of at least one-half its length in 50 per cent of the embryos. Correction for extinction by vitellinc membrane based upon transmission measurements with the photocell and quartz microscope. (O) and (X) are symbols for experimentally determined doses stronger and weaker respectively than the calculated end-point doses (A). analyzed to settle this question. In view of the evidence presented by others on the eft eel ol radiation on avian material, it seems unlikely that radiation has an effect at these small doses. Mayer and Schreiber (1934) found that doses of radiation .several times as large as those used in this investigation were required to inhibit cell division of chick fibroblasts and chondroblasts in tissue culture. Even it mitosis is affected, the energy involved in producing this effect is evidently not involved in preventing folding. This reasoning is based upon the observation that ANALYSIS OF NEURAL TUBE FORMATION 91 doses at wave lengths 2894 and 2967 A which affect mitosis do not influence fold- ing. The same inverse relation is found in the case of the cross-section measure- ments. Consequently, the data which have been analyzed indicate that a decrease in mitosis and volume is not causally related to inhibition of the folding process. This suggests that the incident radiation which is required to prevent folding is not involved in nuclear or volume changes and. consequently, is not absorbed by the cellular material engaged in these changes. From the discussion of the histological observations, mitotic counts, and cross-section measurements, it is evident that a comparable unit of ultraviolet effect on inhibition of the folding process is obtained at different effective wave lengths. In view of this, it is possible to obtain a photo- chemical efficiency curve for this process. In order to determine accurately the photochemical efficiency of different wave lengths, corrections have been made for the energy absorbed by the material which screens the embryo. The two possible sources of error are the albumin and the vitelline membrane. The albumin present above the embryo after a 24 hour in- cubation period is negligible.- It is unlikely that any albumin present is absorbing incident radiation since wave length 2804 A which is most effective in inhibiting folding is most strongly absorbed by albumin. On the other hand, the vitelline membrane absorbs a large per cent of the inci- dent light. As can be seen by an examination of the data for the transmission measurements, different results were obtained by the two methods. With the spectrograph method only the amount of light transmitted by the vitelline mem- brane is measured and the values obtained by calculation of the incident energy on the embryo are minimum. The photocell and quartz microscope measure not only transmitted but some scattered radiation. Since the vitelline membrane lies in contact with the surface of the embryo, most of the scattered radiation is likely ab- sorbed. In view of this the photocell measurements give a better insight into the actual amount of energy incident on the embryo. It should be pointed out, how- ever, that the exact amount of scattered radiation measured with the photocell and quartz microscope is dependent upon certain experimental conditions such as the distance of the object from the objective and the diameter of the opening in the iris diaphragm. For this reason the photocell data can be relied on to give only an approximate value for the incident energy on the embryo. Since the photocell and quartz microscope transmission measurements give a better measure of the incident energy on the embryo, the efficiency curves obtained by correction with the photocell data will be considered in the comparison with ab- sorption spectra. Although the curves drawn between the points determined by interpolation are based upon the assumption that the morphological effect is pro- portional to dose, this does not influence the magnitude of the absorption maxima and minima since the two experimental points are close together at most wave lengths. As was pointed out in the first part of the discussion, the significance of the relative photochemical efficiency curves lies in the fact that they can be used to determine the chemical nature of the irradiated material. This is possible since photochemical efficiency curves are an indirect measure of the amount of energy absorbed when certain fundamental assumptions are fulfilled. These assumptions are: (1) radiation must be transmitted by the absorbing system; (2) the quantum vield for the substance absorbing the energy must be the same for all effective wave 92 JAMES O. DAVIS lengths; and (3) only the radiation which is absorbed and involved in producing the photochemical effect is measured. The validity of this technique was proved both mathematically and experimentally by Warburg (1927) and (1930). Since then, several investigators, namely, Gates (1930), Oster (1935), Giese (1938), and Landen and Uber (1939), have used this method. An examination of the validity of the data in the present investigation with reference to the three assumptions shows that the first assumption is supported by two types of evidence. From cross-section measurements of the notochord, it ap- pears that radiation affects its shape. Since the notochord lies directly beneath the neural plate, it is suggested that radiation is transmitted by the neural plate. This type of reasoning, however, is subject to the criticism that the notochord might be affected indirectly through the action of radiation on the neural plate or other structures. More conclusive evidence is presented by the dark and dis- organized appearance of the mesenchyme cells beneath the neural plate. From the experiments in this study data concerning the constancy of quantum yield with wave length are not available. In order to ascertain the quantum yield of the material engaged in folding it would be necessary to know the particular compound involved. In this investigation no attempt is made to determine the exact chemical compound but only to ascertain to what general group of compounds the material belongs. It is interesting to note that Harris, Bunker, and Mosher (1938) found the quantum yield for ergosterol to be constant for wave lengths 2537, 2652, 2894, 2967, and 3025 A. Bunker, Harris, and Mosher (1940) re- peated this experiment for 7-dehydrocholesterol and found a uniform quantum effi- ciency for all wave lengths except 2967 A which was slightly more efficient. In a study of proteins, Landen (1940) and Hollaender and Duggar (1936) found the quantum yield to be constant for wave lengths between 2400 and 3130 A. Evidence for fulfillment of the third assumption will now be considered. As shown in unpublished work by the author, inhibition of the folding process is al- most exclusively the result of an effect upon the neural plate cells. Furthermore, a very small amount of neural plate material seems to be affected by the energy in- volved in inhibiting folding since effects on mitosis and volume of the neural plate do not appear to influence the folding process. Consequently, it can be concluded that if other molecules are absorbing some of the incident energy required to pre- vent closure of the neural tube, the amount absorbed must be very small. The incident dose at wave length 2804 A required to produce the morphological effect on the neural tube is approximately 71 ergs/mm.2. When the incident energy is expressed in quanta, a value of 1.1 X 1013 quanta/mm.2 is obtained. If this figure is divided by the number of cells in a square millimeter of the neural plate which is nl" the order of 103 to 104, a value of 1.1 X 109 to 1010 quanta is obtained for the incident energy on one cell. This is a smaller dose than that required to produce just a perceptible retardation of cleavage in one sea urchin egg. Giese (1938) found that an incident energy of 623.36 ergs/mm.2 or 3.74 X 1011 quanta per egg at wave length 2537 A was the smallest dose which would retard cleavage and that wave length 2804 A was only twice as effective as wave length 2537 A. In view of this, it can be safely concluded that no more than a very small amount of energy is absorbed by molecules "screening" the inactivated material. Since this technique is valid within the limitations of the data, it is possible to compare the relative photochemical efficiency curves with the absorption spectra of ANALYSIS OF NEURAL TUBE FORMATION 93 chemical compounds. A very extensive series of absorption spectra of the most im- portant biological substances was compiled from the data of several hundred in- vestigators and published by Ellinger (1937; 1938). The major 'groups of com- pounds included were fats, carbohydrates, proteins, sterols, phosphatides, carbonic acid and its derivatives, alkaloids, and glycosides. The photochemical efficiency curves obtained in this study possess absorption maxima at wave lengths 2576 and 1.5 i.o '£ 05 O'" K X- -2 IS -25 E tr UJ a -3.0 - -35 -40 -4.11 2200 2300 2400 2500 2600 2700 2800 2900 WAVE LENGTH IN ANGSTROM UNITS 3000 3100 FIGURE 6. The photochemical efficiency curve for inhibition of the folding process in neural tube formation ( — A — ) compared with the absorption spectra of lumisterol ( ) and 7-dehydrocholesterol (- -O- -)• 2804 A and a minimum at wave lengths 2650-2700 A. Comparison of these curves with the absorption spectra of biological compounds shows that the efficiency curves resemble very closely the absorption spectra of certain sterols. The sterol curves which are very similar to the efficiency curves are for three vitamin D precursors, namely, 7-dehydrocholesterol, ergosterol, and lumisterol. In Figure 6, the photo- chemical efficiency curve for inhibition of closure of the neural tube for a distance of at least one-third its length when correction is made for absorption by the vitel- 94 JAMES O. DAVIS line membrane based upon the quartz microscope and photocell data is compared with the absorption spectra of two of these vitamin D precursors. No attempt is made to compare the efficiency curve with a particular sterol curve since absorption curves of very closely related compounds are known to vary, particularly in the position of the maxima. Von Dimroth (1939) found that the absorption spectra of sterols are dependent upon the number and location of double bonds. The position of the absorption maxima is dependent also upon whether the double bonds are located in one or in two rings. Furthermore, the solvent and pH of the mixture play a role in the position of the absorption bands. It is also possible that the chemical material which has been irradiated is an unidentified sterol or group of sterols. Consequently, an exact match of the curves cannot be expected. Attention is called to the fact that the photochemical efficiency curves do not resemble absorption spectra of single proteins. Most protein curves have only one maximum which is at 2800 A and a minimum around 2500 A which is the position of the secondary maximum in the efficiency curve presented in Figure 6. It should be mentioned, however, that a mixture of nuclear and cytoplasmic proteins might give an absorption curve similar to the photochemical efficiency curves. This is unlikely because it would involve the photochemical inactivation of two different compounds simultaneously. Furthermore, nucleoproteins which are involved in the mitotic mechanism do not seem to be involved in folding because the wave lengths which are most effective in inhibiting mitosis are least effective in inhibit- ing folding. The part that sterols play in early development is discussed by Needham (1942). He calls attention to their possible role in development as "neurogens," i.e. sub- stances which stimulate gastrula ectoderm to neural differentiation. From the present investigation it is evident that sterols are very probably involved at a slightly later stage in development, namely, in neural tube formation. Although the technique used in this study does not allow a final or conclusive identification of a specific compound to be made, it does give (1) evidence that a special com- pound is involved in folding that does not seem to be engaged in other morpho- genetic processes at this time in development and (2) a very strong indication as to the chemical nature of this material. SUMMARY 1. Chick embryos ranging in age from the primitive streak to the 8-Somite stage were irradiated with monochromatic ultraviolet radiation of wave lengths 2483 to 3130 A and subsequently incubated for 30 hours. 2. Histological studies show that radiation inhibits the folding process in neural tube formation, while cell division and volume changes continue. This effect on the neural plate is uniform for all wave lengths except 2967 A ; nevertheless, wave length 2967 A inhibits folding. 3. After correction for absorption by the vitelline membrane, the incident energy on the embryos required to inhibit the folding process was calculated. Folding is affected by all wave lengths except 3130 A and photochemical efficiency curves for the folding process arc presented. 4. In order to obtain information concerning the chemical nature of the mate- ANALYSIS OF NEURAL TUBE FORMATION 95 rial involved in folding, the photochemical efficiency curves which are an indirect measure of absorption were compared to the absorption spectra of biological com- pounds. The validity of this technique is based upon three fundamental assump- tions which are satisfied within the limitations of the available data. 5. The photochemical efficiency curves are very similar to the absorption spectra of sterols, particularly vitamin D precursors. The small doses used in the inhi- bition of folding and the high sensitivity of sterols to ultraviolet light add support to the finding made with absorption measurements that sterols are involved in the folding process. LITERATURE CITED BUNKER, J. W. M., R. S. HARRIS, AND L. M. MOSHER, 1940. Studies in the activation of sterols. Jour. Amcr. Chan. Soc., 62 : 1760-1762. DAVIS, J. O., 1942. Photochemical spectral analysis of neural tube formation. University of Missouri, Doctor's Thesis. VON DIMROTH, K., 1939. Beziehungen zwischen den Absorptions spektren im Ultraviolett mid der Konstitution organischer Verbindungen. Ant/cmmdtc Clicin., 52: 545-560. ELLINGER, F., 1937. Absorptions-Spektroskopie im Ultraviolett. I. Absorptions-Spektra der Eiweiss-Korper, Kohlehydrate und Fette einschliesslich ihre Aufbau und Abbau-Sub- stanzen. Tabulae Bioloi/icac, 12: 291-343. ELLINGER, F., 1938. Absorptions-Spektroskopie im Ultraviolett. II. Absorptionsspektra des Wassers, der Kohlcnsuure und ihrer Derivate, der Urcide, Purinbasen, Pyrimidinbascn. Nucleinsauren und deren Spaltprodukten, Alkaloide einschliesslich dcren Aufbau-und Abbaustoffen, der Glykoside und deren Aufbau-und Spaltprodukten, verschiedener Stoffc mit biologischer oder toxikologischer Bcdeutung, einiger Pflanzenfarstoffe, tierischcr und bakterieller Gifte. Tabulae Biologicae, 16 : 265-354. GATES, F. L., 1930. A study of the bactericidal action of ultraviolet light. III. The absorption of ultraviolet light by bacteria. Jour. Gen. Physiol, 14 : 31-42. GIESE, A. C, 1938. The effect of ultraviolet radiation of wave length 2537A upon cleavage of sea urchin eggs. Biol. Bull., 74: 330-341. HARRIS, R. S., J. W.. M. BUNKER, AND L. M. MOSHER, 1938. Quantitative measurement of the ultraviolet activation of sterols. I. Ergosterol. Jour. Amcr. Chan. Soc., 60: 2579- 2580. HOLLAENDER, A., AND B. M. DuGGAR, 1936. Irradiation of plant viruses and of microorgan- isms with monochromatic light. III. Resistance of the virus of typical tobacco mosaic and Escherichia coli to wave length 2250 A. Proc. Nat. Acad. Scl., 22 : 19-24. LANDEN, E. W., 1940. Quantum yield as a function of wave length for the inactivation of urease. Jour. Amcr. C'licui. Soc., 62: 2465-2468. LANDEN, E. W., AND F. M. UBER, 1939. Ultraviolet absorption spectra of active and inactive yeast. Proc. Soc. E.vf. Biol. and Mcd.. 42: 559-563. MAYER, E., AND H. SCHREIBER, 1934. Die Wellenlangenabhangigheit der Ultravioletwirkung auf Gewebekulturen ("Reinkulturen"). Proto., 21 : 34-61. McKENZiE, C. H., 1941. Diethylstilbestrol. Journal-Lancet, 61: 94-100. NEEDHAM, J., 1942. Biochemistry and Morphogenesis. Cambridge University Press. OSTER, R. H., 1935. Results of irradiating Saccharomyces with monochromatic ultraviolet light. I. Morphological and respiratory changes. Jour. Gen. Physiol., 18: 71-88. UBER, F. M., T. HAYASHI, AND V. ELLS, 1941. Ultraviolet transmission by the vitelline mem- brane of the hen's egg. Sci., 93 : 22-24. UBER, F. M., AND S. JACOBSOHN, 1938. Large quartz monochromator for biophysical research. Rn: Sci. Instr., 9: 150-152. WARBURG, O., 1927. Uber die Wirkung von Kohlenoxyd und Stickoxyd auf Atmung und Garung. B'wclicm. Zcitschr., 189 : 354-380. WARBURG, O., 1930. The enzyme problem and biological oxidations. Johns Hopkins Hasp. Bull, 46: 341-358. SEROLOGICAL RELATIONSHIPS BETWEEN THE MOLLUSCA AND OTHER INVERTEBRATES 1 RAYMOND W. WILHELMI (From the Department of Zoology, University of Missouri, and the Marine Biological Labora- tory, Woods Hole} INTRODUCTION Until recently, opinions regarding classification, interrelationships and phylogeny of organisms have depended primarily on interpretations derived from morpho- logical and developmental data. As a result of the subjective nature of such inter- pretations, concepts have differed and considerable controversy has arisen concerning the phylogenetic status of certain forms and even of entire groups of organisms. In addition to their use in diagnosis and control of many diseases, serological meth- ods have, since the turn of the century, served as an additional source of informa- tion concerning the phylogenetic relationships of plants and of animals. Of the several types of serological reactions available the precipitin test is must suitable for the determination of interrelationships of organisms. Discovered by Kraus (1897), the precipitin reaction was thought to be specific, since the blood sera of goats which had been inoculated with sterile cholera, typhoid or plague culture filtrates caused precipitates only when mixed with the particular bacterial nitrate used for immunization. That foreign proteins other than bacterial ones were antigenic and could cause the the appearance of precipitins in the blood of an injected animal became evident from the research of Bordet (1899) and of Tchistovitch (1899), working inde- pendently. Tchistovitch observed that the serum of rabbits which had been injected with horse or eel serum caused the precipitation of the antigen identical with that used in its production, and, therefore, he claimed that the reaction was specific. Of greater biological significance were the results and conclusions of Bordet, who dis- covered that antichicken serum, produced in the rabbit, reacted not only with chicken serum but also with pigeon serum, although much less strongly in the latter case ; he thus concluded that the precipitin reaction is not strictly specific. These ob- servations have been confirmed and extended by other, more recent researches. The independent discovery by Uhlenhuth (1901) that the precipitin reaction is not strictly specific, since an antiserum against one protein may react with other closely related proteins as well as with the protein used in its production, is the basis of the application of serological reactions to biological problems involving inter- relationships of organisms. He concluded that precipitin reactions are quantita- tively, as well as qualitatively, specific, i.e., any antiserum will react more strongly with the antigenic substance used in its formation (homologous reaction) than with other antigcnirally different, though closely related, substances (heterologous reac- 1 Tliis research \\as conducted with the aid of funds supplied by the University of Missouri Research Council. 96 SEROLOGICAL RELATIONSHIPS 97 tions), the degree of relationship being indicated by the strength of the heterologous reactions as compared to the homologous one. Ascoli (1902), Graham-Smith and Sanger (1903), and Fornet and Mviller (1908, 1910) improved upon the flocculation type of precipitin technique by care- fully introducing the antiserum into the precipitin tube below the antigen so that a definite interfacial boundary was maintained ; at the interphase between the two reagents a layer or "ring" of precipitate formed and the end-point or titer could be determined with greater precision. To Nuttall (1904) belongs the credit for the first extensive application of the precipitin reaction to the problem of animal relationships or phylogeny. Several thousand qualitative and quantitative reactions were performed with sera of ap- proximately six hundred species of animals, and the degree of reaction between any particular antiserum and several different antigens was found to be in proportion to the degree of relationship between the animals concerned, i.e., the results of the precipitin tests confirmed, in general, the classification and relationships to the ani- mals based on morphology and development. During the past two decades, the researches of Professor Alan Boyden and his students have yielded outstanding contributions not only to our knowledge of animal relationships, particularly among the vertebrates, but also, and more significantly, to refinements in techniques and methods. Boyden (1926), employing the "ring" technique, obtained confirmation of his quantitative reactions and titers by re- ciprocal tests, the importance of which he summarized by stating that "it would seem then that this principle of reciprocal relationships could be used as a test of the truth of the values obtained in the precipitin reaction. . . . Only those values which check within the limits of error of the reaction may then be taken." More recently, by employing a highly sensitive instrument, the photronreflectometer (cf. Libby. 1938), for measuring turbidities developed by the precipitin reaction (flocculation technique), Boyden and his students (Chestnut, DeFalco. Gemeroy, Leone) have approached with even greater objectivity the solution of relationship problems within restricted groups of animals. Serological relationships indicate relative physiological and chemical affinities, and presumably, therefore, genetic and phylogenetic relationships, since the protein constitution is the basis for morphological as well as physiological characteristics. With the advent of and improvement upon the precipitin reaction an objective method was provided for solving problems of phylogenetic relationships. How- ever, until recently, the variations in results of precipitin reactions in experiments designed to aid in phylogenetic studies were so great that one could not be sure of interpretations which the results may have suggested. Mez and his students have based a phylogenetic tree of plants upon the results of precipitin reactions and reported agreement between such relationships and those derived from morphological considerations. Zoologists have not yet constructed phylogenetic trees based solely upon results of serological tests. At the present state of our serological information, construction of such trees would be impossible, but, with the improvements in and greater uniformity of techniques and with the accumulation and interpretation of data, such attempts may be made. The relationship of the Mollusca to other invertebrates has long been the source for considerable speculation. Embryological evidence would indicate that annelids, arthropods and mollusks are related by reason of (1) the fate of the blastopore 98 RAYMOND W. WILHELMI which becomes the mouth, placing these animals in a group known as Proterostomia, (2) the formation of the mesoderm which is by proliferation from two mesoblastic cells, the mesodermal bands later splitting to form schizocoelous coelomic spaces, and (3) the type of cleavage, which is, generally speaking, determinant. The present paper will report some experiments involving precipitin reactions designed to determine the serological relationship of the Mollusca to other inverte- brates. MATERIAL AND METHODS Antigens were prepared from representatives of four invertebrate phyla : ANNELIDA, Nereis vircns; ARTHROPODA, Limulus polyphemus; MOLLUSCA, Busycon curica, Busycon candiculatuni, Pcctcn irradians; ECHINODERMATA, Astcrias jorbesi. It will be noted that the habitat of all these animals is marine. Preparation of antigen Preparation of antigens was essentially similar to that reported in earlier pub- lications (Wilhelmi, 1940, 1942). Although numerous difficulties were encount- ered, particularly in the extraction of annelid and molluscan materials, all antigens are extracts of tissues. Following a rather long preliminary period of exposure to filtered sea water to remove certain contaminating organisms and debris, the annelids were washed through several changes of sterile sea water contained in large, covered crystallizing dishes and finally rapidly through sterile distilled water to remove excess salts. In the case of Limulus and the mollusks, following the preliminary treatment of the specimens with filtered sea water and washing with 70 per cent alcohol, exo- skeletal structures were removed. Asterias specimens were treated in a similar man- ner, but before opening specimens the tips of the arms were cut and the coelomic fluid drained into a sterile receptacle to be frozen in dry ice ; then the aboral surfaces of the arms were removed exposing the inner soft tissues. Antigens of the respec- tive animals were prepared by extracting the soft tissues, exclusive of the digestive tract. Materials were handled independently and manipulated carefully to avoid contamination. Tissues of the living animals were minced on a glass plate or in a shallow (Petri) dish and, after placing in a sterile container, frozen rapidly by means of dry ice. When frozen, the material was placed in a vacuum desiccator, evacuated by a Cenco-Hyvac pump, over drierite (anhydrous calcium sulfate) until thoroughly dried. By this method the native chemical structure of antigenic substance is pre- served, since the rapid freezing does not permit deterioration or denaturation of constituent chemicals. The high vacuum causes such rapid evaporation of water that substances remain frozen, the dehydration is more rapid, and oxygen, which might alter the antigens, is removed. Desiccation in the frozen state usually results in dried material with a spongy texture which renders it easy to pulverize. How- ever, in the present experiments considerable difficulty was experienced in effecting complete desiccation of the annelid and molluscan materials in the 24-hour period usually required by the process. When the material appeared to be completely desiccated, it was triturated, using SEROLOGICAL RELATIONSHIPS 99 a mortar and pestle, and then returned to the desiccator under vacuum for an addi- tional 48 hours or until extractions were to be made. Two types of antigens were prepared from the desiccated tissues. One type was a saline extract of the entire animal, and the other, a saline extract of the lipid- free materials of the animal. To prepare the first, desiccated material was weighed and then extracted for 12 to 24 hours with sterile, buffered saline solution main- taining a pH of 7.3, one part by weight of dried substance being mixed with ten parts by volume of extractant. Extraction was done at room temperatures in sterile Erlenmeyer flasks under constant agitation provided by a mechanical shaker. To prepare the second type of antigen, the lipids were removed from the material by repeated extractions with Bloor's mixture, consisting of three parts absolute ether to one part absolute alcohol, one part by weight of powdered material being mixed with one hundred parts by volume of Bloor's mixture. After weighing the lipid- free residue, it was extracted with sterile buffered saline solution using the same methods as employed to obtain saline extracts of the entire animal. After aqueous extraction the material was filtered through sterilized Seitz filters, and the clear, sterile filtrate was transferred to sterile, rubber-stoppered vac- cine bottles and kept in the refrigerator until needed for injections. The residue from the aqueous extraction was desiccated and weighed, and the exact amount of material in solution was determined by calculating the difference in weight of residues before and after the aqueous extraction. Antibody production For injection purposes, healthy rabbits of equivalent weights were used follow- ing a preliminary period of observation. For the experiments reported herein, a longer method of antibody production was employed than that reported in earlier papers. Each rabbit received eight intravenous injections, one every other day. In the series of eight injections, a total of 500 mg. of dry-weight antigen was injected. The 500 mg. was divided into doses of increasing amounts, i.e., ap- proximately 25, 50, 50, 50, 75, 75, 75 and 100 mg., respectively. Ten to 14 days after the last injection, exsanguination of the rabbit was effected by aseptic, intracardial puncture, using a large, sterile hypodermic syringe fitted with a 19-gauge needle, and the blood was placed in large (250 cc.), sterile, cotton- plugged centrifuge tubes and allowed to clot at room temperatures. After standing in the refrigerator for several hours, the clotted blood was centrifuged ; the serum was drawn off and used immediately for tests or filtered through a sterile Seitz filter and placed in sterile vaccine bottles for storage in the refrigerator until needed for the precipitin reaction. Precipitin test In a precipitin tube (8.0 mm., inside diameter), 0.5 cc. of undiluted antiserum was carefully overlaid by 0.5 cc. of appropriately diluted test antigen so that a definite interfacial boundary between the two reagents was maintained. The antigen was brought to the proper concentration by adding sterile saline solution buffered to a pH of 7.3. For most tests a standard dilution of each test antigen was prepared so that 1.0 gram of dry-weight material, either whole-animal or lipid- 100 RAYMOND \V. \YIUIKI. MF tree (depending upon \vhicli was needed), \vuuld be contained in 100 ce. of saline >olution, i.e., 0.01 g. cc. or a dilution of 1 : 100. In homologous tests this standard was diluted to 1 : 1000 for the initial tube, and the dilution in each succeeding tube of a series was doubled. M» that a series of antigen dilutions from 1 : 1000, through 1 : 2000, 1 : 4000. etc. up to 1 : 2.048,000 was employed. To make some of the heter- ologotis precipitin tests it was necessary to evaporate the test solutions to the desired antigenic concentrations, evaporation being effected by placing the extract in sterile Petri dishes in a partial vacuum over drierite. A dilution of 1 ; 10 was employed in the initial tube, and the concentration of antigen was halved in each successive tube. In a positive reaction, a "ring" or layer of white precipitate occurred at the inter- pha.-e of the antiserum and antigen. The tiler of a reaction is the highest antigen dilution which yields a visible pre- cipitate within one hour at 37° C. when tested with an antiserum. either homologous or heterologous, the amount of antigen being determined on the basis of dry weight of antigen material actually in solution. In all cases the readings and titers were taken at the end of one hour at 37° C. Control tests were always conducted and involved the use of (T) normal sera and antigen solutions. (2) immune sera and the extracting salt solution, and (3) normal sera and the extractant. OBSERVATIONS AND RESULTS Table I presents the titers of both homologous and heterologous reactions be- tween the several species of invertebrates employed in these experiments. Each titer represents the average of at least four, and in many cases six or eight, deter- minations at 37° C. Negative reactions are indicated by minus signs ( — ), and tests which were not performed, by blank spaces. The letters u' and c refer to wrhole-animal and lipid-free materials, respectively. The titers of homologous tests cover a wide range, from 1 : 49,000 to 1 : 512,000, results from the use of lipid-free antigens being the only ones considered reliable. Under the conditions of these experiments, the species-specific titer may be desig- nated as 1 : 128,000. However, in two species (Busycon curica and wisterias forbesi) the titers did not fall within the limit of experimental error (plus or minus one dilution tube) allowed in the precipitin reaction, although in each case the titer was only one dilution tube removed from the accepted limit of error. The titers of heterologous reactions never exceeded those of homologous ones. In general, reciprocal tests confirmed the original ones, i.e., were of the same order of magnitude. Such confirmation is imperative in experiments designed to determine phylogenetic relationships by serological methods. Particular attention is directed to the titers of homologous and heterologous re- actions when whole-animal extracts were used as antigens. It will be noted that the variability is exceedingly great and that several inconsistencies appear, particu- larly in the heterologous tests; for example, the titer (1 :4000) of the reaction in- volving antiserum against Liinulus polyphemus whole-animal materials and antigen prepared by extracting whole Uitsvcon carica is higher than that (1:1000 or 1 : 2000) between Pcctcn irradians and Busycon carica materials, although Pecten is undoubtedly more closely related to Busycon than is Limulus. As a matter of fact, the only cross-reactions between arthropod and molluscan materials occurred SI ROLOGICAL RELATIONS! 1 1 1'S 101 TABLE I Homologous and heterologous prccipilin reactions (Number in parenthesis below certain tilers indicates relationship expressed as per cent value of homologous titer) Antiserum Antigen \rrcis virens Limulus Polyphemus Busycon carica Bit *ycon canaliculatum Pecten irradians Asterias forbesi e w e w e w e e e Nereis virens e 106,000 (100) — 1,000 (0.78) (0.000) — 25 (0.0195) 100 (0.0195) Li mill us poly- phemus w 64,000 400 e 500 (0.47) 16,000 128,000 (100) — (0.000) (0.000) (0.000) (0.000) Busycon carica .w 4,000 16,000 e 10 (0.0094) — 2,000 40,000 (100) 49,000 (36) 2,000 (1.887) (0.000) Busy con canal iculat it m w 1,000 8,000 24,000 e 40 (0.038) — (0.000) 2,000 13,333 (27.2) 6,000 128,000 (100) Pecten irradians e (0.000) — (0.000) 1,000 (2.04) 1,000 (0.78) 106,000 (100) (0.000) A sterias forbesi e 40 (0.038) — (0.000) (0.000) — (0.000) (0.000) 512,000 (100) when whole-animal materials had been utilized as antigens in preparation of the antisera. Other interesting comparisons can be made by reference to the Table. The figures in parenthesis represent the titration data converted to percentages of the homologous reaction titers. It will be noted that percentages were calculated only for those reactions involving lipid-free materials. The homologous titer of a particular antiserum is taken as 100 per cent and the titers of all heterologous tests against that antiserum are expressed as percentages of the homologous titers. For example, Asterias forbesi antigen when tested against Nereis virens anti- serum yields a titer .of 1:40; the homologous titer of Nereis virens antiserum 40 being 1 : 106,000, the percentage value becomes i or 0.038 per cent. DISCUSSION Homologous reactions of high titer were thought necessary if heterologous re- actions between such distantly related groups as the Mollusca and other invertebrate phyla were to be obtained. In a series of preliminary experiments in which the same method of producing antibodies was used as was reported previously (Wil- helmi, 1942), i.e., the short method of antibody production, all precipitin tests be- 102 RAYMOND W. WILHELMI twcen materials of Mollusca and other invertebrate phyla were negative, although the homologous reactions yielded titers which were higher and more consistent than those reported herein. Since Wolfe (1935) had reported that the titers of heter- ologous reactions could be increased by extending the series of injections of antigen over a longer period of time, a longer method of antibody production was attempted. The fact that the homologous titer is lower than in previous experiments is a con- firmation of Wolfe's results. Since positive reactions were not obtained in the preliminary experiments men- tioned above, in addition to extending the period of injections, whole-animal ex- tracts, as well as lipid-free materials, were employed as antigens, since in the pre- liminary tests only lipid-free antigenic materials had been used. The purpose of the experiments with lipid-containing antigens was to ascertain whether, in case of failure to produce cross-reactions with lipid-free antigens, positive interphylar tests between Mollusca and materials of other invertebrate phyla could be obtained under any circumstances or conditions. It will be recalled that the only cross-reactions between arthropod and molluscan materials occurred when whole-animal materials had been utilized as antigens in the preparation of antisera. From earlier research and from reports by other authors, Wilhelmi (1942) had concluded that, although the zoological relationships of organisms are closely paralleled by the serological and chemical properties of their respective proteins, i.e., the more distant the phylo- genetic relationships of organisms, the greater the difference between their constitu- ent proteins, the lipids of distantly related species may be very similar, if not identi- cal, immunochemically. In the experiments reported in that paper, lipids had not been employed as antigens in direct tests, but the present experiments lend ample support to the viewpoint that lipids of distantly related species may be very similar immunochemically. It is very apparent that the specificity of antisera produced by injection of lipid-containing antigens is dependent not upon the proteins contained therein but is controlled by the lipids present. That is to say, although the ar- thropod and molluscan proteins are not closely enough related chemically to pro- duce cross-reactions, their haptenic lipids, rendered antigenic when in combination with proteins, are immunochemically so similar that they produce antisera which will yield positive precipitin reactions even at relatively high dilutions. In view of the results of these tests, of the observation of Boyden (1936) in research upon Crustacea that "the convergent reactions obtained with antisera to native serum . . . disappeared following ether extraction," and of the conclusion by Cumley (1939) from experiments with Drosophila spp. that the presence of ether-soluble material in antigens interferes with the reliability of the reactions, it may now be stated that, in all experiments designed to determine the phylogenetic relationships of organisms !)}• means of serological methods, antigens must be prepared from lipid-free mate- rials. Since antisera prepared by injection of lipid-containing antigens yield un- predictable, inconsistent and divergent results, the inescapable conclusion is that, in all experiments dealing with serological relationships of organisms, demonstrable relationships are of doubtful value and resulting interpretations questionable unless experiments are performed with lipid-free materials. Lipid-free antigens, whether the source is blood sera or the various body tissues and organs of organisms, must be prepared and used in all relationship studies employing the precipitin "ring" test. The adult morphology of the Mollusca yields very little in the way of satisfac- tory clue< to the phylogeny of the group, although the arrangement of the ganglia SEROLOGICAL RELATIONSHIPS 103 of the central nervous system may represent a palingenetic structural organization on a segmental plan. In Amphineura and some of the lower Gastropoda there is a ladder-like nervous system, resembling that of some Turbellaria and of the most primitive worm-like arthropods (Peripatus), and a condition considered by some to be a pseudometameric arrangement of organs. When one studies the embryo- logical development in the group, the affinities of the mollusks to other proterostom- ian invertebrates become apparent because of the facts that (1) cleavage is of the determinant type, (2) the blastopore becomes the mouth, (3) the mesoderm is formed by proliferation from two mesoblastic cells, (4) the coelomic spaces arise by schizocoelous methods, and (5) the trochophore, or a modification, the veliger, larva occurs in the life cycle. Lameere (1932) regarded the Mollusca as derived from the nereidiform annelids near the Amphinoinidae. However, Nierstrasz and Stork (1940) contend that they are not phylogenetically closely related to the an- nelids, but that their organization is more readily comparable with that of the Turbellaria, especially with respect to the tendency towards pseudometamerism and the structure of body wall and nervous system. Parker and Haswell (1930) argue that the occurrence of the trochophore larva in the life cycle of the Mollusca need not necessarily be regarded as evidence of their derivation from the Annelida. "In fact the absence of segmental repetition of parts in all, with the exception of Nauti- lus, would seem to indicate the derivation of the phylum from a group in which metamerism had not arisen. It will be readily recognized that the gap between the typical trochophore and certain forms of Turbellarian larvae (Miiller's larva) is not a very wide one, and might be covered by adaptation of the larval Flat-worm to a freer pelagic life. If we were to suppose that the most primitive Mollusca were derived from Turbellarian-like ancestors, the conversion of a larva of the type of Miiller's larva into a larval form like the molluscan trochophore would also have to be postulated. This might involve a common platyhelminth origin for Annulata and Mollusca with subsequent extreme divergence — a divergence in which the re- spective trochophores would take part, though in a limited degree." The results of the present experiments support the conclusion that the Mollusca are serologically more closely related to the Annelida than to any other invertebrate phylum tested, since interphylar reactions did not occur between the representative Mollusca and those of any other phylum except the Annelida when lipid-free anti- genie materials were employed to produce antisera and to make the precipitin test. Under the conditions of the experiments annelid-molluscan titers were of the same order of magnitude as annelid-echinoderm titers. However, it must be remem- bered that, in order to produce reactions between molluscan materials and those of other proterostomian phyla, the longer method of antiserum production had to be used and, as pointed out by Wolfe, such antisera are less specific and give greater group reactions. The fact that no positive reactions were noted in molluscan- arthropod tests is in agreement with results of intradermal tests performed by Tuft and Blumstein (1940) in which it was found that neutralization with one of the members of the Mollusca obliterated reactions to all other members of that group but not to those of the Crustacea. Thus, animals of the Mollusca group and those of the Crustacea group bear no antigenic relationship to each other ; they concluded that there is a common antigen among Mollusca and that a complex molecule with several antigenic fractions is present among the Crustacea. In the present experi- ments intraphylar tests between the molluscan species were done in order that a 1»4 UAVMOXD \V. \VIL11KI.MI inference point would be available for interpretation of the interphylar reactions. The values 36 per cent and 27.2 per cent, representing reciprocal values of tests be- tween materials involving the two species of Busycon (B. carica and B. canallcn- /(//»>;/). are not of the same order of magnitude as the percentage value (85 per cent) reported by Moyden (1943) for members of the same genus, but more nearly approach his values for members of closely related families; however, his data were obtained by the use of the photronreflectometer upon vertebrate materials and perhaps, therefore, are not strictly comparable. In the same paper. Boy den, in ref- erence to unpublished research by Chestnut, reports that the photron'er has been, or is being applied, to the study of Mollusca. Makino (1934), by means of pre- cipitin (flocculation), complement fixation and anaphylactic reactions, was able to differentiate the various genera of Mollusca which he employed and to group them according to their relationships. In the present experiments, the tests between Busycon carica antigen and Pcctcn irradians antiserum yielded a relationship value of 1.887 per cent, and the reciprocal tests, 2.04 per cent, so that interclass tests among Mollusca presumably yield percentages in the vicinity of 1 per cent to 3 per cent. Interphylar values are all less than 0.78 per cent, all between Mollusca and other invertebrate phyla being less than 0.04 per cent. It is apparent that serologically, as well as developmentally, the Mollusca are more closely related to the Annelida than to any other group of invertebrates tested. It should be emphasized, however, that the results of these precipitin tests need not be interpreted to mean that the Mollusca arose from Annelida, since the precipitin reaction is a measure of present chemical relationship of the antigenic constituents of the protoplasm of existing species. These reactions may be interpreted to indi- cate that the Mollusca evolved from animals which also gave rise to present-day Annelida, and perhaps a Turbellarian ancestry such as proposed by Parker and Haswell (1930) will be substantiated by future research and investigation. SUMMARY AND CONCLUSIONS The precipitin reaction was applied to the problem of the relationship of the Mollusca to other invertebrate phyla. From the results, it is apparent that sero- logically, as well as developmentally, the Mollusca are more closely related to the Annelida than to any other group of invertebrates when lipid-free materials are employed as antigens. The results indicate that Mollusca evolved from animals which also gave rise to present-day Annelida. Consistent results, confirmed by reciprocal tests, were obtained by injection of aqueous solutions containing 500 mg. of dried antigenic material from which the lipids had been removed. Injection procedure involved eight intravenous injec- tions of antigen on alternate days and with gradually increasing dosage. To test the influence of the lipids, one group of experiments involved the use of whole- animal, i.e., lipid-containing antigens. The inconsistent and divergent results of the use of autisera resulting from these antigens dictate that, in relationship studies, whether performed with tissue extracts or blood sera, it is essential that lipid-free materials be used in order to obtain accurate and reliable results and to preclude the possibility of group reactions. With two exceptions, the liter of homologous reactions is 1:128,000 plus or minus one dilution tube, and hcterologous titers never exceeded homologous ones. SEROLOGICAL- RELATIONSHIPS 105 LITERATURE CITED ASCOLI, M., 1902. Ueber den Mechanismus der Albuminurie durch Eiereiweiss. Miinchen mcd. Woch., 49: 398-401. BORDET, J., 1899. Le mecanisme de 1'agglutination. Ann. de I'Inst. Pasteur, 13: 225-250. BOYDEN, ALAN, 1926. The precipitin reaction in the study of animal relationships. Rial. Bull., 50 : 73-107. BOYDEN, ALAN, 1936. Serological study of the relationships of some common invertebrata. Ami. Rcf>. Tortugas Lab., Carnegie Inst. Washington, 34 : 82. BOYDEN, ALAN, 1942. Systematic serology : a critical appreciation. Ph\siol. Zool., 15: 109- 145. BOYDEN, ALAN, 1943. Serology and animal systematics. The Amcr. Naturalist, 77 : 234-255. BOYDEN, ALAN, AND RALPH J. DEFALCO, 1943. Report on the use of the photronreflectometer in serological comparisons. Physiol. Zool., 16: 229-241. CUMLEY, RUSSELL W., 1939. The relations among Drosophila species, as determined by the complement fixation reaction using ether-insoluble fractions. Jour. E.vp. Zool., 80: 299-314. DEFALCO, RALPH J., 1942. A serological study of some avian relationships. B'wl. Bull., 83 : 205-218. FORNET, W., AND M. Mi'Li.KK, 1908. Zur Hcrstdlung und Vervvendung pra/ipitierender SCT;I, insbesondere fiir den Nacluvcis von Pferdfleisch. Zcitschr. {. B'wl. Tcchnik und Mcthodik, 1: 201-206. FORNET, W., AND M. MULLER, 1910. Praktische und theoretische Prazipitinuntersuchungen. Zcitschr. f. Hyg. und Infcctionskrankh., 66: 215-243. GRAHAM-SMITH, G. S., AND F. SANGER, 1903. The biological or precipitin test for blood con- sidered mainly from its medico-legal aspect. Jour. Hyg., 3: 258-291. KRAUS, R., 1897. Ueber specifische Reactionen in keimfreien Filtraten aus Cholera, Typhus, und Pestbouillen-culturen, erzeugt durch homologes Serum. Wicn. klin. Wochcn., 10 : 736-738. LAMEERE, A., 1932. Precis dc zoologie. Tome III. Fasc. 1. Les mollusques. H. Cauwen- berg, Bruxelles. LIBBY, RAYMOND L., 1938. The photronreflectometer — an instrument for the measurement of turbid systems. Jour. Innintnol., 34: 71-73. MAKING, KATASHI, 1934. Beobachtungen iiber die Immunitatsreaktionen bei Molluskenarten. Zcitschr. f. Immunitdtsforsch., 81 : 316-335. MEZ, C, 1926. Die Bedeutung der Sero-Diagnostik fiir die stammesgeschichtliche Forschung. Bot. Archiv., 16: 1-23. NIERSTRASZ, H. F., AND H. A. STORK, 1940. Monographic der Solenogastren des Golfes von Neapel. Zoologica (Stuttgart), Heft 99, Band 36, Lief. 5: 1-92. NUTTALL, GEORGE H. F., 1904. Blood immunity and blood relationships. Cambridge Univer- sity Press, Cambridge, England. PARKER, T. J., AND WM. A. HASWELL, 1930. A text-book of zoology, Vol. I. The Macmillan Company, New York. TCHISTOVITCH, TH., 1899. fitudes sur I'immunisation contre le serum d'anguilles. Ann. dc I'Inst. Pasteur, 13: 406-425. TUFT, Louis, AND GEO. I. BLUMSTEIN, 1940. Studies on food allergy. 1. Antigenic relation- ship of shellfish. Jour. Allergy, 11: 475-487. UHLENHUTH, P., 1901. Weitere Mittheilungen iiber meine Methode zum Nachweise von Menschenblut. Deutsche mcd. Wochcn., 27: 260-261. WILHELMI, RAYMOND W., 1940. Serological reactions and species specificity of some hel- minths. Biol. Brill,, 79: 64-90. WILHELMI, RAYMOND W., 1942. The application of the precipitin technique to theories con- cerning the origin of vertebrates. Biol. Bull., 82: 179-189. WOLFE, H. R., 1935. The effect of injection methods on the species specificity of serum pre- cipitins. Jour. Immunol., 29: 1-11. Vol. 87, No. 2 October, 1944 THE . BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY BIOCHEMICAL FACTORS IN THE MAXIMAL GROWTH OF TETRAHYMENA1 VIRGINIA C. DEWEY 2 Arnold Biological Laboratories, Broivn University, Providence, Rhode Island INTRODUCTION The earliest attempt to elucidate the nutritional requirements of the ciliate Tetrahymena gcleii (Glaucoma pirijonnis} by the pure culture technique was made by Lwoff (1924). At that time the failure of the ciliate to grow in solutions of pure ammo acids was attributed to a lack of specific chemical supplements. Later Lwoff (1932) suggested that a requirement for polypeptides was responsi- ble for the lack of growth in such media. This early work indicates immediately that the problem of the nutrition of T. geleii is a dual one. No investigation of the nitrogen requirements may be made without some knowledge of the supplementary factors needed. So far it has been shown that thiamine is important if not absolutely necessary to the nutrition of T. gclc'ii, while riboflavin, pantothenic acid, nicotinic acid and pyri- doxine probably play a part (Elliott, 1935b, 1939; Lwoff and Lwoff, 1937, 19*38; Hall, 1940a, b, 1942; Baker and Johnson, 1941; Kidder and Dewey, 1942). In a preliminary report Dewey (1941) indicated that other factors of unknown nature are required for maximal growth. A great deal less work has been done on the nitrogen nutrition (Elliott, 1935a; Hall and Elliott, 1935; Dewey, 1941; Hall, 1942). None of the results of these investigations is conclusive. The report of Kline (1943) that T. geleii (Colpidiitin striatiim) will grow in an amino acid solu- requires confirmation. An attempt to obtain knowledge of the supplementary requirements necessi- tates the use of a basic medium capable of supplying the nitrogen and carbon needs of the organism and ideally completely lacking in supplementary factors ; the testing of all known growth promoting substances, and the search for and purification of possible unknown growth factors. I take pleasure in acknowledging my indebtedness to Professor G. \V. Kidder of Brown University for his invaluable advice and encouragement. Thanks are due to Dr. R. J. Williams for a gift of pure calcium pantothenate and to Dr. Samuel Lepkovsky for the gift of a concentrate of his factor I. 1 Presented to the Graduate Council of Brown University in partial fulfilment of the re- quirements for the degree of Doctor of Philosophy. - Aided by grants from the Manufacturers' Research Fund for Bacteriology and Proto- zoology and from the Miss Abbott's School Alumnae Fellowship Fund at Brown University. 107 108 VIRGINIA C. DEWEY MATERIAL AND METHODS Preliminary experiments were carried out using four strains of Tetrahymena gclcii and one strain of T. 1'orax. Since only quantitative differences were found between these strains, later work was confined to strain W (Claff, 1940; Kidder, 1941a). This strain was selected because it showed the most rapid growth rate and the greatest resistance to increased salt concentration of the medium. Certain experiments were also carried out using strain H, obtained from Dr. R. H. Hall. All experimental media were prepared with water distilled twice in an all- Pyrex still over potassium permanganate and all the glassware used was Pyrex. Glassware was cleaned by soaking for at least one hour in a hot saturated solution of trisodium phosphate or a preparation sold commercially as "Keego," followed by careful rinsing to remove all traces of the cleaning agent. Stock cultures were maintained in a 2 per cent solution of Difco proteose-peptone and also for a time in a one per cent solution of crude casein (Eimer and Amend). The fact that these ciliates could be maintained in solutions of crude casein sug- gested the use of a highly purified casein as a basic medium. Casein Harris (high- est chemical purity) was used as a one per cent solution. Although this prepara- tion has been called "vitamin-free" casein, this term is somewhat misleading in that traces of certain growth substances appear to be present in amounts sufficient to affect the growth of protozoa. It is, however, still adequate as a basic medium for testing responses to growth stimulants because the growth in it alone is still far from optimal with regard either to the rate of growth or the size of the population supported. For the purposes of this paper the growth in crude casein is regarded as being optimal. Since casein, in order to be available to the ciliates, must be in solution or at least colloidal suspension, the following method was used to disperse it. To the casein was added sufficient alcohol to completely wet it, then water containing 1.0 to 1.5 ml. of normal NaOH per gram of casein was poured in slowly with stirring. With vitamin-free casein it was found necessary to add a balanced salt solution in order to obtain growth. For this purpose the modified Osterhout solution em- ployed by Barker and Taylor (1931) was used in all but the first series of experi- ments. The concentrated stock solutions were added to the water used to make up the medium. The suspension of casein was allowed to stand with occasional stirring until solution was complete. Then normal HC1 was added drop by drop with stirring between drops until the alkali had been neutralized. Care must be taken to add the acid slowly in order to prevent precipitation of the casein. The reaction of the medium was adjusted to pH 6.8-pH 7.0. The medium was then dispensed into tubes in amounts of 5 ml., plugged with cotton and autoclaved for 15 miii. at 15 Ibs. pressure. Elliott's (1939) report that vitamin-free casein re- quired digestion with pepsin before it could be utilized by the ciliates is possibly due to a failure to put the casein into solution. Kxperimental cultures were examined for growth, the results recorded and transplants made at intervals of 4S hours in casein media. The interval was slightly longer in gelatin media. The tube's were kept, however, for from one to two weeks and re-examined at intervals. Such a procedure gives results which indicate the presence or absence of factors necessary to maintain growth at or near a maximum rate. These results cannot, of course, be compared with those obtained by in- FACTORS IN THE GROWTH OF TETRAHYMENA 109 cubating cultures for a week or more before examination, since by that time a slowly growing culture may have reached a concentration equal to that of a rapidly growing culture. In order to eliminate the effects of carry-over only the results of the third transplant in a given medium are considered. Growth is recorded as zero to four plus by comparison with growth in a control medium. Growth re- corded as zero may indicate survival of the inoculum or an increase of one or two divisions, while four plus growth represents a population of from 75,000 to 100,000 organisms per ml. Two or three plus growth is intermediate. The cultures were kept at room temperature (20°-22° C). Transplants were made using an open bacteriological loop and ordinary bacteriological technique. Cultures were also incubated in Kidder culture flasks. The third transplant in a tube was inoculated into the flask and the growth followed by making counts at intervals of 12 or 24 hours (Kidder, 1941a). Gelatin in concentrations of 1 per cent, 1.5 per cent and 2 per cent was also used in certain experiments. Both Harris gelatin (vitamin-free) and Eastman de-ashed gelatin were used. Another medium consisted of 1 per cent silk peptone (Seidenpepton, Hoffman-LeRoche). These media were in some cases supple- mented with amino acids in various concentrations as well as with various growth promoting substances. In other cases solutions of the pure amino acids alone were used. The following amino acids were obtained from the Eastman Kodak Co. : 1-histidine, 1-leucine, dl-threonine, dl-/3-phenylalanine, dl-methionine, d-argi- nine carbonate and d-lysine hydrochloride. From the Hoffman-LaRoche Co. 1-tryptophane, d-isoleucine, dl-valine and glycine were obtained and from Eimer and Amend, tyrosine. The basic medium was supplemented with vitamins and growth factors of known chemical composition as well as with crude extracts of animal and plant material. The known compounds were supplied (with one exception to be noted later) at a level of 0.001 mg. per ml. except in the case of i-inositol, which was used in a con- centration of 0.004 mg. per ml., and biotin, which was used in a concentration of 0.00008 mg. per ml. Thiamine hydrochloride and riboflavin were obtained from the Hoffman-LaRoche Co. A sample of calcium pantothenate was obtained from Dr. R. J. Williams and subsequent calcium pantothenate as well as biotin methyl ester from the S. M. A. Corp. Pyridoxin hydrochloride (first used as factor I concentrate of Lepkovsky) was obtained from Merck and Co. and nicotinic acid, pimelic acid, i-inositol, uracil and p-aminobenzoic acid from the Eastman Kodak Co. Water extracts of crude casein, egg yolk, yeast (Harris), timothy hay and alfalfa meal (Denver Milling Co.) were also used. Only the last two were used in routine culturing and in experiments on fractionation. The extracts were prepared by boiling 50 g. of material with a liter of water for ten min. and filtering with suction using Celite, analytical grade (Johns-Manville), as a filter aid. The timothy extract was used in a dilution of 1:5 and the alfalfa extract in a dilution of 1 : 10. The crude extracts were treated in various ways in an attempt to remove the protein present as a preliminary to a study of the nitrogen requirements of T. gcleii. The results of these fractionations were so interesting that studies on them were continued while the work on the nitrogenous nutrition was in progress. Tests for protein or its degradation products were made by the ninhydrin reaction. One of the first methods tried for the removal of protein was precipitation with 110 VIRGINIA C. DEWEY lead acetate. A 25 per cent solution of normal lead acetate was added to the extract until precipitation was complete. The precipitate was then filtered off with the aid of suction and Celite. Excess lead was removed from the filtrate and the precipitate was decomposed by the use of phosphate. In the first experiments a 5 per cent solution of phosphoric acid was used but later a saturated solution of trisodium phosphate was found to give better results. The precipitated lead phos- phate was then removed by filtration with suction. A similar technique was used in the preparation of the fractions obtained with ferric oxide hydrosol (prepared according to the method of Thomas and Frieden, 1923). Barium hydroxide was used as a precipitant after the addition of three volumes of alcohol to the extract. This latter step was necessary because barium hydroxide alone caused little precipitation when added to the aqueous extract. In this case barium was removed by the use of sulfuric acid and the alcohol by boiling. When phosphotungstic acid was used the extract was first made acid by the addition of sulfuric acid to give a concentration of 50 per cent. The sulfuric acid alone caused the formation of a precipitate which was removed by filtration before the addition of the phosphotungstic acid. After separation of the phosphotungstic acid precipitate both the filtrate and the precipitate were treated with barium hydroxide to remove the sulfuric acid and the phosphotungstic acid. The method given by Peters and Van Slyke (1931) for the removal of carbo- hydrate by the use of copper sulfate and calcium hydroxide was tried and found to remove all the reducing sugars present in the extracts, although protein was not removed. Excess calcium was removed from the filtrate either as the carbonate or the phosphate. Precipitation was also carried out by the use of alcohol or acetone. After add- ing sufficient alcohol or acetone to give the desired concentration the extracts wrere allowed to stand until flocculation was complete. The precipitates were then filtered off. To remove the precipitant the filtrates were boiled or distilled, some- times under reduced pressure. The precipitates were redissolved in water. The above named solvents as well as ether or acetic acid were also used to make extracts of alfalfa meal, using a Soxhlet apparatus except in the case of the acetic acid. The solvents were removed from the extracts by boiling and the residues taken up in water. When it was found that the active material in these extracts was adsorbed on charcoal (Norit) and to some extent on Super Filtrol (Filtrol Corp.), attempts were made to obtain elution. Various concentrations of methyl and ethyl alcohols at various pH's were tested. The most successful eluting agent consisted of 50 per cent ethyl alcohol containing 10 per cent ammonium hydroxide. Both of these substances could be removed from the eluates by boiling. Dialysis of the extracts was carried out in cellophane against distilled water. During the process, which lasted for several days (changing the water outside the cellophane at intervals of 12 hours), a temperature of from 50° to 60° C. was maintained in order to prevent bacterial action. An electric light bulb was used to beat the box in which dialysis was carried out. The diffusate and the dialysate were boiled down or made up to the original volume. The active materials were tested for stability by adjusting the pH of portions ot" the extracts to values ranging from pH 3.0 to pH 10.0 and then heating in the autoclave at 15 Ibs. pressure for one hour. Extracts were also boiled for 24 hours FACTORS IN THE GROWTH OF TETRAHYMENA 111 in the presence of five per cent sulfuric acid in an attempt to remove tannins (Harrison and Roberts, 1939). The results gave an additional test of stability to acid. In all cases the pH of the extracts or of the fractions was adjusted to approxi- mate neutrality in order to avoid changing the pH of the medium or precipitation of the casein upon addition of the extract. RESULTS A. Supplementary F actor s^ None of the strains of Tetrahymena gave growth in the vitamin-free casein medium alone or with the addition of various known growth factors. Upon the addition of 0.08 per cent of a water extract of yeast (Harris) or a water extract of crude casein good growth was obtained. This demonstrated that the casein was available to the ciliates and that the failure to grow was due to a lack of some sub- stance, although the possibility that some toxic substance was neutralized should be kept in mind. It was then found that when the basic medium was made up with the inorganic salt solution of Barker and Taylor growth of the ciliates occurred in the third transplant without the addition of any supplement. This growth was, however, extremely slow, taking a week or ten days to reach a maximum density of about 1000 organisms per ml., which was far below that in controls in crude casein or vitamin-free casein supplemented with hay or alfalfa extract. Addition to the basic medium of thiamine, riboflavin, nicotinic acid or pyridoxine either alone or in com- bination made little or no difference either in the rate or the density of growth. The same may be said of pantothenic acid, p-aminobenzoic acid, uracil, pimelic acid, i-inositol, and biotin methyl ester. These results indicate that some unknown factor (or factors) is required for the maximal growth of Tetrahymena, since crude casein gave far better growth than the basic medium supplemented by any or all of the known compounds mentioned above. The fact that transplantable growth occurs in the unsupplemented vitamin-free casein indicates either that the ciliates are capable of a slow synthesis of all their supplementary requirements or that the casein still contains traces of the required factors. The latter explanation seems more probable in view of the difficulty of obtaining chemically pure proteins. Therefore, until a medium of chemically known composition or one composed en- tirely of synthetic compounds can be formulated the question of the absolute re- quirement for various growth promoting substances will have to remain open. It is clear, however, that for maximal growth the ciliates must be supplied with an outside source of unknown factors. These factors were found to be present in yeast, egg yolk, milk, timothy hay and alfalfa as well as meat (e.g. proteose-pep- tone). The animal sources were much lower in their content of growth promoting material than the plant sources. The former also contain a much larger propor- tion of protein material. For these reasons work was chiefly confined to the plant materials, especially since one of the aims was to obtain protein-free extracts of the growth promoting material in order to study the nitrogen requirements of the ciliates. Preliminary experiments with a water extract of yeast had indicated that treat- ment with lead acetate gave an active precipitate. The procedure when tested on 112 VIRGINIA C. DEWEY extracts of hay or alfalfa gave precipitates which were much reduced in activity while the filtrates were usually inactive. A recombination of the two fractions gave growth very nearly equal to that of the controls (Table I). The indications are that some of the preparations may he slightly toxic, but, more important, that there are at least two factors present in hay or alfalfa which are necessary for the TABLE I Supplements II A or H or I — factor I II— factor II A — untreated alfalfa extract H — untreated hay extract. maintenance of growth at a maximal rate. For convenience the substance present in the material precipitated by lead acetate will be referred to as factor I and the material present in the filtrate as factor II. It was also found that neither factor I nor factor II could be replaced by any one of the known growth supplements nor by a mixture of all the ten tested. This is further evidence for the existence of two substances of unknown structure neces- sary for maximum growth. Further purification or a better separation of the two factors was attempted un- successfully by precipitation with lead acetate from an alkaline solution. Reprecipi- tation of the fractions with lead acetate was also unsuccessful, since the products gave evidence of greatly increased toxicity, possibly due to the increased phosphate concentration. In all these preparations protein was found to be present in the filtrate fraction and since both fractions are required for growth, the method is not useful for the removal of protein. Ferric oxide hydrosol has been used as a protein precipitant. When tested on alfalfa and hay extracts it was found to behave similarly to lead acetate. There was a separation into two fractions, both required for optimum growth. The precipitate contained factor I and the filtrate contained factor II as well as protein (Table II). The results with hay extract and ferric oxide hydrosol are similar to those given above for alfalfa extract but the separation is not so clear cut. A reprecipitation of the iron hydrosol fractions with lead acetate gave a more complete separation, but there was evidence of an increased toxicity of the fractions. The material precipitated by sulfuric acid in preparation of the extracts for the addition of phosphotungstic acid was found to be inert whether alone or in the pre>enre of either factor I or factor II. Upon the addition of phosphotungstic acid there was no clear separation into two active fractions and no removal of protein without appreciable loss of activity. FACTORS IN THE GROWTH OF TETRAHYMENA TABLE II 113 Supplements I II Pel Fell A Pel Fel — iron hydrosol precipitate Fell — iron hydrosol filtrate Other symbols as in Table I. At this point the possibility that carbohydrate might be concerned in the ac- tivity of these fractions arose. Since all the reducing sugar in the extracts was found in the filtrate fraction from the lead acetate treatment, this fraction was treated with copper sulfate and calcium hydroxide. A complete removal of the re- ducing sugars but not of the protein in the preparations was possible without ap- preciable loss in activity. This treatment may be valuable in the further purifica- tion of the factors, since other inert materials appeared to be removed with the sugars. Since heavy metals failed to remove protein, extractions with various organic solvents was tested as a means of obtaining protein-free preparations. Extracts prepared with ether, acetone, alcohol and acetic acid were found to be inactive or even toxic. Dialysis also failed to remove protein or protein breakdown products. Some nitrogenous material of this nature was found to be freely diffusible as were both factor I and factor II. The dialysate in all cases was inert; all activity wras found in the diffusate. The fact that there was some loss of activity from the extracts during dialysis led to the conclusion that one of the factors is destroyed by light. When the electric light bulb used to heat the box in which dialysis was carried out was screened the loss of activity did not occur. This may be correlated with the progressive loss of activity of extracts exposed to ultraviolet radiation for increas- ingly longer intervals (Kidcler and Dewey, 1942, mistakenly state that factor I is affected by the irradiation). The results indicate that factor II is destroyed by light. This is evidence also for the organic nature of the growth promoting ma- terial. Adsorption upon activated charcoal or Fuller's earth followed by selective elution is a well known means of purification of growth factors. When this method was tested it was found that both factors, as well as protein, are readily adsorbed upon Norit and much less readily upon Super Filtrol. The filtrate after the Norit treat- ment was completely inert. Both factors (as well as the protein) appear to be eluted by alkaline alcohol (Table III). The elutions from Super Filtrol were more successful, possibly because the materials are less strongly adsorbed. Although this method may be useful in the purification of the separate fractions after precipitation with lead acetate, it was discarded as a means of protein removal. The tests of the stability of the growth substances to heat at various pH values showed that there was no loss of activity in alfalfa extracts in either acid or alkaline 114 VIRGINIA C. DEWEY TABLE III Supplements it 0 0 II F— filtrate after adsorption E — Eluate from Super Filtrol. solution. On the other hand proteose-peptone treated at an alkaline pH and used to supplement casein was almost inert. By testing it was found that this was due to a loss of factor I during the treatment (Table IV). Factor I from animal sources therefore appears to be heat-labile. The loss of activity in heat-treated proteose-peptone is not due to the destruction of thiamine. The more drastic treat- ment such as that described for the removal of tannins destroyed activity entirely, which is further evidence for the organic nature of the supplements. TABLE IV Supplements HA HPP II Thiamine HA — alfalfa extract heated at high pH HPP — proteose-peptone heated at high pH. The last method tested for the removal of protein was precipitation with organic solvents. Both the whole extracts and the fraction (filtrate) containing the pro- tein after lead acetate precipitation were treated by the addition of alcohol up to a concentration of 75 per cent. This method was successful in the removal of pro- tein from the hay extracts but not from the alfalfa extracts. The precipitates ob- tained from hay were inert and the- activity of the filtrates was unaffected (Table V). Whole hay extract treated in this manner was tised in the experiments on nitrogen nutrition to be described later. The effect of the addition of barium hydroxide plus alcohol was tested on the alfalfa extract in the hope of precipitating the protein. It was found, however, that O per cent alcohol alone precipitated some of both factors along with some of the protein. The addition of barium then had an effect similar to that of lead acetate in that there was a partial separation of the two factors. FACTORS IN THE GROWTH OF TETRAHYMENA 115 Acetone was next considered as a means of removing protein from alfalfa ex- tracts. Its behavior was similar to that of alcohol in that the active substances were precipitated along with the proteinaceous material, the amount increasing as the concentration of the acetone was increased. At 80 per cent factor I was largely precipitated and factor II to a smaller extent. Since protein-free extracts could be obtained readily from hay, the work on alfalfa was discontinued even though it is a richer source of growth-promoting material. TABLE V Supplements 0 I II HP iif H Hp Hf 0 0 + + ± ± ± + + + + ± + + + + I + + + + + * + + + + lip — precipitate from alcohol treatment of factor II fraction I If — nitrate from the same Hp — precipitate from alcohol treatment of hay extract Hf — nitrate from same. From the above the properties of the two factors may be summarized as follows : soluble in water, moderate concentrations of alcohol and in low concentrations of acetone; insoluble in ether; stable to heat (plant sources only in the case of factor I); clialyzable through cellophane; readily adsorbed on charcoal and less readily upon Super Filtrol ; eluted by ammoniated alcohol. Factor II differs from factor I in that the former is not precipitated by the salts of heavy metals and appears to be destroyed by irradiation. When either Harris gelatin or Eastman de-ashed gelatin was used as a basic medium (1.5 per cent solution) the results obtained were similar to those obtained with casein as a basic medium, except that the population density was smaller. In the Harris gelatin alone slight but transplantable growth, which was somewhat im- proved upon the addition of inorganic salts, was obtained. The addition of thia- mine, riboflavin, pantothenic acid or biotin gave little or no improvement in growth, while the addition of hay extract gave a considerable increase in the rate and den- sity of growth. With de-ashed gelatin the addition of inorganic salts was necessary and in their presence without the addition of supplements slight transplantable growth occurred. The addition of thiamine or ribroflavin (0.0001 mg. per ml.) or both together gave no improvement in growth. Growth was increased only upon the addition of both hay extract and riboflavin to the medium. B. Nitrogenous Nutrition The experiments to be described below are exploratory in nature and have served chiefly to suggest further experiments and modes of attack upon the prob- lem. Some of the work of earlier investigators was repeated in the hope that the use of an adequately supplemented medium might give better results than had been obtained. 116 VIRGINIA C. DEWEY The first experiments were carried out upon completely hydrolyzed casein. Such a medium was chosen in the hope of shedding more light upon LwofFs (1932) hypothesis that polypeptides arc required for growth. Acid digestion was used be- cause complete hydrolysis hy enzymatic means is difficult if not impossible and alka- line hydrolysis has a destructive effect on many of the amino acids. No growth oc- curred in the acid hydrolysate even in the presence of what was considered to be adequate supplementation. Attention was then turned to a solution of pure amino acids, also supplemented with protein-free hay extract. This solution was prepared using the ten amino acids found by Rose (1938) to be necessary for the nutrition of the mammal. The amino acids were present in the concentrations found in a one per cent solution of casein. Again no growth occurred in this solution or in various dilutions of it. Such solutions have an osmotic pressure lower than that of salt solutions readily tolerated by the organism. The explanation for the lack of growth must, therefore, l)e sought elsewhere. Three other possible explanations for the lack of growth are, a) that one (or more) amino acid required for the growth of the organism is lack- ing, b) one or more of the amino acids present is toxic or inhibitory, and c) that the organism requires nitrogen in the form of polypeptides. The possibility of the toxicity of the amino acids was considered first. These experiments were to be correlated with others using gelatin as a basic medium and supplemented with one or more of the amino acids known to be lacking from this protein. For this reason those particular amino acids were added to casein as well as to gelatin in the concentrations in which they are found in a one per cent solu- tion of casein. The results in the two media were strikingly different. With casein it was found that the addition of free amino acids had little or no effect on growth. In the case of gelatin (one per cent vitamin-free gelatin Harris) definite inhibition of growth was found in those cultures containing valine, tyrosine or isoleucine. Hydroxyglutamic acid was not then available. When tryptophane was added to the gelatin there was a large increase in the growth and media containing tryptophane in addition to valine or tyrosine gave better growth than similar media lacking tryptophane. It was found that decreasing the concentration of these amino acids to 0.0025 per cent improved the growth in all cases, although tyrosine, valine and isoleucine still showed inhibition of growth. In all cases the media contained pro- tein-free hay extract. These experiments were repeated in tube cultures three or four times, but in order to check the observations cultures were incubated in Kidder culture flasks and the growth followed by making counts at intervals of 12 hours. With casein plus 0.01 per cent tyrosine it was found that the population density at the end of the phase of logarithmic growth (48 hours) was 81,000 organisms per ml. and without tyrosine 75,500 per ml. The figures in the case of tryptophane were quite similar, 84,000 and 70,000 respectively, with and without 0.01 per cent tryptophane. The generation times did not differ significantly in any of the media. The differences in population density in these media represent less than one division per ciliate and are not regarded as being of statistical significance. With one prr cent gelatin as a basic medium it was found that the addition of 0.01 per cent tyrosine, valine or isoleucine gave maximum populations of only a few hundred organisms per ml. Gelatin alone gave 15,000 per ml. and with the addition of tryptophane a maximum of 90,000. When the concentration of added FACTORS IN THE GROWTH OF TETRAHYMENA 117 amino acid was reduced to 0.0025 per cent tyrosine gave a maximum of 12,000; valine, a maximum of 7,100; isoleucine, a maximum of 900 and tryptophane a maxi- mum of 91,000 organisms per ml. In all cases the generation time was lengthened. When two per cent gelatin was used a population of 80,000 organisms per ml. was obtained and when 0.002 per cent tryptophane was added the maximum was 230,000 organisms per ml. In this case the amino acid caused no decrease in inter- divisional time. The above results indicate that certain amino acids are detrimental to the growth of Tetrahymena, but suggest that this inhibition is reduced or absent in the presence of large protein molecules such as casein, or in the presence of tryptophane. The growth in two per cent gelatin with and without tryptophane leads plausibility to the theory that large protein molecules or a sufficient concentration of smaller pro- tein molecules in some way decreases the inhibitory effect of free amino acids upon the ciliates. Time did not permit the testing of the more toxic amino acids with the higher concentration of gelatin. Silk peptone, the only other incomplete protein preparation readily available, gave such good growth when supplemented with hay extract that it was not used as a basic medium for the study of amino acid requirements. DISCUSSION Of the four types of substances generally accepted as being required for growth of an organism (inorganic salts, supplementary substances, carbon and nitrogen compounds) it is evident that Tetrahymena geleii requires inorganic salts (Hall, 1942; Hall and Cosgrove, 1944 and data presented here), supplementary factors and an organic source of nitrogen which supplies the needs for both carbon and nitrogen. The requirement for a source of carbon separate from the source of nitrogen has never been demonstrated. At present the question of the supplementary factor requirements of Tetra- hymena remains unsettled. So far the claims that thiamine is a growth factor (i.e. an absolute requirement for growth) have not been substantiated. Indeed under certain conditions it is not even to be regarded as a growth stimulant (Kidder and Dewey, 1942). The work of Hall and Cosgrove (1944) fails to refute this claim. Hall (1942) claims that riboflavin is also a growth factor for Tetrahymena (Colpidimn cauip\luui) . This work could not be confirmed, although stimulation of growth could be obtained with both thiamine and riboflavin under certain condi- tions. In any case the growth stimulation obtained with the two unknown factors described above is far more powerful than that caused by either of these compounds. Elliott (1935b) reports an increase in the maximum population density of cul- tures when pantothenic acid was added to tryptone media. Since he was using a crude preparation of pantothenic acid this effect may have been due to other sub- stances in the preparations. Pantothenic acid has subsequently been found to have have no effect on growth when added to a casein medium. So far as can be de- termined from the data published (Hall, 1939; 1942) pimelic acid has no "accelera- tory" effect on growth. The effect appears to be due to the introduction of in- organic salts. Certain secondary effects have been attributed to thiamine, riboflavin and other known growth substances (Hall, 1940a; Hall and Shottenfeld, 1941; Baker and 118 VIRGINIA C. DEWEY Johnson, 1941). These are concerned with the death and decline phases of growth and are not df immediate interest here. It is of more importance to the prob- lem under consideration that none of the known growth-promoting substances will permit maintenance of growth at the maximum rate and of a maximum density. For sucli growth at least two substances of unknown nature are required. Whether or not some of the known compounds may also be required for such growth can- not be decided until pure preparations of these substances and a basic medium known to contain no growth supplements are available. The use of purified gelatin for a basic medium may give information of some value, but it is not truly suitable, since its use introduces the complication that it does not satisfy a possible require- ment for one or more of the amino acids it lacks. Although the so-called vitamin- free casein is not altogether ideal because it appears to contain traces of growth promoting materials, it is nevertheless an adequate basic medium for a study of growth stimulation. Unsupplemented, the growth it supports is far from maximal. The fact that growth of these ciliates can be obtained in gelatin solutions when properly supplemented, as pointed out by Hall (1942), would indicate that the ciliate requires for growtli none of the amino acids lacking from that protein. In other words T. yclcii must synthesize tryptophane, valine, hydroxyglutamic acid, isoleucine and possibly tyrosine unless its protoplasm does not contain these amino acids. This latter hypothesis seems most unlikely especially in view of the fact that tryptophane increases the growth so remarkably. It is difficult to explain however, why tryptophane increases the maximum concentration of organisms ob- tained rather than the growth rate. The failure of other investigators to obtain growth with solutions of amino acids or incomplete proteins supplemented with amino acids is now understandable. In some cases (Lwoff, 1932; Elliott, 1935b) the media contained none of the supple- mentary factors now known to have a profound effect on growth. Nor can the claims of Hall and Elliott (1935) regarding the effects of certain amino acids be regarded as conclusive, since their results were expressed as ,r .r,,. As Kidder (1941b) has pointed out, this method of representation may give an entirely false conception of the results obtained. Another source of possible error in the earlier work may lie in the use of concentrations of amino acids which may now lie re- garded as inhibitory to growth. It is possible that this difficulty may be overcome by the adsorption of amino acids upon inert colloids and by the use of tryptophane, which appears to decrease the toxicity of other free amino acids. In view of the inhibitory effect of free amino acids and of the ability of T. gclcii to grow in an incomplete protein such as gelatin, it is difficult to understand the report of Kline (1943) that T. i/clcii (Colpidiuin striatttiu) will grow in a solution of 15 amino acids with the addition of various supplements. It is possible that the explanation lies in the fact that different strains of T. c/eleii were used. Xo definite decision can as yet be made between the three suggested possibilities for the lack of growth of T. f/clcii in amino acid solutions. The evidence on hand, however, suggests that the factor of toxicity of free amino acids is of some im- portance. This effect, rather than a requirement for polypeptidcs, is a possible explanation for the decreasing growth obtained by Lwoff (1932) as the degree of hydrolysis of the medium used was increased. This would be true whether or not adequate supplements were present. FACTORS IN THE GROWTH OF TETRAHYMENA 119 SUMMARY 1. The known growth promoting substances alone or in various combinations are not sufficient for the growth of Tetrahymena gclcii at a maximal rate and density. 2. At least two unknown substances (factor I and factor II), present in both plant and animal materials, are required for such growth. 3. Factor I is distinguished from factor II by the fact that the former is pre- cipitated by heavy metal salts while the latter is not. 4. Active protein-free preparations of these factors may be prepared from ex- tracts of timothy hay by treatment with ethyl alcohol. 5. Growth of the ciliate could not be obtained -in acid digests of casein or in solutions of free amino acids supplemented with the protein-free extract. 6. Tyrosine, valine, and isoleucine were found to be inhibitory in the presence of gelatin, but not in the presence of casein. 7. A large increase in the population density occurred in the presence of trypto- phane and gelatin but not with tryptophane and casein. LITERATURE CITED BAKER, E. G. S., AND W. H. JOHNSON, 1941. Some effects of vitamin B components on popu- lations of Tetrahymena geleii. Anat. Rcc., 81 (suppl.) : 68. BARKER, H. A., AND C. V. TAYLOR, 1931. A study of the conditions of encystment of Colpoda cucullus. Physiol. Zool., 4 : 620-634. CLAFF, C. L., 1940. A migration dilution apparatus for the sterilization of protozoa. Physiol. Zool., 13: 334-341. DEWEY, V. C., 1941. The nutrition of Tetrahymena geleii (Protozoa, Ciliata). Proc. Soc. E.vpcr. Blol. and Med., 46 : 482-484. ELLIOTT, A. M., 1935a. Effects of certain organic acids and protein derivatives on the growth of Colpidium. Arch, f Protistcnk., 84: 472-494. ELLIOTT, A. M., 1935b. The influence of pantothenic acid on the growth of protozoa. Biol. Bull., 68: 82-92. ELLIOTT, A. M., 1939. The vitamin B complex and the growth of Colpidium striatum. Plivsiol. Zool., 12 : 363-373. HALL, R. P., 1939. Pimelic acid as a growth stimulant for Colpidium campylum. Arch. f. Protistcnk., 92: 315-319. HALL, R. P., 1940a. Relation of vitamins to the growth and decline of populations in Glau- coma piriformis. Anat. Rec., 78 (suppl.) : 164. HALL, R. P., 1940b. Riboflavin and thiamin as growth factors for the cilite, Colopidium cam- pylum. Anat. Rcc., 78 (suppl.) : 164. HALL, R. P., 1942. Incomplete proteins as nitrogen sources and their relation to vitamin re- quirements in Colpodium campylum. Physiol. Zool., 15: 95-107. HALL, R. P., AND W. B. COSGROVE, 1944. The question of the synthesis of thiamin by the ciliate, Glaucoma piriformis. Biol. Bull., 86: 31-40. HALL, R. P., AND A. M. ELLIOTT, 1935. Growth of Colpidium in relation to certain incomplete proteins and amino acids. Arch. f. Protistcnk., 85: 443-450. HALL, R. P., AND A. SHOTTENFELD, 1941. Maximal density and phases of death in populations of Glaucoma piriformis. Physiol. Zool., 14: 384-393. HARRISON, C. J., AND E. A. H. ROBERTS, 1939. The fermentation process in tea manufacture. IV. Tea tannin and its fermentation products. Bioclicin. Jour., 33 : 1408-1420. KIDDER, G. W., 1941a. Growth studies on ciliates. V. The acceleration and inhibition of ciliate growth in biologically conditioned medium. Physiol. Zool., 14 : 209-225. KIDDER, G. W., 1941b. Growth studies on ciliates. VII. Comparative growth characteristics of four species of sterile ciliates. Biol. Bull., 80 : 50-68. 120 VIRGINIA C. DEWEY KIDDER, . The result- ing hydrolysate was biuret negative. This hydrolysate was used in a 0.5 per cent concentration (calculated from the original amount of casein used). 4. Gelatin — This was used in a two per cent solution. 5. Heated gelatin — Four per cent gelatin was autoclaved one hour at pH 10. After cooling the pH was adjusted to 6.8 and the solution was diluted to a concen- tration of two per cent. 6. Gelatin hydrolysate — Four per cent gelatin was refluxed for five hours in a 24 per cent solution of HL,SO4. The sulphate was removed by Ba(OH)2. This hydrolysate was biuret negative and was used in one per cent concentration (calcu- lated from the original amount of gelatin used). Alfalfa extract was prepared as described previously (Kidder and Dewey, 1942). After heat and alkali treatment at pH 10 it was adjusted to pH 6.8 and added in a dilution of 1 : 10 final concentration. This dethiaminized extract is designated A. Thiamine hydrochloride was added where indicated in the concentration of one microgram per ml. of medium. To all of the media used in the following experiments were added just before sterilization the following inorganic salts (Hall and Cosgrove, 1944) : 0.02 per cent MgSO4-7HoO; 0.02 per cent K.,HPO4 ; 0.01 per cent^CaCl.,-2H.,O ; 0.00025 per cent FeCL-6H,,O ; 0.00001 per cent MnCL-4H,O ; 0.00001 per cent ZnCL. To all media was also added 0.1 microgram per ml. of riboflavin. Tryptophane was added to the hydrolysed casein (to compensate for loss in hydrolysis) and to all gelatin and gelatin hydrolysates to a concentration of 0.0025 per cent. Experiments with ami no acid mixtures now being conducted show that tryptophane is essential to the growth of Tetrahymena. All media were used at pH 6.8-6.9. A number of preliminary experiments were carried out with each medium in tubes in serial transplants. Each tube contained five ml. of medium. All tube series were inoculated with a bacteriological loop delivering approximately 0.008 ml. Tube series were grown through at least three transplants before any conclusions THIAMINE AND TETRAHYMENA 123 were drawn, this to eliminate the possibility of carry over of medium from the stock cultures. Tube cultures were incubated at room temperature and transplants were made every 48 hours, except where very slow growth occurred in the early trans- plants, where longer times were allowed. The quantitative studies were made using the culture flasks described earlier (Kidder, 1941b). These flasks contained 100 ml. of media. Inoculations were made from third transplant tubes of like media so that the flask cultures represent fourth transplant series. Sterile serological pipettes were used for the inoculations and from 0.1 ml. to 0.5 ml. was added, depending upon the density of the popula- tion in the tube from which the inoculation was made. After the first few experi- ments inoculations were made from cultures within the exponential growth phase and the inoculations were calculated to give an initial count of as near 100 cells per ml. as possible. Flask cultures were incubated at 24.5° C. All flask experiments were repeated at least once. Our method of counting cells from culture has been described elsewhere (Kid- der, 1941b), but it should be noted here that this method gives only viable counts, hence our population counts tend to be lower in the stationary phase and phase of decline than where methods involving the counting of killed cells is employed. These differences are well illustrated in the work of Johnson and Baker (1943). Generation time (g) was calculated by the use of the formula „ = - Mog2 log b — log a where t -- the time in hours during which the population has been increasing ex- ponentially, a -- the number of cells per unit volume at the beginning, and b — the number of cells at the end of time, t. EXPERIMENTAL Population Studies Casein and casein hydrolysate — When a solution of casein is adjusted to pH 10 and autoclaved for one hour to render it thiamine-free, a number of changes take place which make it very inferior to unheated casein as a basic medium for Tetra- hymena. Hall and Cosgrove (1944) state that factors in addition to thiamine must have been destroyed, because even upon the addition of thiamine poor growth re- sulted. With our strain of T. gclcii heat treated casein plus thiamine (also salts and riboflavin, as mentioned above) inhibited growth even in the first transplant, and second transplants were almost invariably negative. In no case was growth obtained in the third transplant. However, if the insoluble precipitate resulting from such treatment is filtered off and the concentration (originally one per cent before filtration) is reduced by one-half then low but transplantable growth results. The addition of thiamine has no significant effect upon the generation time, length of the logarithmic phase, maximum yield or survival up to the limit of our experi- ment (Table I ; Fig. 1). This indicates that the heat treatment has produced toxic substances which, when reduced in concentration do not inhibit growth entirely. It also shows that Tetrahymena can reproduce without an exogenous source of thiamine. 124 G. W. KIDDER AND V. C. DEWEY TABLE I Medium Generation time in hours Population per ml. at end of log. phase Maximum yield cells/ml. Population per ml. at 1 1 days Heated casein 0.5 per cent + A 5.02 18,000 160,000 61,000 Heated casein 0.5 per cent + A + B! 4.80 17,500 ' 182,000 82,000 Filtered heated casein 8.35 1,600 5,500 2,000 Filtered heated casein + B, 8.96 1,800 8,000 3,200 Filtered heated casein +A 4.43 42,000 100,000 57,000 Filtered heated casein + A + BI 4.27 38,000 110,000 86,000 A = heat and alkali treated alfalfa extract; B, = thiamine 1 microgram/ml. All media contains salts and riboflavin (0.1 micrograms/ml.). 2 4 tr uj 0. t/l UJ . // // »/ w ? 1 C e s> i i HEATED CASEIN FILTRATE, ii ii ii it ii ii ii ii ii 1 1 i + Bi + A + A + Bi 1 l 1 1 1 0 24 48 72 96 120 144 HO U RS 168 192 216 240 264 FIGURE 1 When dethiaminized alfalfa extract is added to the heated casein or to the filtered heated casein the response is striking. Rapid growth now occurs in the heated casein while in the filtered heated casein the generation time is reduced by nearly one-half and the population at the end of the logarithmic growth phase is in- 11 < ased from around 1,600 to over 40.000 per nil. The maximum yield is increased from ahout 5,000 to approximately 100,000 per nil. and a much higher population is maintained for at least 11 days (over 50,000 as compared to 2,000 per ml.) (Table THIAMINE AND TETRAHYMENA 125 I ; Fig. 1 ) . This would seem to indicate that, in addition to supplying stimulatory factors (Dewey, 1944) and the synthesizing factor, the alfalfa extract counteracts the toxic effects of the heat treatment on the casein. There is indication in the shape of the growth curve that the greater toxicity of the heated casein has not heen as successfully counteracted as that of the filtered heated casein. The generation time is approximately 0.5 hour longer in the former and the population begins to fall off sooner. The maximum yield, however, is higher (160,000 as compared to 100,000 per ml.) in the unfiltered casein. This last may be due to the higher con- centration of available protein. TABLE II Medium Generation time in hours Population per ml. at end of log. phase Maximum yield cells/ml. Population per ml. at 11 days Casein hydrolysate 0.5 per cent + A 4.27 50,000 122,000 49,000 Casein hydrolysate 0.5 per cent + A + Bt 4.36 31,000 191,000 94,000 Heated casein hydrol. 0.5 per cent + A 4.37 29,000 182,000 54,000 Heated casein hydrol. 0.5 per cent + A + B! 4.46 32,000 171,000 73,000 A = heat and alkali treated alfalfa extract; Bi = thiamine 1 microgram/ml. All media contains salts, riboflavin (0.1 microgram/ml.) and 1-tryptophane (0.0025 per cent). 24 (T Ul o. tn LJ O o o t o O CASEIN HYDROLYSATE + A • " " + A + Bi CHEATED CASEIN HYDROLYSATE + A O " " " + A+ Bi -1 1 I 1 1 1 1 1 1 1 1 0 24 48 72 96 120 144 168 HOURS 192 216 240 264 FIGURE 2 126 (,. \Y. KIDDER AND V. C. DEWEY The addition of thiainine to either the heated casein or filtered heated casein plus dethiaminized alfalfa extract has very little effect. The cultures maintain a slightly higher level at 11 days duration but the shape of the growth curves are almost identical. "When a 0.5 per cent solution of a biuret negative casein hydrolysate plus 0.0025 per cent 1-tryptophane was used as a basic medium it was found that growth was impossible beyond the first transplant, even when thiamine wras added. Inasmuch as unheated casein and filtered heated casein give slow but indefinitely transplantable growth the acid hydrolysis must have destroyed some factor or factors, other than thiamine, necessary for growth. Excellent growth resulted, however, when de- TABLE III Medium Generation time in hours Population per ml. at end of log. phase Maximum yield cells/ml. Population per ml. at 11 days Size of cells at 1 1 days (av. 20 measurements) Gelatin 2 per cent 5.68 2,600 12,000 2,100 22M X 16.5M Gelatin 2 per cent + B! 5.59 2,400 67,000 31,000 91. SM X 34M Gelatin 2 per cent + A 3.21 18,000 140,500 47,000 51ju X 24M Gelatin 2 per cent + A + B! 3.22 17,200 161,000 52,000 86.5ju X 22M A = heat and alkali treated alfalfa extract; BI = thiamine 1 microgram/ml. All media contains salts, riboflavin (0.1 microgram/ml.) and 1-tryptophane (0.0025 per cent). 48 72 96 120 144 HOURS 168 192 216 240 264 FIGURE 3 THIAMINE AND TETRAHYMENA 127 TABLE IV Medium Generation time in hours Population per ml. at end of log. phase Maximum yield cells/ml. Population per ml. at 1 1 days Size of cells at 11 days (av. 20 measurements) Heated gelatin 2 per cent 5.31 2,750 11,700 200 20.u X 16M Heated gelatin 2 per cent + B! 5.72 2,000 68,500 44,000 85M X 30/z Heated gelatin 2 per cent + A 3.08 19,500 82,000 40,000 47M X 20M Heated gelatin 2 per cent + A + Bj 3.42 17,500 96,000 72,000 89,u X 36.5ji A = heat and alkali treated alfalfa extract; BI = thiamine 1 microgram/ml. All media contains salts, riboflavin (0.1 microgram/ml.) and 1-tryptophane (0.0025 per cent). K U Q. irijormis in peptone solution was at first reduced as much as 80 per cent by KCN, but later returned to normal or nearly so. Peters (1929) found that M/500 KCN did not inhibit the respiration of Colpidmm colpoda. Pitts (1932) found that the respiration of Colpidium campylnni was slightly reduced by cyanide but that this effect was only temporary. Hall (1941), using an improved technique to avoid the loss of cyanide from the test solution, definitely confirmed the fact that Colpidium campylum is sensitive to cyanide. So far as is known, there are no published investigations dealing with the mechanism of respiration in amoeboid organisms. Pclomy.va carolinensis Wilson (Chaos chaos Schaeffer), a multinucleate rhizopod, is favorable for physiological studies because it is relatively large and can easily be grown in the laboratory. MATERIAL AND METHODS The specimens of Pclomyxa carolincnsis used in these experiments were of the same strain as those used by Belda (1942) and Pace and Belda (1944). They were grown in Hahnert (1932) solution and were fed by adding paramecia to the cultures. Prior to each experiment, however, the pelomyxae were kept for about a week in a culture solution buffered to maintain a hydrogen-ion concentration of pH 6.8 (Pace and Belda, 1944, Table I). Portions of a centrifuged culture of Paramcciiun caiidatitin were added every second or third day. The pelomyxae grew well under these conditions and usually contained numerous food vacuoles. The rate of oxygen consumption was measured by means of a Barcroft-Warburg apparatus. Preliminary tests (Pace and Belda, 1944) had shown that there was no measurable difference in the rate of oxygen consumption between pelomyxae tested in flasks which contained 100, 200, or 300 organisms. In the present series "i experiments usually 200 specimens were put into each flask; in a few cases 150 or 300 specimens were used. 1 With the support of a tyrant from the American Philosophical Society. 138 RESPIRATION IN PELOMYXA 139 A typical experiment was carried out in the following manner: a 0.4 ml. por- tion of 10 per cent KOH was put into the inset and a 0.3 ml. portion of 3 N HC1 into the onset of 3 of the flasks. These 3 flasks were used as controls. Pelomyxae of uniform size were removed from the buffered culture medium with a capillary pipette under a binocular dissecting microscope and washed in 3 separate portions of fresh sterile culture medium. A 5 ml. portion of sterile culture medium con- taining the proper number of pelomyxae was then put into each of the 3 Warburg flasks. A 0.4 ml. portion of a KOH-KCN absorption solution - was put into the inset of the 3 remaining flasks and a 0.3 ml. portion of 3 N HC1 was put into the onset. Pelomyxae were removed from the buffered culture medium and washed in 3 separate portions of fresh culture medium plus either potassium cyanide, potassium arsenite, or ethyl urethane. The Barcroft-Warburg apparatus included a total of 7 manometers and flasks. Of these, 6 were prepared as above. A 5 ml. portion of sterile culture solution without pelomyxae was put into the remaining flask which was used as a thermo- barometer. The water bath of the apparatus was kept at 25° ±0.05° C. The shaking mechanism was operated at the rate of 124 complete cycles per minute through an amplitude of 3 cm. After the manometers and flasks had been put into place with the stopcocks open, the shaking mechanism was run for one hour in order to equalize the temperature of the flasks with that of the water bath. All stopcocks were then closed, and manometer readings were recorded at intervals of one hour. RESULTS /. The effect of potassium cyanide on respiration. In order to ascertain the possible effects of cyanide on the structure and activity of Pelomyxa, several dozen specimens were put into Columbia dishes containing buffered culture solution plus different concentrations of KCN. The specimens were observed carefully under the microscope and compared with other specimens kept in culture solution without KCN. Practically all the food vacuoles disappear in pelomyxae kept for twelve hours or longer in a solution containing 10~2 M KCN. In addition there is a reduction in number or size of both the bipyramidal crystals and the cytoplasmic granules, so that the organisms now appear highly transparent. A number of large vacuoles containing clear fluid are produced in the cytoplasm. Large masses of gelated cytoplasm are found occasionally, both in the interior of the organisms and near the tips of the pseudopodia. Only intermittent movement of the plasmasol can be seen. The hyaline layer appears well-defined, and is much thicker than in normal 2 The KCN and KOH concentrations of the absorption solutions suggested by Krebs (1935) vary with the KCN concentration of the experimental culture fluid as shown : Molar concentration of Absorbing solution in KCN in culture solution inner cup (inset) 10-2 10 ml. 2N KCN + 0.2 ml. N KOH lO-3 10 ml. N KCN+ 1.0 ml. N KOH lO-4 5 ml. N KCN + 5.0 ml. N KOH 10-5 1 ml. N KCN + 10.0 ml. N KOH 140 I). M. PACE AND \V. H. BELDA specimens. 'I'he surface of the pclomyxae is covered with small protuberances. After 24 hours in the solution, the pelomyxae have long, thread-like pseudo- podia. Additional clear vacuoles make their appearance and movement of the cytoplasm practically ceases. The plasmagel layer appears to be very thin and it is difficult to handle the organisms without breaking the outer protoplasmic layers. If the outer layer is ruptured no new membrane is formed in the region of rupture, and the cytoplasm flows out into the surrounding culture medium. In lower con- centrations of KCN, namely. 1O3, 10~4, and 10~5 M, similar effects occur, but in progressively less degree. When KCN was added to the buffered culture solution in the higher concen- trations used (10 - and 10'" M), the hydrogen-ion concentration was reduced. IIC1 was added to restore the hydrogen-ion concentration to the value of pH 6.8. TABLE I The effect of potassium cyanide on oxygen consumption in Pelomyxa carolinensis. Tempera- ture 25° C.; hydrogen-ion concentration, pH 6.8. In most of the tests, 200 pelomyxae were used in each flask; in a few tests, 150 and 300 were used. Average volume of one million pelomyxae, 32,000 cubic millimeters. Molar concentration of KCN Number of tests Duration of tests Average Oj consumption in mm.s per hour per million organisms Average O: consumption in mm.3 per hour per mm.3 cell substance Per cent inhibition 0 (Control) io-6 8 8 3 to 6 hours 9045 ±595 3132±387 0.282±0.018 0.098±0.012 65.4 0 (Control) lO-4 9 9 3 to 5 hours 8962 ±641 3220±302 0.280±0.020 0.100±0.009 63.1 0 (Control) IO-3 7 7 4 to 5 hours 8718±548 2840±345 0.272±0.017 0.089±0.010 67.5 0 (Control) IO-2 8 8 3 to 4 hours 9478±567 2930±248 0.296±0.018 0.092 ±0.008 69.1 The results of the experiments with cyanide are shown in Table I. This table shows that there is a reduction of 63 to 69 per cent in the rate of oxygen consump- tion in Pelomyxa when KCN is present in the culture solution in concentrations of 10~r', 10~4, 10~:f, and 10"- M. The highest concentration of KCN. however, pro- duced only slightly greater inhibition than the lowest. These results indicate that cellular oxidation in Pelomyxa is regulated principally, but not entirely, by the cytochrome-cytochrome oxidase system. //. The cl) eel oj potassium arsenate. There is some evidence in the results obtained by Szent-Gyorgyi and Banga ( 1933), Korr ( I'tf5). atid Cohen and Gerard (1937), that arsenite may inhibit the activity not only of glutathione but also of dehydrogenases. If this be true, the re- sults with IVlomyxa indicate that arsenite inhibits the action of dehydrogenases much le>s than does urethane. RESPIRATION IN PELOMYXA 143 The degree of inhibition of respiration by ethyl urethane in Pelomyxa is ap- proximately equal to that in Glaucoma (Lwoff, 1934), but in approximately equal concentrations of arsenite, respiration in Pelomyxa is inhibited initially by 29 per cent, compared to 75-80 per cent in Glaucoma. If, as supposed, arsenites inhibit the activity of glutathione but not that of respiratory enzymes, it appears that glutathione is much less important in the respiration of Pelomyxa than of Glaucoma. These results indicate that the respiratory mechanism of Pelomyxa carolincnsis differs considerably from that of some of the Ciliata. Whether or not the mecha- nism of respiration of Pelomyxa carolincnsis resembles that of other free-living Rhizopoda must await further investigation. SUMMARY 1. In 10~5 M KCN respiration in Pelomyxa carolincnsis is inhibited by 63 per cent. In much higher concentrations of KCN, up to 10~2 M, only slightly greater inhibition occurs. 2. Pelomyxae which have been exposed to potassium cyanide (10~s to 10~2 M), for 12 to 24 hours, show many changes in protoplasmic structure. 3. In 5 X 10~3 M potassium arsenite the maximum inhibition of respiration in Pelomyxa carolincnsis is 35 per cent; this effect, however, is only temporary. 4. In 0.17 M (1.5 per cent) ethyl urethane the respiration of Pelomyxa caro- linensis is inhibited by 65 per cent. 5. Respiration in Pelomyxa carolincnsis appears to occur chiefly through a cytochrome-cytochrome oxidase system, and partly through a mechanism involving glutathione. 6. The respiratory mechanism of Pelomyxa carolincnsis, a rhizopod, differs con- siderably from that of a number of the ciliates. LITERATURE CITED BELDA, W. H., 1942. Permeability to water in Pelomyxa carolinensis. I. Changes in volume of Pelomyxa carolinensis in solutions of different osmotic concentration. The Sale- siamiin, 37 : 68-81. COHEN, R. A., AND R. W. GERARD, 1937. Hyperthyroidism and brain oxidations. /. Cell, and Comp. Physiol, 10 : 223-240. GERARD, R. W., AND L. H. HYMAN, 1931. The cyanide sensitivity of Paramecium. Amcr. J. Physiol., 97 : 524-525. HAHNERT, W. F., 1932. Studies on the chemical needs of Amoeba proteus : a culture method. Biol. Bull., 62: 205-211. HALL, R. H., 1941. The effect of cyanide on oxygen consumption of Colpidium campylum. Physiol. Zool, 14: 193-208. KORR, I. M., 1935. An electrometric study of the reducing intensity of luminous bacteria in the presence of agents affecting oxidations. /. Cell, and Comp. Physiol., 6 : 181-216. KREBS, H. A., 1935. Metabolism of amino-acids. III. Deamination of amino acids. Biochcm. J., 29: 1620-1644. LUND, E. J., 1918. Rate of oxidation in P. caudatum and its independence of the toxic action of KCN. Amer. J. Physiol., 45 : 365-373. LWOFF, M., 1934. Sur la respiration du Cilie Glaucoma piriformis. C. R. soc. Biol. Paris, 115: 237-241. PACE, D. M., AND W. H. BELDA, 1944. The effect of food content and temperature on respira- tion in Pelomyxa carolinensis Wilson. Biol. Bull., 86: 146-153. 144 D. M. PACK AND W. H. BELDA PETERS, R. A., 1929. Observations on the oxygen consumption of Colpidium colpoda. /. Physiol.. 68 : ii-iii. PITTS, R. F., 1('32. KlTirt of cyanide on respiration of the protozoan, Colpidium campylum. Proc. Soc. K.rf. li'wl. N. )'., 29: 542. SAITO, T., AND H. TAMIVA, 1937. Uber die Atmungsfarbstoffe von Paramecium. Cytologia. Fujii Jubik-c Volunic, i>p. 1133-1138. SHOUP, C. S., AND J. T. BOVKIN, 1931. The sensitivity of Paramecium to cyanide and effects of iron on respiration. /. Gen. Physiol., 15: 107-118. SzENT-GvORGYI, A.. AND I. BANGA, 1933. Uber das Co-ferment der Milchsauer-oxydation. Z. Phys. Chem., 217: 39-49. THE EXTERNAL MORPHOLOGY OF THE THIRD AND FOURTH ZOEAL STAGES OF THE BLUE CRAB, CALLINECTES SAPIDUS RATHBUN 1 SEWELL H. HOPKINS Agricultural and Mechanical College of Texas and Virginia Fisheries Laboratory For the past two years, workers at the Virginia Fisheries Laboratory, Williams- burg, have been attempting to rear larvae of the commercially important blue crab from the egg through all zoeal stages. In 1941 Dr. Margaret S. Lochhead worked out a successful method of hatching the eggs (Lochhead, Lochhead and Newcombe, 1942) and reared the larvae to the "second zoea" stage. During the summers of 1942 and 1943 this work was continued by Mrs. Mildred Sandoz and Miss Rosalie Rogers, who succeeded in rearing a number of individuals to the "third zoea" stage. The anatomy of the first and second zoeal stages was described in detail by Hopkins (1943). Churchill (1942) described five zoeal stages found in plankton tows at the mouth of Chesapeake Bay. Churchill's first and second zoeae seem to be identi- cal with those reared from blue crab eggs at the Virginia Fisheries Laboratory, but his third zoea is markedly different from the third zoea reared at this laboratory, as reported by Sandoz and Hopkins (1944). It is now realized, by the agencies concerned with regulation of the crab fishery in Chesapeake Bay, that a more detailed knowledge of the biology of the crab is necessary. Studies of the numbers and seasonal and geographic distribution of larvae in all stages are important means of locating the breeding grounds and de- termining the length of the larval period, the migrations of the larvae, the percent- age of survival under natural conditions, etc. Obviously these studies will be worthless unless the blue crab larvae are correctly identified, and distinguished from the other species of the same family (Portunidae) found in this region. The fol- lowing paragraphs attempt to give an accurate and detailed description of all fea- tures which may be of importance in separating larvae of different species. A number of individuals in the "fourth zoea" stage have been found in plankton tows made by the Virginia Fisheries Laboratory in the mouth of the Bay near Cape Henry. These are very similar to the blue crab "third zoea" raised in the labora- tory and found in plankton, but quite different from Churchill's "fourth zoea" (which was also found in our tows). The first and second zoeal stages have been restudied, and a few minor correc- tions of my 1943 description seem necessary. In the first zoea, there are six setae on the endopodite of the first maxilla, two in one group and four in the other. In the second zoea, there are normally three apical setae on the scaphognathite of the second maxilla, although only two can be seen in some specimens. The shorter seta on the fourth segment of the endopodite of the first maxilliped is usually longer than indicated in my 1943 report. 1 Joint contribution from the Virginia Fisheries Laboratory of the College of William and Mary and Commission of Fisheries (Number 20) : and from the A. and M. College of Texas. 145 146 SKWFLI, H. HOPKINS THE THIRD ZOEA Three specimens reared in the laboratory and six found in plankton tows (Cape Henry, August 14. 1('41 ) were dissected and mounted in glycerine. In addition, a number of specimens from plankton and one specimen reared in the laborator were mounted and studied entire, and still others were studied in formalin without mount- ing. Even after the most detailed study, no difference could be found between laboratory-reared and plankton specimens, except that one laboratory-reared speci- men seemed to have moulted precociously (the telson lacked the fourth or inner pair of setae, the second maxilla was of the second zoea type, and the size was be- low normal, although the maxillipeds each bore eight swimming hairs). The total length of the body is difficult to measure accurately because of the bent position of most specimens, but varies between 1.40 and 1.65 mm., measured from front of carapace between eyes to tips of telson. The carapace is exactly as in the second zoea except for larger size and the presence of a single (occasionally two) seta with setules on the posterior edge of the carapace. The dorsal spine is 0.40 to 0.50 mm. long, the lateral spines are 0.09 to 0.10 mm., and the rostrum is 0.33 to 0.36 mm. (measured from lower edges of bases of eyestalks). The eyes have very short stalks and are 0.21 to 0.26 mm. in diameter (Figs. 1 and 2). The antennule is unchanged from its form in the second zoea except that the setae or aesthetes are less uniform in width; the largest aesthete is about twice as wide as the second, the second is almost twice as wide as the third, and the third is about twice as wide as the fourth ; there is also a very short bristle or seta which usually cannot be seen. The peduncle of the antennule is 0.12 to 0.15 mm. long and the length of the longest aesthete is about 0.20 mm. (Fig. 3). The antenna is 0.30 to 0.35 mm. long. The spinous process bears 12 to 18 hooklike spines on each side. The exopodite is about 0.01 mm. long and bears two setae of unequal length, the longer 0.03 to 0.04 mm. long. There is a very slight ridge or bulge on the antenna near the exopodite which represents the first rudi- ment of the endopodite. The labrum bears a distinct chromatophore. The mandible is 0.12 to 0.19 mm. long and 0.08 to 0.11 mm. wide, and bears a very large chromatophore usually di- vided into three parts. The maxillule or first maxilla is 0.16 to 0.22 mm. long, from base to end of endopodite. The outer edge of the basipodite bears a seta about 0.04 mm. long. The distal (basal) endite of the protopodite bears eight setae and the proximal (coxal) endite bears six or seven setae. The endopodite is two-segmented; the PLATE I The third zoeal stage of Callincctcs sapidtts. All figures are camera lucida drawings Scale line A represents 0.5 mm. in Figures 1 and 2 ; scale line B represents 0.2 mm. in Figures 3-7. FIGURE 1. Third zoea reared from egg in laboratory, slightly flattened under cover glass. Fi(;rKK 2. Third zoea from plankton tow, Cape Henry. Fn.rKK 3. Antennule, third zoea from plankton. !,•]•: 4. First maxilla, third zoea reared in laboratory. i. 5. Second maxilla, third zoe.i from plankton. I'i'.ri'i 6. Kndopodite of first maxilliped, third zoea reared in laboratory. 1'H, i RI. 7. Endopodite of second maxilliped, third zoea reared in laboratory. THIRD AND FOURTH ZOEAE OF BLUE CRAB 147 PLATE I 148 SKWKLL H. HOPKINS distal segment is bifurcated and bears two groups of setae, four in one group and two in the other. The setae on the basal and coxal endites are 0.03 to 0.06 mm. long and those on the endopodite are 0.04 to 0.08 mm. long ; all setae on the first maxilla have setules (Fig. 4). The second maxilla seems t»> be subject to considerable variation in size and number of setae. The length of the second maxilla varies from 0.12 to 0.17 mm., measured from the base to the tip of the endopodite. The scaphognathite measures from 0.13 to 0.16 mm. through its longest dimension, and bears seven or eight setae along the distal part of the outer margin plus three to five apical setae around the proximal tip, making a total of ten to twelve setae. The endopodite has a single segment with bilobed tip bearing six setae in two groups, two in one group and four in the other. The basal and coxal endites are both bilobed; the distal (basal) endite bears nine setae and the proximal (coxal) one bears six or seven setae. All setae on the second maxilla bear setules and there are also setules around the mar- gins of the endites ( Fig. 5 ) . The basipodite of the first maxilliped is 0.20 to 0.27 mm. long; its posterior side bears ten setae with setules, and there is a very distinct chromatophore near its distal end. The endopodite is 0.18 to 0.23 mm. long, and has five segments; the first (proximal) segment bears two setae, the next bears two, the third has none, the fourth bears two, and the fifth bears five setae, four terminal and one short lateral seta (Fig. 6). The exopodite is divided into two segments; its length is 0.16 to 0.18 mm.; the eight terminal setae or "swimming hairs" are of unequal lengths, varying from 0.13 to 0.23 mm. The basipodite of the second maxilliped is 0.20 to 0.24 mm. long and bears four setae with setules near its posterior margin ; there is no chromatophore in the ba- sipodite, but there is a rather small chromatophore in the coxopodite. The en- dopodite is 0.06 to 0.09 mm. long and consists of three segments ; the first (proxi- mal) and second segment each bear one seta and the third bears five unequal terminal setae (Fig. 7). The exopodite is 0.18 to 0.23 mm. long, is divided into two segments, and bears eight terminal setae or "swimming hairs" of unequal lengths, 0.15 to 0.27 mm. The abdomen is 1.05 to 1.20 mm. long and has six segments, not counting the telson. The lateral spines or hooks on the second and third segments are exactly as in the second zoea. The third, fourth, and fifth segments have lateral spines projecting posteriad from the posterior margin of each segment. The second, third, fourth, and fifth segments each have a pair of dorsal setae projecting from the posterior margin, but there is no sign of a dorsal spine on any segment. The first PLATE II The fourth zocal stage of Call'mcctcs sapidus. All figures are camera lucida drawings. Scale line A represents 0.5 mm. in Figure 8; scale line B represents 0.2 mm. in Figures 9-14. FIGURE N. Fourth zoea from plankton tow, Cape Henry. FIGURE 9. Antennule. Fn.rki-; 10. Antenna. RE 11. First maxilla. Fn.rkK 12. Second maxilla. :•}•: 13. Kndnpodite of first maxilliped. 14. Endopodite of second maxilliped. THIRD AND FOURTH ZOEAE OF BLUE CRAB 149 PLATE II 150 SI.WKLL H. HOPKINS abdominal segment bears a large median chromatophore ; each of the other segments bears a chromatophore lying across the extreme anterior end near the ventral sur- face. F.xcept for slightly larger size, the telson is unchanged from its condition in the second zoea; that is, the dorsal side of each furcus bears a large spine directed dorsolaterally, and a smaller spine, farther back, directed dorsally, and the inner side1 of each furcus bears close to its base a small spine without setules in addition to the original three setae with setules (Fig. 1). The chromatophores probably cannot be described accurately without using living specimens, but the following chromatophores have been seen in the best pre- served specimens: (1) one chromatophore in the front of the head, between the eyes; (2) one chromatophore dorsal to the anterior part of the alimentary canal; (3) a pair of large chromatophores dorsal to the gut in the posterior part of the cephalothorax ; (4) a pair of chromatophores just ventral to the anterior part of the alimentary canal; (5) a small chromatophore, not always visible, below the base of the dorsal carapacial spine; (6) a large chromatophore in the center of the first abdominal segment; (7) chromatophores in the ventral anterior margin of the third, fourth, fifth, and sixth abdominal segments; (8) a large chromatophore in the labrum ; (9) a large chromatophore in each mandible; (10) a chromatophore in the distal end of the basipodite of each first maxilliped ; (11) a chromatophore in the coxopodite of each second maxilliped. THE FOURTH ZOEA Several specimens which are tentatively identified as the "fourth zoea" of the blue crab were found in surface plankton tows near Cape Henry, August 14, 1941. Five of these specimens have been studied in formalin solution, and two have been dissected for more detailed study of appendages. These specimens are identical with the third zoea of the blue crab except for larger size, a better developed en- dopodite bud on the antenna, additional setae on the maxillae and maxillipeds, and more setae on the posterior edge of the carapace. Identification of this zoea should, of course, be checked by comparison with laboratory-reared fourth zoeae when these can be obtained, but, in the meantime, I have little or no doubt that this is actually the fourth zoea of the blue crab. The length of the body, from midway between eyes to tips of the telson, is 1.75 to 1.95 mm. The dorsal spine on the carapace is 0.54 to 0.60 mm. long and the lateral spines are about 0.12 mm. long. The length of the rostrum is from 0.45 to 0.50 mm. The eyes are about 0.35 mm. in diameter. The posterior edge of the carapace on each side bears three large setae with setules (Fig. 8). The antennule is almost exactly as in the third zoea; the peduncle is 0.16 mm. long and the longest aesthete is 0.20 mm. (Fig. 9). The antenna is from 0.35 to 0.42 mm. long and is unchanged from the third zoeal stage except that the bud of the emlopodite is now very distinct (Fig. 10) ; it is noteworthy that the exopodite is still of exactly the same form as in the first zoea. The labrum and mandible are as in the third zoea except for slightly larger size. The maxillule or first maxilla differs from the third zoeal stage only in slightly larger size and in having nine or ten setae on the distal ( basal J endite where the third zoea had eight ( Fig. 11). The scaphognathite of the second maxilla hears about sixteen setae, with only a slight gap between the setae of the outer margin and the apical setae, which are now similar THIRD AND FOURTH ZOEAE OF BLUE CRAB 151 in form. The basal and coxal enclites of the second maxilla are as in the third zoea except that the distal (basal) endite bears nine or ten setae and the proximal (coxal) one has seven (Fig. 12). The first maxilliped bears eight setae or "swimming hairs" on the exopodite. The endopodite is five-segmented ; the first segment bears two setae, the second bears two, the third bears one (lacking in some specimens), the fourth has two, and the terminal segment bears six setae of which four are terminal (Fig. 13). The basipodite of the first maxilliped bears ten setae, located mostly along its pos- terior edge. The basipodite is 0.28 mm. long, the exopodite is 0.23 to 0.26 mm., and the endopodite is 0.25 to 0.31 mm. The longest seta on the exopodite is 0.32 mm. long. As in previous stages, there is a chromatophore in the distal end of the basipodite (Fig. 8). The second maxilliped bears ten terminal setae on the exopodite. The endopo- dite is three-segmented ; the first and second segments each bear one seta and the terminal segment has five setae (Fig. 14). The basipodite bears four setae and contains no chromatophore, but the coxopodite does have a chromatophore. The basipodite is 0.30 mm. long, the exopodite 0.24 mm., and the endopodite 0.09 to 0.11 mm. The longest seta on the exopodite is 0.31 to 0.34 mm. long. The abdomen, including the telson, is identical in every detail with the abdomen of the third zoea, except for larger size. DISCUSSION Churchill (1941, 1942) has described five zoeal stages which he assigned to CaUincctcs sap id us. His first zoea and second zoea do belong to this species ; they agree with our laboratory-reared zoeae in every detail except that Churchill ap- parently overlooked the little dorsal spine on each furcus of the telson, which is present in all blue crab zoeae from the first. However, Churchill's third and fourth zoeae, and presumably his fifth zoea also, belong to a different species of crab and have nothing to do with C. sapidus. Churchill's third zoea is slightly larger than our laboratory-reared blue crab third zoea, but the most striking differences are the presence of prominent dorsal spines on the fifth abdominal segment (lacking in C. sapidus) and the much greater length of the antennal exopodite (which in C. sapidus is unchanged from the second zoeal stage). Churchill's third zoea has six swimming hairs on the exopodite of the first maxilliped, and seven on the second, while C. sapidus has eight on each maxilliped. There are also minor differences, including the number of setae on the carapace, the degree of development of the ap- pendages following the second maxilliped, and perhaps the numbers of setae on the endopodites of the maxillipeds. I did not find Churchill's third zoea in my plankton tows, but I did find one specimen of his fourth zoea (in a surface tow off Ocean View, Va., Aug. 14, 1941) and can confirm the accuracy of his excellent figure and description of this stage. Churchill's fourth zoea is much more robust than the fourth zoea which I have as- signed to C. sapidus; the larger size is obvious to the naked eye. In addition to the possession of dorsal spines on the fifth abdominal segment and the elongated anten- nal exopodite, his species differs from mine by having seven (instead of three) setae on the posterior edge of the carapace, nine swimming hairs (instead of ten) on the exopodite of the second maxilliped, and greater development of the buds of 152 SKYVKLL H. HOPKINS appendages posterior t" the second maxilliped. The distal pair of spines on the furci of the telson are much smaller than the corresponding spines in blue crab zoeae and are on the inner side of the furci rather than the dorsal side as in C. sapidus. Hoth maxillipeds contain chromatophores in the basipodites in Churchill's species, while only the first maxillipeds of C. sapidus have chromatophores in the basipodite. Churchill's description of live zoeal stages which he assigned to C. sapidus was based entirely on specimens found in plankton, with the exception of the first zoea. The Virginia Fisheries Laboratory has hatched thousands of blue crab eggs in the laboratory and has reared a few specimens through the second and third zoeal stages, so that the identity of these stages is known beyond question. The identity of the fourth zoea, described and assigned to C. sapidus in the present paper, has not been confirmed in this way. It is certain that Churchill's third zoea does not belong to C. sapidus; probably it is a zoeal stage of some other crab of the family Portunidae. The zoeal stages of Ovalipcs occllatus, Arcnacns cribrarius, Bathyncctcs supci'ba, and Callincctcs oniatus (species which live in or near the mouth of Chesapeake Bay) have never been de- scribed, so it is possible that Churchill's species belongs to one of these. It seems probable that his fourth and fifth zoeae belong to the same species as his third zoea, and it is practically certain that they do not belong to Callincctcs sapidus. The zoeae of several species of Portnnns described by Lebour (1928) have a long anten- nal exopodite like Churchill's zoea, but none of them have dorsal spines on the fifth (or any other) abdominal segment. It seems probable that C. sapidus has a fifth zoeal stage which has not yet been seen by anyone. LITERATURE CITED CHURCHILL, E. P., 1941. The zoeal stages of the blue crab, Callinectes sapidus. Anat. Rec., 81 (Suppl.) : 37-38 (Abstract). CHURCHILL, E. P., 1942. The zoeal stages of the blue crab, Callinectes sapidus Rathbun. Publ. No. 49, Chesapeake Biol. Lab. HOPKINS, S. H., 1943. The external morphology of the first and second zoeal stages of the blue crab, Callinectes sapidus Rathbun. Trans. Atner. Micros. Soc., 62: 85-90. LEBOUR, M., 1928. The larval stages of the Plymouth Brachyura. Proc. Zool. Soc. London, July, 1928: 473-560. LOCH HEAD, M. S., J. H. LOCHHEAD, AND C. L. NswcoMBE, 1942. Hatching of the blue crab, Callinectes sapidus Rathbun. Science, 95 : 382-383. LOCHHEAD, M. S., AND C. L. NEWCOMBE, 1942. Methods of hatching eggs of the blue crab. Virginia Jour. Sci., 3 : 76-86. SAN no/, M., AND S. H. HOPKINS, 1944. Zoeal larvae of the blue crab Callinectes sapidus Kathbun. Jour. Wash. .-lead. Sci.. 34: 132-133. PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS PRESENTED AT THE MARINE BIOLOGICAL LABORATORY, SUMMER OF 1944 JULY 18 Melanophore control of the sexual dimorphism of jeathcr pigmentation in the Barred Plymouth Rock. B. H. Willier. This report deals with the effects of sex-linked genes on the expression of melanophores in the formation of sexual differences in the pigmentation pattern of feathers of the Barred Rock fowl. Melanoblasts from Barred Rock individuals of known sex were introduced into feather germs of the same or of a different breed of fowl, either by grafting them directly into the wing bud of a host embryo or by allowing them to migrate into melanoblast-free wing skin grafts. All possible combinations between melanoblasts of male and female genotypes and the sex of the host and sex genotype of the grafted skin were obtained. A black and white barred pattern was produced in host or graft contour feathers which was invariably in accordance with the sex genotype of the melanoblast introduced. The melanoblasts of the male which are homozygous for two dominant sex-linked genes, barring and silver, and for a dominant autosomal gene, extension, produced a narrow gray-black band and a wide, almost pure white band. Melano- blasts of the female which are homozygous for extension and hemizygous for barring and silver produced a wide black band and a narrow gray-white band. Thus the melanophores of male and female genotypes are provided with different properties for controlling (a) the relative width of the light and dark bands, and (b) the intensity of their pigmentation. The difference in expression appears to be determined by the number of sex-linked bar genes interacting with the rest of the genotype (mainly with extension and silver). Furthermore, this difference in expression of melanophores of male and female genotypes is manifested independently of sex hormones of the host and of the sex genotype of the feather germs of the grafted skin. It be- comes clear, therefore, that sexual dimorphism of barring is a manifestation of differences in expression of male and female melanophores as provided by the number of sex-linked bar genes interacting with the rest of the genotype. Since melanophores of the same genotype (from same individual), either male or female, produced a barred pattern which varies in quality from feather to feather in the host or grafted skin, it is evident that the individual feather germ has a more or less specific modifying influence on the rhythmic production of pigment by the melanophores. Growth rate appears to be one of the modifying factors since (1) in flight feathers the variations in barred patterns and in growth rate are roughly parallel, and (2) the differences in barred patterns produced in homologous flight feathers of the host and its donor by melanophores of the same genotype appear to be correlated with breed differences in growth rate, being generally higher in the white Leghorn host than in its barred donor. An increase in growth rate is usually correlated with an increase in the relative amount of black pigment in the vane and with a decrease in the distinctness of barring. Other modifying factors acting simultaneously in the feather germ are likewise involved. The effects of peripheral factors on motor neuron differentiation in the chid: embryo. Viktor Hamburger. If a wing or leg primordium is extripated in a two- or three-day embryo, the spinal ganglia as well as the lateral motor columns of the corresponding parts of the spinal cord become hypo- plastic. If the central nervous system is overloaded by implantation of a supernumerary limb, a hyperplastic effect is observed. The question arises whether the periphery affects the growth of the central nervous system by controlling the mitotic activity or the cellular differentiation. The lateral motor column was selected for an analysis of this problem. 153 154 PROGRAM AXD ABSTRACTS OF SCIENTIFIC PAPERS Mitotic counts were made in 20 wing extirpation cases, during the peak of mitotic activity (in most cases on the fifth and sixth day of incubation). A slight depression of the mitotic activity was Found <>n the operated side. However, the observations to be reported presently show that this response is transient and not related to the hypo- or hyperplasia of the lateral motor column. If cellular proliferation is not affected by wing extirpation or transplantation, then the sum total of all cells of the spinal cord should remain identical on the operated and the unoperated sides of the cord. This expectation was borne out by cell counts of the motor cells, and, separately, non -motor cells in the ventral half of the cord of five older embryos (two cases of wing-bud extirpation, two cases of leg bud extirpation, and one case of wing implantation). In all instances the total cell numbers were strikingly similar on the two sides. In all instances of hypoplasia, moreover, the deficit of large motor neurons was almost exactly compensated by an excess of small non-motor cells; in the case of hyperplasia, a surplus of motor cells was accompanied by a smaller number of non-motor cells. These data give convincing evidence that the proliferative activity of the cord is not permanently impaired by the operation ; they leave no doubt but that the peripheral factors control the process of differentiation of small indifferent cells into large motor neurons. An inductive effect is postulated which emanates from a small group of pioneer motor neurons, and which spreads over adjacent indifferent cells, inducing them to differentiate. The newly recruited neurons are added to the lateral motor column and thus increase its inductive capacity. This process of augmentation is not a self-perpetuating mechanism, however, but is under the "remote control" of conditions prevailing at the periphery. It is cut off at the moment when the periphery is saturated with nerve supply. Under experi- mental conditions, this would happen earlier (hypoplasia) or later (hyperplasia) than under normal conditions. The superficial yd layer and its role in development. Warren H. Lewis. Probably every cell and egg and ameboid organism has a superficial gel layer. It exerts continuous contractile tension, a fundamental property of gelled protoplasm. Gel layer and endoplasm are reversible states of the same cytoplasm that readily changes from one state to the other. Local increases and decreases of its contractile tension are responsible for many changes of cell form, extension of nerve axones, cell locomotion, flow of endoplasm, and cleavage of cells and eggs ; and during development for the infiltration and interpenetration by migration of individual cells among others of their own type, nerve cells for example, and among others of different types, capillary endothelial spreading, fibroblast infiltrations and myoblast migrations. The contractile tension of the gel layer over the yolk of the zebra fish egg squeezes endo- plasm out of the yolk to form the blastodisc and compresses the yolk globules into polyhedrons. After cleavage this gel layer pulls the blastodisc over the yolk (gastrulation). Mechanics of imagination. A relative increase of the contractile tensions of the gel layers on one surface of a series of adherent epithelial cells will result in a concave depression (in- vagination) on the side of the greater tension. Imagination is the resultant of two forces, a distorting one (the contractile tension of the adherent gel layers) and a resisting one (resistance of the cells to distortion). The cells suffer less distortion because of the invagination. In amphibia, an increase in the contraction of the adherent gel layers of the outer surfaces of the blastophore cells is responsible for one phase of blastopore invagination ( gastrulation). This contraction also pulls the presumptive endoderm and mesoderm into the walls of the primitive arrhenteron and the presumptive ectoderm and neural plate towards the blastopore in spite of an opposing contractile tension exerted by the adherent gel layers of the surface cells over the rest of the egg. The same mechanical principle is involved in the neural tube formation and its subsequent bendin.L's; optic vesicle evagination and invagination to form the optic cup; imaginations of the lens, otic vesicle, nasal pits, and probably in the evaginations of the thyroid, lungs, liver, pan- creas, etc. It also plays a leading and revealing role in the wound healing of eggs. The contractile tensions exerted by the superficial gel layer of cells thus plays a leading role in early morphogenesis and probably in later stages also, but the role it plays is dependent upon cytoplasmogenesis, cell division, cell growth (increase in protoplasmic mass), cell adhesions, accumulation of intercellular products, etc. PRESENTED AT MARINE BIOLOGICAL LABORATORY 155 JULY 25 Fcrritin and iron metabolism. L. Michaelis. A considerable amount of iron is stored in mammalian tissues which is not in the form of any iron-porphyrin compound. 60 years ago Schmiedeberg prepared an iron-containing protein from liver, which he designated as ferratin. It was an ill-defined, not easily reproducible sub- stance. In 1935 Lauffberger discovered that a protein containing as much as over 20 per cent of iron can be obtained as a well-crystallized compound in the form of its cadmium salt from horse's spleen and many other organs of various species. It was designated as ferritin. In 1942, S. Granick and L. Michaelis elaborated a method of separating the iron from the protein. This protein was designated as apoferritin. It resembles the globulins and is not, and does not contain, any nucleoprotein. Whereas ferritin, in spite of its high ability to crystallize, is non- homogeneous on ultracentrifugation, apoferritin is a perfectly homogeneous protein of molecular weight 500,000, as determined by Dr. Rothen. This protein is highly antigenetic. The pre- cipitin reaction shows that it is essentially different from all other known proteins, that it is species-specific, but not organic-specific. Ferritin and apoferritin cannot be distinguished by the precipitin test. Ferritin is brown, apoferritin colorless. In spite of the fact that about 20 per cent of its weight is withdrawn from ferritin, in the form of Fe, on converting it to apoferritin (or about 30 per cent of its weight in the form of ferric hydroxide), the crystal form of apoferritin and ferritin are alike even to such an extent that Dr. Fankuchen could find no essential difference in the X-ray diffraction pattern of the two, except for the fact that diffrac- tion lines are stronger in ferritin than in apoferritin. The method of preparing ferritin depends on the fact that in an aqueous organ extract, on heating at 80° C. (but not higher) most of the proteins are coagulated, but not ferritin. From the filtrate the remaining protein is salted out by 30 per cent ammonium sulfate, the coagulum redissolved in water and CdSO4 is added. Crystallization proceeds rapidly. The crystals are always isotropic, in the cubic system, either octahedra, often twinned octahedra (horse), or cubes, or tetrahedra in other animals. Human organs yield crystals with slightly curved faces, less regularly shaped. Ferritin can be prepared from many mammals, but not so far from cats or deer, or non-mammalian vertebrates, or invertebrates. It is found in spleen, liver, and bone marrow, and to a small extent (and usually with a smaller iron content, resembling apofer- ritin), in kidney and testes, but not in blood or muscle. When radioactive iron is injected into anemic dogs, the iron can be retraced after a short time as ferritin in the liver, and under certain circumstances also in the spleen. The iron of ferritin is always in the ferric state, it is paramagnetic and has a magnetic susceptibility of a characteristic magnitude which does not occur in any other normally occurring iron compound of the organism. It corresponds to a magnetic moment, per gram-atom Fe, of 3.8 Bohr magnetons, which according to theory indicates the presence of three unpaired elec- trons, whereas in other ferric compounds the number of unpaired electrons in one iron atom is usually either five or one. The ferric hydroxide precipitated from ferritin by NaOH has the same magnetic characteristic. The facts known so far are scarcely compatible with the assumption that one has to deal with a stoichiometrically well defined iron salt, or iron complex compound, of apoferritin. Very likely, micelles of colloidal iron hydroxide of composition mainly FeOOH, containing small amounts of ferric phosphate in addition, are interspersed in the open spaces of the very loose crystalline structure of apoferritin. Why a special protein is needed for the storage of the iron can not be explained as yet. It is however obvious that iron is accumulated in the form of ferritin as hemoglobin is broken down, and that ferritin furnishes iron lor the formation of fresh hemoglobin. It is remarkable that the iron of ferritin is always in the ferric state, that of hemoglobin however in the ferrous state. Theory of uietachromatic staining. L. Michaelis. Methods of differential staining, as used in histological technique, may be divided into two classes. The substrate to be stained may be subjected to two (or more) different dyestuffs, either simultaneously or successively, whereby different histological elements absorb selectively the one or the other dye : selective staining. Or, one individual dycstuff may be used to stain various histological elements in different shades of colors : metachromatic staining. Among 156 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS dyestuffs exhibiting tin- inetachrotnatic effect, toluidine blue and thionine are best known. Tbey stain, for instance, all nuclei, or the cytoplasm of lymphocytes, blue; however the same dyes stain the granula of basophilic leucocytes, or mucus, or "amyloid" pathologically occurring in liver and other organs, purple. This difference in color does not depend on pH within wide limits and should be well distinguished from the pH-effect of indicator dyes. All substrates which stain "metachromatically," instead of "normally," are half-esters of sulfuric acid with high-molecular carbohydrates, either linked to a protein or not, e.g. chondroitin sulfuric acid, mucoproteins, and many vegetable colloids among which agar is best known. Metachromatic dyes, therefore, represent a specific chemical reagent that can be used microscopically in situ. Many more dyes exhibit this effect, if the color analysis is carried out spectrophotometrically. It is a more accidental property of the two dyes mentioned above, to reveal the metachromatic effect directly to the unaided dye. On spectrophotometric observation, methylene blue shows this effect at least to the same extent, though it is not very obvious for the unaided eye. A comparative study of many dyestuffs has revealed that all those dyestuffs stain meta- chromatically which, in aqueous solution, have the property of forming polymers of the dyestuff molecules in equilibrium with the monomeric dyestuff molecule. Even in rather dilute solution, such dyestuffs form dimeric molecular aggregates, which can be recognized by the fact that the dimers have an absorption band at a wave length different from that of the monomer. Since the percentage of dimerised dye molecules increases with increasing concentration, the molar absorption coefficient, plotted against wave length, varies with the concentration. Such dye stuffs are said to "disobey Beer's law," according to which the molar absorption coefficient should be independent of concentration. The absorption curve of the metachromatic color, however, is not the same as that of the dimeric dye, rather is the absorption band still more displaced. There is evidence that the metachromatic color is due to high-polymers of the dye. The difference between normal and metachromatic staining, then, consists in the fact, that the surface of "normally staining" sub- strates adsorbs a monomolecular layer of the dye, and the surface of metachromatically staining "substrate" absorbs a polymolecular layer. All dyestuff molecules are long, flat, almost two- dimensional molecules. Di- and polymerisation consists in piling up these flat molecules plane to plane. The forces which bring about polymerisation are, in a loose sense, comparable to what we may call exaggerated van der Waal's forces. In every case, the metachromatic color turns to the normal color by increasing the temperature, in a reversible way, due to the fact that thermal motion disrupts the aggregates of the dye molecules. The correlation of the polychromatic effect of a dyestuff with its chemical structure will be discussed in some other place. The following characteristic example may be mentioned. Thionine contains two amino groups at the ends of the elongated molecule, an electric charge oscillating from the one to the other amino group ("resonance"). It is highly metachromatic. .Substituting O for S (oxomine) almost abolishes metachromasy. Eliminating one of the two .amino groups (monoaminothiazine) destroys metachromasy entirely. formula of thionin (univalent cation, as existing in neutral or slightly acid solution) The chemical organization of the cytoplasm. Arnold Lazarcnv. The organization within the cell permits the coexistence of otherwise incompatible sub- staiu ( - -mil as phosphatase and certain coenzymes. Although the mechanism is little under- stood, nil organization is determined, in part, by the localization of enzymes within the cell. 'I he cytoplasm ot the liver cell contains several granular components. These may be separated by differential centrifugatioii after cell fragmentation. The larger particles are the mitochondria. The smaller particles, which are submicroscopic in si/e, are of two distinct types one is particulate glycogeii, the other is a lipo-protein complex which can be differentiated Irom the mitochondria by quantitative chemical analysis. Most of the eytoplasmic lipids are PRESENTED AT MARINE BIOLOGICAL LABORATORY 157 concentrated in the mitochondria and submicroscopic lipo-protcin participate. Both contain ribose nucleic acid. These particulates serve as centers of enzymatic localization within the cell. Since both mitochondria and the submicroscopic lipo-protein particulate oxidize succinic acid, they must contain at least three of the respiratory enzymes — succino dehydrogenase, cytochrome C, and cytochrome oxidase. The mitochondria in addition contain glutamic dehydrogenase. Spatial orientation of the components of the respiratory chain, within the particle, may serve to direct metabolic activity. The manifestations of a reversible structural framework within the cytoplasm, as evidenced by thixotropy, birefringence (at times), and sol-gel transformations may be explained by the existence of a thread-like micelle similar in type to the tobacco mosaic virus. These elongated micelles can produce a rigid structure even though they are spatially separated by I50A (a distance several times the diameter of an albumen molecule). Thus although cytoplasm is organized into particulate components (significant for the localization of some of the respiratory enzymes), micellar components (which may give rise to a reversible structural framework), it nevertheless has a continuous aqueous phase. AUGUST 1 Native protein crystallography and diffraction patterns. Dorothy Wrinch. Any attempt to understand how native proteins operate in living systems must be based on a knowledge of their atomic organization and of the architectural patterns which characterize them. Unfortunately the instability of these molecules precludes all but the most delicate methods. So far the best and perhaps the only techniques that do not destroy the structure we wish to study are those of physics, particularly X-ray diffraction investigations of protein crystals. While the protein molecule is not known to be biologically active in crystalline form, a full analysis of insulin crystals (for example) would indicate the nature of the surface pattern of the constituent molecular units. It is confirmed by recent crystallographic work (Fankuchen, Ann. N. Y. Ac. Sci. 41 : 157; 1941) that the surface pattern of each protein species, as indicated by the interlinking of units in the crystal, is a highly specific and individualistic tapestry of atoms and electrons. Such molecules can crystallize with vastly different complements of foreign molecules (often water) and for their stability depend on this foreign population. Moreover, irrespective of the foreign element, the protein molecule interlinks with its fellows in character- istic ways and maintains intact its own skeletal structure. In striking contrast to many com- plex organic crystals, the native proteins in general form crystals of high symmetry. While these studies have yielded important information about molecular weights and some information about shapes and sizes, they have not as yet uncovered the atomic structure of the native protein particle. To obtain any light on this problem from these studies, it is necessary to interpret the intensity data obtained from X-ray observations. This is a matter of the great- est difficulty. One line of attack is to study the nature of diffraction patterns in general, a problem of absorbing interest to astronomers for more than a hundred years. Diffraction patterns of apertures of various shapes, repeated in various arrangements, are available for study. It is suggested that the study of such patterns as these may enable us to begin to learn the language of diffraction patterns. A more fundamental attack is also suggested, namely the methodo- logical study of the diffraction patterns of distributions of each and every type. An introduction to such studies by the present writer is in course of publication. The aim is to obtain a clear picture as to how structural features of a known distribution manifest themselves in structural features of its diffraction pattern. This correlation of distributions with their diffraction pat- terns constitutes an assemblage of mathematical facts essential for the adequate exploitation of the experimental data. The role of adenylpyrophosphatase in alcoholic fermentation of yeast. Otto Meyerhof. The Naples Station still lives! Ernst Scharrer. 158 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS AUGUST 8 On the energy source oj the nerve action potential. David Nachmansohn. In earlier theories acetylcholine (ACh) was supposed to be a "synaptic" transmitter, i.e. a substance released at the nerve ending and acting direcly on a second neuron or on the effector cell. According to the new concept the release and the removal of ACh is an intra- ccllular process occurring everywhere at the neuronal surface and directly connected with the nerve action potential. The action of ACh may be pictured in the following way: The nerve is surrounded by a polarized membrane. The polarized state of the membrane is due to a selective permeability to K, which is present in different concentrations on either side of the membrane. During the passage of the impulse the resistance of the membrane is decreased and the permea- bility to all ions increased. Hereby a local depolarization occurs. This change in permeability appears to be produced by the rapid appearance and removel of ACh. The polarized point becomes negative- to the adjacent region and flow of current results. This flow of current stimulates the next following point. There again ACh is released and the whole process re- peated. The impulse is thus propagated along the axon. At the nerve ending the surface is incicast.'i, the resistance therefore decreased. This leads to a greater flow of current which enables the impulse to cross the non-conducting gap. The transmitting agent is always the electric current, the action potential, but the current is generated by ACh. The picture is con- sistent with the idea of propagated impulses as developed by Keith Lucas and Adrian. It makes unnecessary to assume that the transmission along axon and across synapses differs funda- mentally. If the release and the removal of ACh are responsible for the alterations of the nerve mem- brane during the transmission of the nerve impulse, chemical reactions must supply the energy for the resynthesis of ACh. The electric organ of Rlectrophorous clcctriciis offers a suitable material for comparing electrical and chemical changes connected with the action potential since both are in the range of possible measurement. Such measurements were carried out during the last two years (Nachmansohn, Cox, Coates and Machado). The electric energy released per gram and impulse was found to be eight microcalories, the total electric energy about 48 microcalories. The energy released by phos- phocreatine breakdown is about 32 microcalories, that by lactic acid formation 15-18 micro- calories. Since the energy of lactic acid formation is probably used, as in muscle, to phos- phorylate creatine ("Parnas reaction"), these figures are consistent with the conclusion that phosphate bonds may yield the energy for ACh synthesis. The amounts of ACh which may be split by one gram of electric tissue during one discharge is about 5 X l()~G millimole. The amount of phosphocreatine actually split per gram and im- pulse is about 3 X 10 -° millimole. Thus the amounts of ACh and phosphocreatine metabolized seem to be of the same order of magnitude. One of the facts supporting the new concept is the extremely high concentration everywhere at the neuronal surface indicating a rate of ACh metabolism sufficiently high to parallel the electric changes. In electric tissue the rate may be at least 100,000 times but probably one million times as high as that of respiration. But we have to distinguish between the possible rate and the absolute amounts metabolized. ACh is released and hydrolyzed within a very short period. The recovery requires one to two hours during which the rate of respiration may be increased. If the absolute amounts are compared, a satisfactory picture is obtained. A whole chain of reactions connected with the nerve potential could be established. Since it is initiated by the release of ACh, it has been called the "acetylcholine cycle." As a result of these investigations a new enzyme, choline acetylasc, could be extracted from brain which in presence ot adenosine triphosphate under strictly anaerobic conditions and in cell- tree solution synthesizes ACh. Current, v»lfa thereby restored, and the blood pressure returned to normal. Gelatin as a plasma substitute, with special reference to pseudo-agglutination. Richard G. Abell and William M. Parkins. That gelatin possesses many of the properties desirable in a plasma substitute has recently been demonstrated by Parkins, Koop, Riegel, Vars and Lockwood (1943). The present in- vestigations on pseudo-agglutination are a part of a larger program being carried out at the University of Pennsylvania by Dr. Parkins and others in which various properties of gelatin in reference to its use as a plasma substitute are being further investigated. The results to be described in this abstract have previously been mentkmd briefly in the paper by Parkins et al. referred to above. It is well known that the addition of gelatin to erythrocytes in vitro causes these cells to adhere in clumps. This phenomenon, known as pseudo-agglutination, has been reported to follow infusions of gelatin by Hanzlik and Karner (1920), Amberson (1937), Stein, Grodins and Button (1943) and Grodins (1943). If such clumps are formed within the blood vessels, it is important to know whether they interfere with the blood flow. Six rabbits and five dogs were infused with six per cent solutions of biological gelatin (courtesy of Dr. D. Tourtellotte, Charles B. Knox Gelatin Co., and Kind and Knox Gelatin Co.) in 0.85 per cent saline, and the effect upon the blood vessels and blood flow observed directly with the microscope in the living animal. In the case of the dogs, the vessels studied were inclosed within intestinal-mesenteric chambers, modified to fit the dog from the original type described by Zintel (1936). Following acute hemorrhage to the point of reducing the blood pressure to 30 mm. Hg, the blood in many of the capillaries and venules became station- ary ; in the arterioles the flow became sluggish. Replacement of the blood lost by hemorrhage with an equal volume of gelatin caused the blood flow to return to its control rate. In the normal rapid flow in the dog's mesentery no clumps could be detected following gelatin infusion. Only when the rate of flow was decreased and stagnation induced (by further hemorrhage) could such clumps be seen. In order to secure further evidence on the manner in which gelatin causes pseudo-ag- glutination of erythrocytes, transparent moat chambers (Abell, 1932) were inserted in rabbits' ears. After these chambers became vascularized, six per cent solutions of gelatin were injected intravenously (15 cc./kg. ) and the effect studied with the microscope. No pseudo-agglutination could be observed until the rate of flow was reduced by squeezing the main artery of the ear, which supplied the vessels in the chambers. When this was done, it could be seen that erythrocytes that came together side by side adhered to each other to form uroiips of cells in rouleaux. In control experiments this occurred normally, in the absence of gelatin. Following gelatin infusions, however, the individual rouleaux groups adhered to each other to form larger clumps, made up of several rouleaux groups. Following gelatin injections, the erythrocytes were seen to pass from the arterioles into the capillaries as separate cells. They did not adhere to each other to form pseudo-agglutinated clumps to any appreciable extent, until they reached the venules. Such clumps did not inter- Eere with the blood llo\\. When a clump approached a vessel smaller than it was, it separated into its constituent cells, and hence did not block the vessel. The clumps floated in the plasma, vhich carried them along, and which was always between them and the walls of the vessels. 'I here was no evidence of increase in viscosity of the blood. I bus, although pseudo-agglutination of erythrocytes does occur following intravenous Injections of <.'clatin, such pseudo-agglutinated cells do not block the vessels or interfere with the ( onsequcntly, from the standpoint of flow, pseudo-agglutination does not contra- indicate the use of gelatin as a plasma substitute. PRESENTED AT MARINE BIOLOGICAL LABORATORY 163 AUGUST 24 Experimental studies on the cytoloyy of alliiiin. C. A. Berger. The cytological effects of a number of chemical agents were studied in the root tips of Allium ccpa and compared with the well known effects of colchicine. Acenaphthene, veratrine, sulfanilamide, chloral hydrate and benzene all induced polyploidy and produced cytological effects similar to those of colchicine. The primary effects of all these substances are the prevention of the formation of an effec- tive spindle and a delay in the division of the spindle attachment regions of the chromosomes. As a result of these primary effects the chromosomes are held at metaphase and become shorter and thicker than normal metaphase chromosomes. After a longer or shorter delay the spindle attachment regions divide, but in the absence of an effective spindle no anaphase movement takes place and the whole group of chromosomes undergoes a revision process giving rise to a tetraploid resting nucleus. After a sufficient period of recovery these cells undergo mitosis as polyploid cells. Root tips grown in an atmosphere lacking oxygen were found to show similar cytological effects and to produce tetraploid cells. The tentative conclusion is advanced that these effects are not specific to any of the chemicals in question but are general effects common to all the substances in question and interfering with some fundamental metabolic process concerned with the formation of the spindle and the division of the spindle attachment region. Naphthalene-acetic acid was also used. This substance does not affect the meristem but induces polyploid divisions in the older, differentiated regions of the root. Naphthalene-acetic acid differs from the other substances in that it does not affect cells in division, but causes a double reproduction of the chromosomes in the resting nuclei. At metaphase tetrachromosomes are found. These are four chromosomes held together at a common undivided spindle attach- ment region. After a slight delay at metaphase two successive divisions of the spindle attach- ment regions occur. Naphthalene-acetic acid does not interfere with the formation of the spindle and anaphase separation occurs, resulting in two tetraploid cells. Studies on the chemical basis of fever. Valy Menkin, M.D.3 Fever is usually associated with some form of cell injury. Inflammation is the complex vascular, lymphatic, and tissue response in vertebrates to the presence of an irritant and as such, it represents a manifestation of severe cellular injury. The pattern of injury in inflammation has been shown in earlier studies to be referable to a thermolabile, non-diffusible substance located in the euglobulin fraction of exudates. (Arch. Path., 1943; 36: 269.) This substance, termed necrosin, appears to be either a proteolytic enzyme or else to have proteolytic activity associated with it. Necrosin, in the form of the toxic euglobulin of exudates, is pyrogenic to both dogs and rabbits. (Proc. Soc. Exp. Biol. and Med., 1943; 54: 184; Fed. Proc., 1944; 3: No. 1.) Its formation at the site of injury and its absorption into the circulation offers a reasonable explanation for the basic mechanism of fever accompanying numerous inflammatory processes. Recent studies indicate that this toxic euglobulin contains a component in turn insoluble in the presence of NaCl or SO4=. This component is essentially the fever-inducing factor or at least it is associated with that fraction. It can be dissociated from necrosin by treating the exudate with ammonium sulphate at one-third saturation. The precipitate formed is treated with distilled water prior to dialysis of the SO4= ions. A true euglobulin enters into the aqueous phase containing the SO4= ions. This is necrosin in a further state of purification. It is toxic to mice and is capable of inducing a severe cutaneous inflammation ; but it is non-pyrogenic. The pyrogenic factor seems primarily associated with the precipitate which has failed to dissolve in the aqueous phase containing the SO4= ions. This highly fever-inducing substance, readily dried by freezing, is termed "pyrexin." Its presence offers a satisfactory explanation for the basic mechanism of fever with inflammation. It is thermostable. Boiling fails to inactivate pyrexin. Ashing destroys it. Incubation of the non-pyrogenic purified necrosin favors the formation of pyrexin. This suggests that pyrexin may be an end product of proteolysis associated with enzymatic activity in the necrosin fraction. It is absent in non-hemolyzed serum, but it is present 3 Fellow of the Guggenheim Research Foundation. 164 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS to some extent in hemolyzed scrum and in serum from an animal with a concomitant acute inflammation. It is absent in tin- pseudogMmlin and albumin fractions of exudates. Pyrexin is excreted, at least in part, in urine. The N and P contents of pyrexin are about ten per cent and one per cent respectively. The material is Biuret negative but Ninhydrin positive, except in the fraction recovered from urine which is usually also Ninhydrin negative. It is Molisch negative. It is insoluble in ether and 95 per cent alcohol, but apparently soluble in relatively \\eak alkali. The possibility of a peptide attached to a nucleic acid derivative is not precluded by the available data. The exact chemical nature of pyrexin is, however, unknown, and will therefore require further studies. Evidence with barbiturates and antipyretics indicates that the possible mode of action of pyrexin is on the fever centers in the hypothalamic region. In vitro fertilization and cleavage of human ovarian eggs. John Rock and Miriam F. Menkin. (This paper has already appeared in Science, 100: 105-107, August 4, 1944.) AUGUST 29 Phosphoprotein plwsphatase, a new enzyme from the frog egg. Daniel L. Harris.4 Immediately following homogenization of the ovarian eggs of the leopard frog, Rana piplcns, in 0.1 M sodium citrate, there is a rapid liberation of inorganic phosphate from some ester within the eggs. The reaction takes place over a broad pH range in neutral or acid solutions, but there seems to be relatively little hydrolysis in an alkaline medium. There is a pronounced optimum at or near pH 5.0. At pH 5.0 the inorganic phosphate rises from about 23 mg. per cent to 250 mg. per cent in 5 minutes. The speed of the reaction as well as the pH optimum indicates that the hydrolysis is due to an enzyme rather than to the acid conditions. Further- more, the activity is destroyed by heating. An analysis of the changes in distribution of phosphate in the brei following "autolysis" as compared with the control in which the reaction was prevented by the addition of trichlo- racetic acid showed the following : There was no significant change in phospholipid, an increase instead of the expected decrease in organic acid-soluble esters, but a profound decrease in phosphate bound to protein. This decrease in protein phosphate was sufficient to account for the increase in inorganic phosphate as well as the increase in organic acid-soluble phosphate esters. The latter esters have not been identified as yet, but they are known to be resistant compounds, withstanding 1 N HC1 for three hours at 100 ° C. Nucleoproteins and the phosphoproteins of the yolk platelets are the main proteins con- taining phosphate in the frog egg. To distinguish between these two possible substrates, tests were made using nucleic acid and casein, a typical phosphoprotein, as substrates. There was a rapid liberation of phosphate from casein, but little or none from nucleic acid. Indeed, nucleic acid appeared to inhibit the action of the enzyme on the natural substrate. Vitellin, isolated from the yolk of hen's eggs, was attacked as were yolk platelets which were denatured by heat and added in excess. Sodium /3-glycerophosphate and disodium phenylphosphate were hydro- lyzed but slowly if at all. The enzyme is, therefore, regarded as a phosphoprotein phosphatase, and the natural substrate within the frog egg is yolk. In the developing embryo the enzyme is thought to liberate inorganic phosphate as needed for carbohydrate metabolism. The enzyme is bound, in part at least, to the yolk platelets. Lack of contact between the enzyme and substrate can not, therefore, explain the fact that the hydrolytic action of the enzyme is held in check in the ovarian eggs which remain essentially unchanged throughout many months. Behavior and tube building habits of Polydora ligni. Edith Mortensen and Paul S. Galtsoff. Polychaete worms of the genus Polydora live on mud bottoms where they cause profound changes by gathering and depositing huge quantities of mud, often covering and smothering other •' National Research Council Fellow in the Natural Sciences. PRESENTED AT MARINE BIOLOGICAL LABORATORY 165 inhabitants \vith a thick layer of material formed of their loosely constructed mud tubes and excreta. The worms also penetrate between the mantle and the shell of oysters where so-called mud blisters are formed. Thus they sometimes change from a free living to a commensal existence. In tube building the mud is gathered in a deep ciliated groove extending the length of the inner margin of each of a pair of tentacular cirri as they lash about and secrete a sticky mucus. As the worm turns on its long axis, the mucus-covered mud is passed to the basal end of the tentacles and then is dropped, forming a ring about the anterior end of the body. The mucus is sufficiently adhesive to cause the particles of mud to stick together readily and immediately. There is no systematic placing of the mud in a rotary fashion at the edge of the tube, nor is the material packed together in any way. That tubes are essential for the survival of Polydora liyni was demonstrated in experi- ments in which two groups of 20 animals were kept under identical conditions with the exception that one group was kept in glass tubing whereas the other was placed free in sea water. On the 23rd day of the experiment, all of the 20 worms left free had perished while 16 of the 20 kept in tubes were still alive. Rejection of material unsuitable for tube building or food is accomplished by the reversal of ciliary motion along the tentacular groove. When an animal is given substances such as corn starch, Chinese ink, finely ground glass, and sand grains of various sizes, the number of reversals recorded per unit of time is greater than when the same animal is given mud from its natural environment. Likewise the number of reversals occurring when mud soaked in a M/40 KCL solution is used is greater than with mud alone. A 0.01 per cent lactic acid solution added to corn starch gives more reversals than corn starch alone. Thus physical and chemical factors control the reversal of ciliary motion. The funnel shaped pygidial structure at the posterior end of the body is not a sucker as has been suggested by previous authors but probably a plunger for clearing the tube of excreta. The conditions which stimulate the entrance of the worms 'into oysters and cause the free living animal to assume a commensal habit remain unknown. Apparently the worms are not attracted by the oysters and, as laboratory experiments show, may remain in close association with them without penerating their shells. Infestation of oysters by Polydora may be a purely accidental phenomenon. The click mechanism in clatcrid beetles. J. B. Buck. Several thousand tests on four species of elaterids showred that in click -jumping from an initial dorsum-down position the beetles come to rest on their feet about twice as often as on their dorsal surfaces. In testing the possibility that this is achieved by controlling the number of aerial loops or twists so that an upright landing results, it was found that : ( 1 ) Varying the ratio of height jumped to distance fallen caused no consistent change in the percentage of "successful" (up- right) jumps. (2) A tabulation of the direction in which the beetles were facing after jumping indicated that a position in the original plane of the longitudinal axis is favored, and that among the successful jumps those which end with the beetle facing the same direction as originally are somewhat more common than expected. Such a position could only result from a jump including n+ l/i twists (rotation on the longitudinal axis) and n + Vz loops (rotation on the transverse axis). In testing the alternative possibility that the excess of successful jumps is due to body shape or weight distribution, beetles were shaken in a box and dropped or thrown on to a level surface. Beetles dropped in such random fashion land upright about as frequently as after normal jumps, and moreover it makes no difference whether the beetle is dead or alive. Suc- cess is enhanced — sometimes to 85 or 90 per cent — by jumping or dropping the beetles on in- clined planes, indicating that anything which increases the probability of rolling or bouncing increases their chance of reaching the upright position. The heights reached in jumping are distributed normally, and the proportion of successful jumps is the same in each height class, so that if success depends on completing a particular number of aerial loops or twists, that number is a constant, independent of the height reached. Changes in temperature do not affect significantly the proportion of successful jumps. 166 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS The evidence summari/ed indicates that although there may be some path-selection in the mechanism ot the jump, the major lactor in the preponderance of successful jumps is a pre- disposition toward the attainment of the upright position, probably due to the external shape of the body. Mi'tujuorf>li<>sis in /lie lari'a of flic Tunicate, Stycla partita. L. M. Bertholf and S. O. Mast. The organisms used in this investigation were kept continuously at 20-21° C. The average length of larval life (i.e. the time between hatching and the retraction of the tail) varies with the season. It increased from approximately five hours late in June to a maxi- mum of SO hours early in August and then decreased somewhat. There is great individual variation. It ranged from less than one hour to more than eight days during the course of the summer. In sea-water in which a few hundred larvae per cc. had previously metamorphosed, the length of larval life is much shorter than it is in normal sea-water. As the number of larvae \\hich had metamorphosed in a given quantity of sea-water increases, the rate of metamorphosis in fresh larvae in this sea-water increases to a maximum and then decreases to zero, i.e. the sea-water becomes so toxic that it kills the larvae before the tail is retracted. Cupric chloride (2 :< 1CH3 M.) and Janus green (one part in 250,000 of sea-water), each acting for 3 minutes, greatly accelerate metamorphosis. Neutral red, extract of the muscle of a rabbit killed by x-rays, and concentrated sea-water accelerate it somewhat. Increases in hydrogen-ion concentration from pH 8.05 to pH 7.47, dilution of sea-water up to 40 per cent, crowding of the larvae (1000 per cc. of sea-water) have no measurable effect. \\ e postulate the following hypothesis of metamorphosis in this animal: It is known that the larva consists of organized adult tissue and organized larval tissue. The larval tissue prob- ably produces a substance which in low concentration augments metabolism and in high concen- trations retards it. This substance, then, would be more concentrated in the larval tissue, where it is produced, than in the adult tissue, into which it diffuses. It therefore, at a certain concen- tration, would retard metabolism in the former and augment it in the latter to such an extent that metamorphosis is initiated. This substance doubtless diffuses out of the larval tissue into the surrounding medium. Sea-water in which metamorphosis has occurred, therefore, contains some of this substance, which diffuses into fresh larvae put into it, and consequently increases the substance in the larval tissue and thus accelerates metamorphosis. Cupric chloride, Janus green, and other compounds which accelerate metamorphosis probably merely increase the retarding effect of the substance produced by the larval tissue. Vol. 87, No. 3 December, 1944 THE BIOLOGICAL BULLETI PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY A PRIMITIVE COCCID CHROMOSOME CYCLE IN PUTO SP. SALLY HUGHES-SCHRADER Department of Zooloyy, Columbia l*nivcrsit\. AYri' York INTRODUCTION Every coccid thus far studied presents such striking peculiarities in its meiosis that we are confronted with the paradox of regarding a simple and orthodox maturation process as of especial interest. Puto sp. of the family Pseudococcidae reveals a primitive chromosome cycle possibly archetypal for coccids. Only in the llaveiine tribe of the family Margarodidae have partially comparable conditions been encountered. Thus in Llarcia bonvari we find as probably primitive traits a sex ratio which approaches equality, no trace of parthenogenesis nor of hermaphro- ditism, and an XX - XO sex chromosome mechanism. But even in this relatively generalized species a highly specialized achromatic figure has been evolved in male meiosis, and asynapsis of one pair of autosomes is already established as a constant and normal feature in a certain percentage of the spermatocytes. Moreover, in Llaveia the secondary pairing of homologous chromosomes just prior to the second meiotic division provides a mechanism which ensures segregation without previous synapsis — an essential preliminary step to the successful operation of the com- pletely asynaptic habit as encountered in the related genus Protortonia. Puto, while it shares with Llaveia the primitive traits listed above, shows none of the specializations just enumerated. A survey of its cytology discloses a primitive, typically hemipteran pattern and further permits the recognition of certain phenomena as basic coccid characteristics independent of the specialized modi- fications encountered in the different groups. MATERIAL AND METHODS Specimens of this coccid have been deposited with Dr. Harold Morrison of the Bureau of Entomology and Plant Quarantine, U. S. Department of Agriculture. Washington, D. C., to whom grateful acknowledgment is made for his assistance. Dr. Morrison reports that the genus Puto is in so confused a state taxonomically as to preclude a specific identification at the present time. The material was collected near the village of Tequisistlan, Oaxaca, Mexico, in November 1933. All instars were represented at this time, as again in more sporadic infestations found near Tehuantepec, Oaxaca, in December, 1938. The favorite host plant was the stinging-haired Jatropha known locally as the "mala mujer." 167 168 SALLY HUGHES-SCHRADER Male nymphs of the third and fourth instars and adult females with eggs and embryos were dissected in Allen's Bouin. This fixative gave good results in embryonic and late meiotic stages but proved unsuitable for early meiosis. Male material was sectioned at four and female at six micra, and stained in Iron Hema- toxylin. Chromosome Complement The chromosomes of the female Puto are 14 in number and comprise 7 pairs differing slightly in length (Fig. 1). The male diploid set numbers 13, of which the next to shortest element is the unpaired sex chromosome (Fig. 2). Somatic Mitosis Somatic mitosis conforms to the hemipteran type. Its most characteristic features derive from the possession by the chromosomes of a diffuse, in contrast to a localized, kinetochore. Thus the whole body of the chromosome orients at metaphase, chromosomal fibers form from the poleward surface of each chromatid along its entire length, and anaphasic disjunction is parallel (Figs. 3, 4, and 5). In Puto the chromosomal fibers converge to division centers in which a minute centriole may often be discerned. Neither astral rays nor continuous fibers are present. The association between the constituent chromatids of the chromosome is closer throughout the mitotic cycle than in most coccids. (This effect is en- hanced in the present material by the stain used.) Thus in the metaphase chromosome the two daughter chromatids only are usually distinguishable, although in an occasional end view a four-parted structure is suggested (Fig. 4). Anaphasic disjunction is parallel for about one third of the inter-center distance (Fig. 5) ; in late anaphase, as in most coccids, each chromosome curves toward the division center (Fig. 6). Female Meiosis The ovary of the young female conforms in structure to the usual coccid type. There is no trace of hermaphroditism in any instar. Meiosis is completely normal throughout its course. Seven normal bivalents are formed, and invariably two polar bodies are successively given off. This has been confirmed in many eggs, from several different females. Fusion of male and female pronuclei, while both polar bodies or their derivatives are still recognizable peripherally, has been observed in several eggs. Furthermore no haploid embryos have been found among some hundred checked. Thus, although no final conclusion is justified without the confirmation of breeding experiments, all the cytological evidence indicates the absence of parthenogenesis of either diploid or haploid type. Male Meiosis a. Propliascs; structure and orientation of bivalents. From diakinesis on, the major features of male meiosis can be followed with adequate clarity. Earlier stages fix too poorly for a detailed analysis but appear to be entirely normal. There is no evidence of any anomalous behavior, such as A PRIMITIVE CO.CCID CHROMOSOME CYCLE 169 vesicle formation, or a difference in rate of condensation between haploid sets of chromosomes, such as is associated with the variant degrees of asynapsis en- countered in other coccids. A normal synapsis may safely be assumed. This is confirmed by the diakinetic bivalents. They are invariably six in number ; no univalents other than the sex chromosome are present. At diakinesis the autosomal bivalents and the sex chromosome are found peripherally distributed, closely underlying the nuclear membrane. The course of the constituent chromatids cannot be followed throughout the bivalents but open X--, FIGURES 1-6. Somatic mitosis. (All drawings made with camera lucida at table level with Zeiss 2 mm., 1.3 n.a. obj. and 20 X oc. ; enlarged with pantograph; magnification as repro- duced 2700 X.) FIGURE 1. Polar aspect of metaphase, female. FIGURE 2. Same, male. FIGURE 3. Lateral aspect of metaphase ; entire chromosome oriented, chromosomal fibers from entire length of chromosome. FIGURE 4. Same — one focal level only drawn — showing ends of a group of chromosomes. FIGURE 5. Early anaphase — disjunction parallel. FIGURE 6. Late anaphase — chromosomes curve toward centers. cross configurations (center, Fig. 7) suggest the resolution of a chiasma by rotation of the arms. Bivalent C of figure 8 would similarly be interpreted as a later stage in the same process. But the question of chiasmata aside, it is evident that in the marginally placed bivalents of figure 7 and in A and B of figure 8, the homologous chromosomes of each bivalent are assuming an end to end juxta- position. In the interpretation of these bivalents it must be remembered that no localized kinetochore is present in these chromosomes. The median knots in bivalents such as A and B in figure 8 thus represent chromosome ends and not, as might be assumed on superficial scrutiny, kinetochores. Similarly, the large central 170 SALLY HUdHES-SCHRADER aperture of the bivalent separates originally sister chroniatids in the vertical, and homologous chromosomes in the horizontal arms. b. Metaphase I. Shortening and thickening of the chromosomes proceed rapidly and with no change in the position of the homologues in relation to each other. By metaphase each bivalent is a compact, superficially four parted body; but extreme as is the condensation undergone, a polar view of an early metaphase plate (Fig. 9) still gives evidence of the end to end alignment of homologues in some of the bivalents. In the metaphase orientation, as expected under the influence of the diffuse kineto- chore. the long axis of each chromosome lies at right angles to the spindle axis. The constriction visible in each bivalent from the polar view is therefore the primary split — in this case the point of contact between the ends of the two homologous chromosomes. The constriction visible from the lateral aspect (Fig 10), which becomes the plane of separation in the ensuing division, is accordingly the secondary split. The first division, patently equational for the X chromosome, is thus, disregarding crossing-over, basically equational in character for the autosomal bivalents also. In structure and orientation for the first division, therefore, the bivalents of Puto conform to the coccid and aphid type, as analysed by Ris (1942). r. First mciotic division. Chromosomal fibers form from the entire poleward surface of each chromatid and converge toward the division center. The four wefts of fibers thus produced in relation with each bivalent are visible in the obliquely viewed, leftmost bivalent of figure 10. No centrioles are visible and the chromosomal fibers tend to fade out distally, but the pole is nevertheless well demarcated and the spindle as a whole is of normal form. As in somatic mitosis no continuous fibers nor astral rays are present. Anaphasic disjunction is normal and regular with no trace of differential rates among the bivalents or their constituent elements. Delicate interzonal connectives form between the separating chromosomal elements, but the small size precludes an analysis of their structure. d. Interkincsis: separation and secondary pairing oj chroniatids. Already in the telophase of the first division the two chroniatids derived from each metaphase bivalent begin to separate (Fig. 11). In the ensuing interkinetic interval this movement is continued until frequently the separation is complete and the full diploid number of chromatids may be counted as in figure 12. This separation is not interpretable as an extreme expression of that "repulsion" between chromatids characteristic of most organisms immediately prior to the second meiotic division. The chromatids here involved are not originally sister strands held together by a joint kinetochore region, but represent, again disregarding crossing-over, equational halves of the two homologous chromosomes of the meta- phase bivalent. Their separation thus indicates simply the lapse of the terminal attraction or association which held the homologues together after the terminaliza- tion of any chiasmata which may have been present — an association ordinarily broken at first anaphase. Little or no unravelling of the chromatids has thus far occurred. They remain throughout interkinesis as compact centers with only a slight irregularity of outline A PRIMITIVE COCCID CHROMOSOME CYCLE 171 (Figs. 12 and 13). The nuclear membrane now reforms, and it is of interest that therewith the chromatids assume once more, as previously in diakinesis and later in the spermatid nucleus, a peripheral distribution underlying the membrane. Simultaneously with this orientation the chromatids begin to reassociate in pairs (Fig. 13). Size differences show this pairing to be between homologous chroma- tids. Although the long axis of these compact chromatids cannot now be deter- <> B 8 FIGURES 7-10. Diakinesis and Metaphasc I in male. FIGURE 7. Diakinesis — (only four bivalents drawn) ; three bivalents show homologues assuming end to end position, one open cross. FIGURE 8. Diakinetic bivalents ; A — early assumption of end to end position of homo- logues ; B and C- — stages in opening of cross configuration. FIGURE 9. First meiotic metaphase. polar view ; six bivalents and univalent X ; con- striction in bivalents is primary split. FIGURE 10. Same, lateral view; constriction in plane of separation is secondary split. mined with accuracy, there is little doubt that the realignment results in a side by side lengthwise, association. This assumption is supported by the close parallellism obtaining between the realignment here and in the corresponding chromatids of Nautococcus ( Hughes- Schrader, 1942) in which the long axis is persistently recognizable. Moreover, as the newly formed dyads orient for the second metaphase the plane of contact between the homologous chromatids comes to lie at right angles to the spindle axis and forms the plane of separation for the 172 SALLY HUGHES-SCHRADEK second division. Chromosomal fibers then form from the entire poleward surface of each chromatid further identifying this as the long axis. The second division is thus reductional for all non-crossover regions. It should be emphasized that the seriation of the interkinetic stages just described can be positively established. Cell and nuclear size are in series with those of first anaphase and telophase on the one side, and second metaphase on the other. Moreover, in the first telophase alone is a heavily staining midbody developed in the interzonal connectives (Figs. 11 and 12). This midbody is re- tained, with decreasing sharpness of staining reaction, in the interkinetic cells and thus confirms their identification. c. Second inciotic division. The spindle for the second division resembles that of the first, but with more sharply delimited and acuminate ends. Again chromosomal fibers alone are formed, neither continuous fibers nor astral rays being present (Figs. 15 and 16). The sex chromosome comes to lie either on the edge of the equatorial plate, or more frequently nearer to one pole, but always close to the spindle. It shows no tendency toward division, produces no chromosomal fibers, and passes usually in advance of the autosomes undivided to one pole. Second telophases show the expected two categories of spermatid nuclei — those with six autosomes and the X (Fig. 17) and those with six autosomes only (Fig. 18). /. Quadrinucleate spermatids. With the formation of a membrane around the spermatid nucleus the chromo- somes again assume a peripheral position under it (Fig. 19). A cytoplasmic fusion of the spermatid cells in groups of four now takes place. Each resultant quadrinucleate spermatid contains, invariably, two of the six-chromosome and two of the seven-chromosome nuclei (Fig. 19). It follows that the four nuclei involved thus almost certainly represent the products of a single primary spermatocyte. The short centrally directed stalk often visible in the body of the quadrinucleate cell further suggests that the meiotic divisions may not have been quite complete cytoplasmically. If the products of the two divisions retain a connection through a common stalk a limiting factor in the fusion would be provided. Such a mech- anism, it will be recalled, has been demonstrated in other coccids (F. Schrader, 1931, Hughes-Schrader, 1931). Later stages in sperm formation show the pro- gressive development of all four components of the quadrinucleate spermatid. There is no evidence of any degeneration or loss of nuclei and apparently four normal sperm are formed from each quadrinucleate spermatic!. COMMENT ( ytologically I'nlo sp. stands out as a persistently primitive type among coccids thus far studied. This is evident in the absence of hermaproditism and of parthenogenesis and in the retention by both male and female of a normal meiosis. Its relatively generalized chromosome cycle is most nearly approached by the more primitive species of the llaveiine tribe of the family Margarodidae. Taxonomically the Margarodidae and the ( Mheziidae constitute the most primitive subdivision of existing coccids; they are set off from all other families by such primitive traits A PRIMITIVE COCCID CHROMOSOME CYCLE 173 FIGURES 11-19. Intcrkincsis and second meiotic division in male. FIGURE 11. Late telophasc I with separation of chromatids underway; spindle rest of first division at lower left. FIGURE 12. Early interkinesis ; complete separation of chromatids. FIGURE 13. Reassociation of chromatids in pairs ; spindle rest usually present at this stage not included in section. FIGURE 14. Polar view of second metaphase. FIGURE 15. Lateral view of same; X chromosome close to spindle at equator. FIGURE 16. Early second anaphase ; X chromosome near spindle, off equator. FIGURE 17. Second telophase, with 6 autosomes and X chromosome. FIGURE 18. Same, with 6 autosomes only. FIGURE 19. Quadrinuclcate spermatid. with two nuclei showing 7 and two 6 chromosomal masses. 174 SALLY HUGHES-SCHRADER as the retention (with a few specialized exceptions) of abdominal spiracles in all stages and well developed compound eyes in the adult males. Their closest rela- tives among other coccids are to be found in the Pseudococcidae — of which Puto constitutes the probably most primitive genus — linking the pseudococcid and ortheziid stems (Morrison, 1928, and personal communication). The persistence in the family Pseudococcidae of so primitive a type of chromo- some cycle as that of Puto has especial interest in view of the highly specialized male meiosis of the other pseudococcids thus far investigated. These comprise several species of Pseudococcus (Schrader, 1921, 1923a and b), and Phcnacoccus ttccriccld (Hughes-Schrader, 1935). These are jointly characterized by a per- sistent heteropycnosis of one haploid set of chromosomes in the male, by segregation without synapsis, and by the degeneration of the spermatid nuclei derived from the heteropycnotic complement. Similar conditions are encountered in Gossyparia spuria of the family Kermidae (Schrader, 1929) and in Lccanhtin hesperidum and L. hemisphaericum of the Cocciclae (Thomsen, 1927, Suomalainen, 1940). While I'uto throws no light on the origin of these specializations, the existence of the XX - XO sex chromosome mechanism in a primitive pseudococcid is highly signifi- cant, indicating that the male is primarily the heterogametic sex in this group. Its presence alike in Puto and the primitive llaveiines may well mean that it also represents the primitive condition for coccids as a whole. Its loss, and the substitu- tion of alternative mechanisms — (haplo-diploidy in the leery ini and the as yet unsolved sex determining mechanism of Pseudococcus, Phenacoccus, Gossyparia, and Lecanium in which both sexes originate from eggs fertilized by one class of sperm only) — have occurred in all other forms thus far investigated. Homogamety of the female relative to sex is indicated by the fact that in all cases the eggs of diploid-parthenogenetic females and of self-fertilized hermaphrodites (basically female in constitution) give rise exclusively to females. The curious cytoplasmic fusion of spermatids in groups of four appears to be of very early origin in the coccid stem for it is found in every species thus far studied. Even the haploid males of the Iceryini with only one meiotic division retain the habit, producing binucleate spermatids. Multinucleate spermatids have been de- scribed in certain spiders by Wagner (1896). He reports variation within the individual and among species; the binucleate and quadrinucleate condition is frequent and higher multiples are occasionally encountered. Later authors have not dealt with the problem in detail but incidental observations (Wallace 1905, IVisenherg, 1905. and dickering and Hard, 1935) indicate that cell bridges < "niaining spindle remnants frequently persist between spermatids. Incomplete cytoplasmic division and multinucleate cells probably form the basis for certain of Warren's (1928, 1931) claims of amitosis in spider spermatogenesis. In the coccids no variation among species nor within the species or the individual has been observed. The fusion is always limited to the derivatives of each primary spermatocyte. The limiting factor appears to be the persistence, in the radially arranged cells of each cyst, of a centrally directed stalk from each spermatocyte— a stalk never completely severed during the meiotic cell divisions. A significant feature of male meiosis in Puto is the complete separation and subsequent realignment of the chromatids during interkinesis. As already pointed out, tin's separation breaks the terminal association between homologous chromo- A PRIMITIVE COCCID CHROMOSOME CYCLE 175 somes which persists throughout the first division, and the realignment side by side ensures segregation at the second division. This type of meiosis, with its characteristic and essential orientation of the hivalents at first metaphase, is found in all the more primitive of the llaveiine coccids (Llaveia, Llaveiella, and Nautococcus — Hughes-Schrader, 1931, 1940, 1942). Although the meiotic figures of the female Puto are too small for critical analysis, it is significant that in the females of Pscudococcus citri (Schrader, 1923a) and Lccanium hcsperidnni (Thomsen, 1927), which in contrast to their highly specialized males retain an otherwise orthodox meiosis, the same separation and realignment of chromatids for the second division take place. Its occurrence in the unspecialized Puto male further confirms the conclusion that this type of meiosis is a primitive character for the coccids as a whole. Ris (1942), who first pointed out the significance of these phenomena, presents convincing evidence that the same type of meiosis obtains in aphids also, and must thus have differentiated after the Sternorhyncha had sep- arated from the auchenorhynchous Homoptera. Incidentally it is of interest that the secondary pairing involved in this type of meiosis may well have played a role in the evolution of asynapsis among the llaveiine coccids. With the renewed operation of the pairing force just prior to the second division, asynaptic chromo- somes which have divided separately and equationally during the first division, are brought together briefly at second metaphase and undergo a normal segregation. Thus in the llaveiine coccids asynapsis has been free to evolve without its usual sequelae of meiotic irregularities. Secondary pairing, while completely independent in its origin, here incidentally operates as a mechanism stabilizing asynapsis. SUMMARY A primitive chromosomal cycle possibly archetypal for coccids is reported for Puto sp. of the family Pseudococcidae. There is no hermaphroditism nor any cytological evidence for parthenogenesis. The diploid chromosome number is 14 in the female, 13 in the male. Somatic mitosis is of the type characteristic for chromosomes with diffuse kinetochore. Meiosis is regular in both sexes. In the male it can be demonstrated to adhere to the coccid-aphid type, with: (a) the first division equational for non-crossover regions; (b) separation of chromatids and their secondary pairing during interkinesis, and (c) segregation of non-crossover regions in the second division. An XX -female, XO -male sex determining mechanism is present. Quadrinucleate spermatids are formed. This is the only coccid thus far reported with a simple and orthodox meiosis in both sexes. LITERATURE CITED BOSENBERG, H., 1905. Beitrage zur Kenntniss der Spermatogenese bei den Arachnoiden. Zool. Jahrb., 21 : 515-570. CHICKERING, A. M., AND W. HARD, 1935. Notes on the spermatogenesis of spiders. Papers Mich. Acad. Sci., Arts, Letters, 20: 589-595. HUGHES-SCHRADER, S., 1931. A study of the chromosome cycle and the meiotic division figure in Llaveia bouvari — a primitive coccid. Zeitschr. Zellf. mikr. Anat., 13 : 742-769. HUGHES-SCHRADER, S., 1935. The chromosome cycle of Phenacoccus (Coccidae). Biol. Bull., 59: 462-468. HUGHES-SCHRADER, S., 1940. The meiotic chromosome of the male Llaveiella taenechina Morrison (Coccidae) and the question of the tertiary split. Biol. Bull., 78: 312-337. 170 SALLY HUGHKS SCHRADER HUGHES-SCHRADER, S., In42. The chromosomes of Xautococcus schraderae Vays., and the meiotic division figure of male llaveiine coccids. Jour. Morph., 70: 261-299. MOKKISOX. II., 192S. A classification of tlie higher groups and genera of the coccid family Margarodidae. Tech. Hull., ('. S. Z'r/1'- Agric., 52. Ris, H., 1942. A cytologiral and experimental analysis of the mciotic behavior of the univalent x chromosome in the hearberry aphid Tamalia ( = Phyllaphis) coweni (Ckll.). Jour. J-.rp. ZooL, 90: 267-330. SCHRADER, F., 1921. The chromosomes of Pseudococcus nipae. Biol. Bull., 40: 259-270. Si'HKAUKk. F., 1923a. The sex ratio and oogenesis of Pseudococcns citri. Zcitschr. ind. Absi. u. Ver., 30: 164-182. SCHRADER, F., 1923b. A study of the chromosomes in three species of Pseudococcns. Arch. Zcllf., 17: 45-62. SCHRADER, F., 1929. Experimental and cytological investigations of the life-cycle of Gossy- paria spnria (Coccidae) and their bearing on the problem of haploidy in males. Zcitschr. wiss. Zoo!.. 134: 150-179. Si uKAUKR, F., 1931. The chromosome cycle of Protortonia primitiva (Coccidae) and a consideration of the meiotic division apparatus in the male. Zcitschr. u'iss. ZooL, 138 : 386-408. SUOMALAIXEN, E., 1940. Bcitrage zur Zytologie der parthenogenetischen Insekten II. Lecanium hemisphaericum (Coccidae). Ann. Acad. Sci. Fcnn., 57: 3-30. THOMSEX, M., 1927. Studien iiber die Parthenogenese bei einigen Cocciden und Aleurodiden. Zcitschr. Zcllf. tnikr. Aunt., 5: l-llo. WAGNER, J., 1896. Beitrage zur Kenntniss der Spermatogenese bei den Spinnen. Arbeit. Kais. Xaturf. Gcs. St. Petersburg, 26 : 81-98. U'ALLACE, L. B., 1905. The spermatogenesis of the spider. Biol. Bull, 8 : 169-188. \\~ARREN, E., 1928. The comparative histology of the testis and the origin of the spermatozoa in certain South African spiders. Ann. Natal Mus., 6 : 1-88. \\"ARREX, E., 1931. The multiple origin of spermatozoa from spermatids in certain South African spiders. Ann. Natal Mus., 6: 451-458. ECOLOGICAL OBSERVATIONS ON TWO PUERTO-RICAN ECHINODERMS, MELLITA LATA AND ASTROPECTEN MARGINATUS ROMAN KENK Department of Biology, University of Puerto Rico I. Mellita lata H. L. Clark In 1941 Dr. Henry van cler Schalie collected two tests of keyhole urchins (Mellita) on the beach near Loiza Vieja, Puerto Rico. The tests were later ex- amined by Dr. Hubert Lyman Clark,1 who recognized them as Mellita lata H. L. Clark, a' species recently described by him (Clark, 1940) and known previously only from two localities — Puerto Limon, Costa Rica, and La Mancha, Veracruz, Mexico. According to Clark' (1933) two species of Mellita occur in Puerto Rico— M. quinquiesperforata (Leske) and M. sexiesperjorata (Leske). The latter, how- ever, is now placed in the genus Leodia Gray. The records of M. quinquiesper- forata are all based on the collections of the "Fish Hawk" expedition which took a total of ten specimens at Ponce, Arroyo, Mayagriez, Puerto Real, and in San Juan Harbor. The collection was studied by Clark (1901) and the specimens were as- signed to the species Mellita testudinata Klein (a synonym of M. quinquiesper- forata) which name was, at that time, used for all five-lunuled members of the genus Mellita from the eastern coasts of the Americas. Clark's recent revision (1940) of the genus Mellita segregates several new forms from the old group of M. quinquiesperforata. In the light of this critical study, the Puerto-Rican form is now to be transferred to the species M. lata. This species ranks among the most common echinoderms of Puerto Rico. It occurs on sandy beaches along the entire circumference of the island. In addition to the localities listed by Clark (1901, p. 254), it has been found in the following places : (1) Beach east of the mouth of Herrera River, east of Loiza Vieja, P. R. Several tests on the sand (coll. H. van der Schalie and the author). (MCZ No. 7972-}. (2) Beach about l/, mile west of Punta Embarcadero, northwest of Luquillo, P. R. Very numerous in shallow water. (MCZ No. 7997). (3) Beach about ^ mile southwest of Punta Santiago, Playa de Humacao, east of Humacao, P. R. Dry tests on shore and living animals in shallow water. (MCZ No. 7984). (4) Beach at Las Mareas, 4 miles southwest of Guayama, P. R. (or 1 mile 1 This study was begun at the suggestion of Dr. Clark, who desired to obtain additional specimens of this seemingly rare form from Puerto Rico, and asked me to gather more information on its distribution and ecology. I am grateful to Dr. Clark for his continued interest in this work. 2 Collection of the Museum of Comparative Zoology, Harvard University. 177 178 ROMAN KI-.XK northeast of Puntu Ola Grande). Numerous specimens, close to shore (coll. Gloria Fernandez). (MCZ No. 7998). (5) Beach near Central Boca Chica, Barrio Cintrona, about 6 miles east of Ponce, P. R. Four specimens, rather fresh, dead on the beach. (MCZ No. 7986). (6) Playa dc- Muni, 3 miles north-northwest of Mayagiiez Harbor, P. R. Five fresh tests on the sand. (MCZ No. 7988). (7) Beach at Punta Cadena, 6t/-> miles west of Ariasco, P. R., in shallow water (coll. Carlos F. Blanco). (MCZ No. 7990). (8) Columbus Park, % mile south of Aguadilla, P. R. Very numerous in shallow water, 1% to 2 feet. (MCZ No. 7989). (9) Catano Beach in San Juan Bay, 2 miles southwest of San Juan, P. R. (MCZ No. 7999). (10) Isla Verde, east of San Juan, P. R. Three specimens from the collection of the Department of Biology, University of Puerto Rico. (MCZ No. 7973). In several of these localities, the animals are exceedingly numerous. This may be said, in particular, of Luquillo Beach where they were found to be most abundant close to shore, just below the zone of moving sand, at depths of from one to three feet. Up to 16 animals were counted in a square-foot area. The people of the island coast, including fishermen, pay little attention to the animal and have no particular name for it. In two places, Humacao and Aguadilla, I heard them referring to Mellita as "estrella" which means star and is also the name used generally for sea-stars. Apparently the use of the name is due to the radial pattern of the oral surface. A general description of the morphology of Mellita lata was given by Clark (1940, pp. 437-438, and pi. 60, fig. 1 ; pi. 61, fig. 1 ; pi. 62, figs. 1,2). The species is characterized mainly by its elliptical shape, the width exceeding the length con- siderably ; by the anterior situation of the apex ; by the dimensions and the shape of the lunules ; and by the large heads of the capitate aboral primary spines. Color of the living animal. — The aboral surface is dark grayish olive-green. The oral side (Fig. 2) has a remarkable color pattern. In animals from Luquillo, the ambulacral areas are usually dark wine-red, occasionally dark purple, or more rarely a lighter shade of red. Lighter (pinkish), narrow, somewhat branched bands radiate from the peristomial margin towards the inner ends of the paired lunules (I, II, IV, and V) and similarly in the anterior midline (III) towards the anterior margin. More irregular light patches and stripes, extending in a trans- verse direction, occur on both sides of the unpaired lunule. In specimens taken at Aguadilla, however, the ambulacral areas were brown, the shade ranging from deep yellow-brown to red-brown, and the lighter bands and patches were in a light brown hue. The interambulacra are covered with silvery, translucent spines and appear whitish or light pink. The color of the dark ureas of the oral surface is due mainly to the coloration of the numerous tube feet and that of the periproct with the anal tube. In other places, the epidermis has a light pink or yellowish color. After the spines and the epidermis of a fresh specimen are removed, the following FIGURE 1. Behavior of Mcllita lata meeting obstacles while digging into the sand: (a), successive positions at running against stick placed vertically in the path of movement, to one side of the midline ; (fc), stick placed vertically in the center of path; (c), vertical board placed transversally to the direction of movement; (d), vertical board placed obliquely in the path of movement. ISO ROMAN KI-XK color pattern is seen on the exposed te>t : Aboral side — blue-green; oral side— amhulacral areas brown and tin- interambulacra white, pink, or light yellow. The dark coloring matter of the oral side is extracted by the common preservatives (alcohol, diluted formalin) in a very short time. The anus is situated as the end of a short tube (about three millimeters long in fully grown animals) and is surrounded by small, irregular papillae. The tube feet are closely packed in the ambulacral areas, except in the radial bands of lighter color. Moreover they are found along the margins of all lunules, including the posterior or interambulacral lunule, and also among the marginal spines of the entire circumference. The oral side has three kinds of spines which may be roughly grouped as follows: Long, slender spines of the interambulacra (these are the ones responsible, together with the marginal spines, for the locomotion of the animal) ; medium-, sized spines scattered over the ambulacral areas ; and short spines of the ambulacral surfaces, particularly flanking the ambulacral furrows. No pedicellariae were seen. Ecology. — The animals live buried in the uppermost layer of the sand in such a manner that usually only the posterior lunule, with a small part of the posterior surface and of the posterior margin, is visible. Occasionally, also lunules I and V may be exposed. Only exceptionally is the contour of the entire animal discernible. They move continuously through the sand, their speed varying from 11 to 26 millimeters per minute. Experiments on locomotion. — Several individuals were placed on a layer of sand in a flat pan filled with sea water. At first the animals remained quiet and no move- ment of the marginal spines was seen from the aboral side. After a short time— a few seconds up to perhaps one minute — they began to move forward, first in small jerks and later in continuous movement. They dug obliquely into the sand and disappeared from the surface within a short time, in from one to about four minutes (see Table I and Figs. 4a to 4c ). TABLE I Time required by Mellita lata to dig in completely. Temperature, 28° C. Time required Specimen Length in millimeters First trial Second trial A 25 1 min. 20 sec. 1 min. 30 sec. B 26 1 min. 25 sec. C 36 2 min. 30 sec. 1 min. 45 sec. I) 53 2 min. 05 sec. E 60 2 min. 40 sec. F 70 3 min. 20 sec. 3 min. 20 sec. G 71 3 min. 05 sec As the animals move in a forward direction, they need a "runway" of a certain length to dig in successfully. Two specimens placed in a jar with a diameter of about 90 millimeters, were moving around continuously, for hours, but were always on the surface of the sand. During the digging proces>, the animals do not react to such mechanical TWO PUERTO-RICAN ECHINODERMS 181 > I FIGURE 2. Mcllitn htta H. L. Clark, oral sick-, natural size. FIGURE 3. Astropcctcn wtiryiinitus Gray, aboral side, % natural size. L82 ROM AX KKNK stimuli as tapping the exposed surface with a wooden stick, hut continue their locomotion without interruption and at an even rate. Tt the tapping is so strong as to dislodge the animal from its hold in the sand, the movement is stopped for a short time, to he >ooii again resumed in the usual way. The following four experiments have been conducted repeatedly to determine the behavior of a digging animal meeting obstacles placed in its way: (a i If a >tick is placed vertically before the animal, to one side of the midline (Fig. la), locomotion is almost stopped when the animal reaches the obstacle, but the rhythmical movements of the spines continue. Slowly the animal works it> way around the stick keeping its original orientation all the time {i.e., the antero-posterior axis remains parallel to the original direction of motion as the animal shifts to the right or left, depending on the location of the obstacle with respect to the midline). i /• ) When a stick is placed vertically at the center of the frontal margin ( Fig. I/1), the animal at first stops its locomotion, but continues to move its spines. Within a short time, the pushing of the spines of one side — the right or the left - prevails and the animal turns slowly towards the weaker side. It then continues to move with an orientation of the axis which is at a slight angle to the original direction. ( c i If a vertical board is placed transversely in the path of movement ( Fig. lr), the animal continues the movement of the spines upon reaching the plane. Slowly it turns to the right or left, for approximately 16 degrees. Then it continues moving laterally along the plane, at a very slow rate, but retains the axis constantly in the new < >nentation. (d) \\hen a vertical board is placed obliquely (at an angle of less than 74 degrees) in the path of the movement (Fig. !42. p. 150) mi Mcllita qiiinquiesperforata (Leske) at Beaufort, North Carolina, indicated a hehavior rather different from that exhibited by our species: "The sea urchins. Moira and Mellita. also move almost directly downward by rapidly waving their spines and tube feet so as to move sand from underneath their tests toward the w^' d FIGURE 5. Photographs ol .Istrcpcctcn imiranclla tincta used were collected by Dr. A. I).'i\\-Mii in a pond containing much Elodca in the vicinity of New York City and shipped to Woods Hole where they lived well in the laboratory for more than a 188 FOOD-VACUOLK IX PKRITRICHA 189 week. Nearly all were attached to the stems and leaves of Elodca and most of them were single. Campanclla umbellaria was found in abundance in a shallow ditch in a peat marsh adjoining a small lake known as Sol's pond about one mile northeast of Falmouth. Mass. The water in this ditch was covered with duckweeds (Lcnina) and was distinctly acid (pH 6.2) but clear. Ehrenberg (1838) and Greeff (1870-71) called this organism Epistylis flaracans. The two species of Campanclla studied were practically the same in form, gross structure, size, formation of colonies and behavior, but the former had six double rows of cilia on the peristome, was grayish in color, owing to numerous conspicuous granules, and was found chiefly on Elodca, while the latter had only four double rows of cilia, was distinctly yellowish in color, had no conspicuous granules, and was found chiefly on Lctnna. Vorticclla similis \vas found in abundance attached to cluck-weeds in a pond which contained all sorts of refuse including much ashes. The water in this pond was continuously distinctly alkaline, usually pH 8.2. Most of the specimens used were, however, obtained from laboratory cultures. They thrive indefinitely in boiled tap-water containing crushed hemp seeds (two seeds in 50 cc. in a finger bowl) if the solution is renewed about once a week. In the pond they were always found very near the surface and in the laboratory they grew well only in shallow water. They apparently require an abundance of oxygen. STRUCTURE OF THE FEEDING APPARATUS Introduction All observers agree that the feeding apparatus in the Peritricha contains a ciliated tube which is connected with the peristome, that this tube consists of an outer part in which the cilia produce an ingoing and an outgoing current and an inner part in which they produce only an ingoing current, and that the fecal sub- stance and the content of the contractile vacuole are discharged into the outer part. There is, however, much variation in the names applied to these two parts and great diversity of opinion concerning the structure of the cytoplasm beyond the distal end of the inner part, as set forth in the following paragraphs. The outer part is called "buccal cavity" by some, "vestibulum" by others and "vestibule" by still others. The outer opening of this part is called "mouth" by some, and the inner opening is called "mouth" by others. The inner part is called "pharynx" by some, "cytopharynx" by others and "oesophagus" or gullet by still others. Some hold that it opens directly into the cytoplasm, others that it does not. We shall call the outer part "vestibulum," the inner part "pharynx," and the outer opening of the vestibulum "mouth." Ehrenberg (1838) concludes, on the basis of the direction of the movement of the food-vacuoles through the cytoplasm, that the pharynx opens into a tube or gut which extends through the cytoplasm to the anus in the wall of the vestibulum. Koehring (1930, p. 55) supports this conclusion. She could not see a differentiated tube in the cytoplasm but she says that the "orderly course" of food-vacuoles in Vorticclla sp., and other evidence, indicates that there is a "digestive system in ciliates, comparable to the digestive system of many metazoan organisms." Greeff (1870) could find no evidence of a digestive system in the peritricha but he main- 190 MAST AXD BOWEN FIG. 3 | UK 1. Camera outline of Camfanclhi unibcllaria. PC, peristome containing six double rows of cilia, not shown ; V , Vestihulum ; P, pharynx ; PR, pharyngeal ring ; OS, oesophageal sac ; OF, oesophageal fibers ; N , nucleus ; CV , contractile vacuole. FIGURE 2. Camera sketch of a portion of the feeding apparatus in Campanclla umb ell aria greatly compressed. P, pharynx ; PR, pharyngeal ring ; OS, oesophageal sac ; OP, oesophageal FOOD-VACUOLE IN PERITRICHA 191 tains that in at least some of them (Campanella nmbcllaria) the pharynx opens into a funnel-shaped structure ("der Trichter") which in turn opens into a tube ("Oesophagus") but that this tube opens directly into the cytoplasm near the posterior end of the body, not into the vestibulum near the anterior end. Schroder (1906) says he observed such a tube in Epistylis plicatilis and Vorticclla monilata as well as in Campanella ninbellaria and Kahl (1935) concludes that it is present in all the Peritricha. Greenwood (1894) and Kitching (1938) were however unable to find any indication of it in any of some ten species studied. Material and methods Observations were made on the structure of the feeding apparatus in all the species listed above but certain parts of it could be more clearly seen in the two species of Campanella than in any of the others. No difference was found in the feeding apparatus in these two species. They were consequently used indiscrimi- nately. Both contain so much opaque substance that their internal structure cannot be made out under normal conditions. It was found, however, that, owing to their tough elastic surface membrane, they can be greatly compressed without injury and that this greatly facilitates observations on their structure. The observations were made as follows : A small unattached colony in tap-water containing a little powdered carmine, was mounted under a cover-glass supported by two small parallel ridges of vaseline. Water was then very slowly removed with a strip of filter paper until the campanellae were compressed as much as desired. During this process they were closely observed under low and high magnification. In some preparations the organisms were fixed by drawing Schaudinn fluid and alcohol under the cover-glass. However, this did not facilitate observations on the structure. Compensating oculars (10, 15 and 20x), apochromatic objectives (10, 20 and 40x dry and 60x oil immersion, n. ap. 1.4), an achromatic condenser and a concentrated filament lamp, with ground glass ray-filter and an iris diaphragm, were used in all the observations. Results The results obtained are presented in Figures 1 and 2 and the following para- graphs : These figures show that the wall of the pharynx in Campanella is con- siderably thicker at the distal end than elsewhere and that attached to this end there are several fibers which converge as they proceed and soon form a bundle which extends through the cytoplasm nearly to the posterior end of the body. The thickened end of the wall of the pharynx forms a definite ring which is highly refractive and distinctly yellowish in color. We have designated it the pharyngeal ring and the fibers attached to it, the oesophageal fibers (Fig. 2). The oesophageal fibers can be seen near the ring only under occasional circumstances and then not fibers; CV ', contractile vacuole; F-i-F3, food-vacuoles, showing change in shape and size; W , membrane at the surface of the body. FIGURE 3. Camera outline of an optical section of Vorticella similis. V. vestibulum; P, pharynx ; PR, pharyngeal ring ; OS, oesophageal sac ; OF, oesophageal fibers ; Ft - F5, food vacuoles, showing change in form and size ; small dots, bacteria and granules ; large dots, yeast-cells (The body contains numerous food-vacuoles and granules not represented). H>2 MAST AND I'.OYYKX very distinctly. They can however be seen definitely in the bundle but they cannot be clearly differentiated because they are superimposed and close together, seven were definitely seen in one bundle and three to five in others. There probably are a few more than" seven and they probably are equally spaced in their attachment to the pharyngeal ring. In some specimens the bundle was spread out considerably at the end forming a brush. In specimens which have been compressed and killed under a cover-glass the oesophageal fibers remain intact for several days if the preparation is sealed with vaseline and kept in a damp chamber and they do not decrease appreciably in distinctness for at least two days. No activity was seen in the oesophageal fibers except in one specimen. This specimen was greatly compressed. The cytoplasm in it had gathered around an irregular cavity at the end of the pharynx. The cilia in the pharynx were still active and were forcing fluid into this cavity, which was abnormally large so that the oesopageal libers were much distorted in their arrangement and in their connection with the pharyngeal ring. Three of these fibers, only slightly separated from each other, extended through this cavity near one side and then joined the rest in the bundle. Waves were definitely seen to pass synchronously along these three fibers, from their attachment to the ring, on into the bundle. This activity continued, however, only a few moments after which the entire organism appeared to be dead. Numerous attempts were made to reproduce the conditions under which this was seen but without success. In one specimen, however, in which an irregular cavity had formed at the end of the pharynx, six inactive oesophageal fibers were seen to extend from the pharyngeal ring through the cavity. The physiological state necessary for activity in these fibers probably continues such a short time after the campanellae are compressed that it is rarely encountered. Oesophageal fibers were seen in all the other species studied and a pharyngeal ring in several. The fibers were fairly distinct in Vorticclla snuilis (Fig. 3). Kpistylis plicatilis and Ophrydium cctatitin (Mast, 1944), but they could not be counted with certainty in any of them, although seven were distinctly seen in one ophrydium and five in one vorticella. There doubtless are more, probably about ten. Numerous specimens of Vorticella similis were fixed (some in hot Schaudinn and others in hot Bouin fluid) stained with Heidenhain haematoxylin, and sec- tioned (3, 5, 7 and lO/*). Those fixed in Bouin fluid were much better than those fixed in Schaudinn, but the oesophageal fibers could not be as distinctly seen in either as in living specimens. There was no indication of an oesophageal tube in any of the species studied. If there actually is such a tube the fibers observed must be in its wall. There is, however, considerable evidence (presented later) which opposes this supposition. There is, then, in the results obtained no support for the views of Ehrenberg and Koehring or Greeff and Schroder presented above. Fibers extending from the pharynx have been seen by Schuberg (1890) in S tent or, Sharp (1914) in Diplodinium, Andrews (1923) in FoUicullna and Bozler ( 1^24) and Lund (1941) in Parauiccinm. Schuberg and Andrews maintain that the fibers are in the wall of an oesophageal tube. Sharp, Bozler and Lund main- tain that they extend directly through the cytoplasm. The views concerning their function vary greatly. FOOD-VACUOLE IN PERITRICHA 193 FORMATION AND MOVEMENT OF THE FOOD-VACUOLES It is well known that in the peritricha the food-particles aggregate at the distal end of the pharynx, but opinions differ as to how the food-vacuoles are formed and transported through the cytoplasm. Numerous observations were made on the process of feeding in many speci- mens of Campanclla iinibellaria. and tlncta and Vorticclla siinilis under various conditions, and on a few specimens of each of the other species listed above. The results obtained in the observations on Campanclla led to the following conclusions : When the organisms are not feeding there is at the distal end of the pharynx a cone-shaped space filled with culture fluid and particles suspended in it. At the surface of this space there is a membrane in the form of a cone-shaped sac which we shall call the oesophageal sac (Fig. 2). This membrane is doubtless produced by the interaction between the fluid in the space and the adjoining cytoplasm. Pharyngeal cilia project into the sac and the oesophageal fibers pass from the pharyngeal ring over its surface to its apex where they unite to form a bundle which passes on into the cytoplasm. When feeding begins the pharyngeal cilia force more culture fluid and particles into the oesophageal sac. This stretches the membrane around it. but continuous interaction between the fluid in it and the adjoining cytoplasm prevents this membrane from becoming too thin. As the sac enlarges it becomes spindle-shaped, owing to unequal pressure of the oesophageal fibers and possibly the adjoining cytoplasm on different regions of its surface. Under normal conditions enlargement continues until the sac is nearly twice as wide as the pharynx, then a constriction begins to form near the pharyngeal ring. This constriction increases until a spindle-shaped portion of the sac is pinched off, leaving a cone-shaped portion attached to the pharynx, the same in shape and size as that which obtained before feeding began (Fig. 4 A-E). The spindle-shaped portion is a new food-vacuole. There is no perceptible change in size of the pharynx or the pharyngeal ring during this process. These structures are consequently not directly involved in the formation of the food-vacuole. The newly formed food-vacuole moves rapidly through the cytoplasm to the distal end of the oesophageal fibers. Here it remains a few moments, usually turning sharply, then it proceeds slowly with the cytoplasm on an indefinite course, ending in the lower part of the vestibulum where its indigestible content is discharged. Its slow movement is obviously due to the movement of the cytoplasm in which it is suspended, i.e. to cytoplasmic streaming, but during its rapid movement definite currents are produced in the adjoining cytoplasm, showing very clearly that this movement is not due to cytoplasmic streaming. The constriction in the oesophageal sac is probably due to simultaneous inward pressure, in the same region, of the oesophageal fibers on its surface ; and the food- vacuole is probably transported from the pharynx to the posterior end of the body within the bundle of oesophageal fibers by waves passing synchronously along these fibers and from the posterior end of the body to the vestibulum by streaming movement in the cytoplasm (cyclosis). The formation and transportation of the food-vacuoles in Vorticclla similis and all the other species studied is in full harmony with this description. In all, the food-vacuole is formed by pinching off a portion of a cone-shaped sac attached to the pharynx and in all the food-vacuole is spindle-shaped and passes rapidly 194 MAST AND BOWEN FIG. 4 I'i<,rRE 4. Outlines showing a portion of the feeding apparatus and the formation and movement of food-vacuoles in (.'ani^ouclla. A and ./,, feeding apparatus in a specimen not ing or immediately after a food-vacuole has been formed; B-E, successive stages in the nation <>t a food-vacuole under normal conditions; B^-E,, same in a compressed individual with jx ristome closed and its cilia inactive. P, pharynx; PR, pharyngeal ring; OS, oesophageal sac; 0 '.phagel fibers; F, food-vacuole. Note that under normal conditions the cone shaped oesophageal sac enlarges greatly and FOOD-VACUOLE IN PERITRICHA 195 through the cytoplasm to the posterior end of the body and then slowly with the cytoplasm through the body. No evidence of an oesophageal tube was observed in any of these species. These conclusions and others are strongly supported by the results obtained in detailed observations on variations in the formation and the movement of food- vacuoles in several specimens. These observations are considered in the following paragraphs : 1. In a specimen of Vorticclla siinilis mounted in tap-water but not compressed, it was observed that the food-vacuoles had, immediately after they were formed, a long projection at one end. One of these vacuoles was continuously studied under the oil-immersion objective during the entire process of formation and for some time after, and the following observed : The constriction in the oesophageal sac did not completely separate the food- vacuole from it. When the vacuole moved away this connection was drawn out until it had formed a strand fully as long as the vacuole ; then it broke at the apex of the sac. The vacuole with this strand attached now moved rapidly to the posterior end of the body, then turned sharply ; after which the strand folded over, came in contact with the surface of the vacuole, and fused with it ; then the vacuole moved on slowly and very slowly rounded up. The membrane on the surface of this vacuole appeared to be very thick and viscous. In other specimens of this species under the same conditions, but with powdered carmine added to the tap-water, some of the food-vacuoles remained spindle-shaped for at least one hour after they had reached the posterior end of the body and in some specimens of Ophrydium ectatmn more than two hours, whereas they ordi- narily round up in a few moments. Obviously either the membrane at the surface of these vacuoles was thicker and more viscous than ordinarily or their entire content was more viscous, probably the latter. The results presented above show that constriction in the oesophageal sac is not the only factor involved in the formation of the food-vacuoles, that is, that in con- nection with this constriction there must be a mechanism which forces the vacuole toward the posterior end of the body so as to stretch out and break its connection with the oesophageal sac. They also show that the membrane at the surface of the food-vacuole is formed while it is still a part of the oesophageal sac, not after it has reached the posterior end of the body as some maintain. They show, moreover, that the membrane at the surface of the vacuole varies greatly in thickness and in viscosity and that the entire content of the vacuole probably also varies greatly in viscosity. 2. A specimen of Cainpanella uinbcllaria was greatly compressed and then con- tinuously observed under the oil-immersion objective. The peristome was inverted and the cilia on it were inactive but those in the vestibulum and the pharynx were active and food-vacuoles were formed at intervals of about 45 seconds ; but after nine had been given off all ciliary action ceased. All these vacuoles were spindle- shaped, but much smaller and relatively much longer than those formed under becomes spindle-shaped, that a portion of this sac is constricted off to form the food-vacuole, and that the constriction begins at the base of the sac near the pharyngeal ring ; but that under abnormal conditions the sac enlarges but little, that only a small portion is constricted off and that the constriction begins near the tip of the sac. Under both conditions the formed food-vacuole usually moves rapidly to the end of the oesophageal fibers. 196 MAST AXI) BOYYF.X normal conditions. The minor axis of the first one formed was about half as long as the diameter of the pharyngeal ring and that of the last one not more than one-sixth; whereas it usually is nearly twice as long in normal food-vacuoles. In their formation the oesophageal sac enlarged slightly, then a constriction appeared near its apex and soon a small portion of the sac was pinched off (Fig. 4 A^-E^). This passed rapidly to the posterior end of the body then almost immediately rounded up. after which it moved slowly, decreased rapidly in size and seemed to disappear entirely. There were no visible particles in any of these food-vacuoles. Ciliary activity in the pharynx was seen in nearly all the compressed cam- panellae examined, but food-vacuoles formed in only a small percentage of them. In all but a few of these the formation of food-vacuoles ceased immediately after ciliary activity in the pharynx had ceased and in these few only one vacuole formed after this. These results indicate that the enlargement of the oesophageal sac is dependent upon activity of the cilia in the pharynx but not upon activity of those on the peri- stome, and they show that the formation of the food-vacuole is not specifically dependent upon ciliary action in the pharynx or the size of the oesophaeal sac or the presence of particles in suspension in the fluid in it. 3. In a compressed specimen of Catnpanella tincta five small food-vacuoles were formed in succession and rapidly transported to the posterior end of the body ; then there suddenly occurred a very violent upheaval in the cytoplasm, after which a large food-vacuole was formed and transported, but very slowly and only a short distance, after which it turned sharply, nearly stopped moving and soon rounded up. Two more large vacuoles were formed after this and these also moved slowly and only a short distance, then stopped and rounded up. The large food-vacuoles were more than 20 times as large as the small ones ; they moved much more slowly than the small ones and not more than half as far before they stopped and rounded up.1 Similar results were obtained in observations on several other compressed specimens. In one of these a very large food-vacuole formed, slowly moved back a short distance, turned sharply in its course, rounded up and stopped. Then a very small vacuole formed and moved rapidly, past the large one, nearly to the posterior end of the body after which it moved very slowly and rounded up. This was followed by the formation of two more small vacuoles, both of which moved rapidly past the large one to the posterior end of the body (Fig. 5). The large vacuole was closely observed under the oil-immersion objective. It did not move appreciably but rapidly decreased in size and disappeared entirely in three minutes. The tact that some of these food-vacuoles went only about one-fourth as far as others before their rate of speed rapidly decreased cannot be understood on the assumption that they passed through a tube and were propelled by peristalsis in it. It can however, be readily understood on the assumption, postulated above, that their movement was due to the action of fibers which can move freely and are not fixed in their spacial interrelationship. 4. A specimen of I'orticclla siinilis was mounted in tap-water and the oesopha- geal sac measured at maximum size. Then the tap-water was replaced by distilled 1 IK- junior author asserts that in his observations on the effect ot" various chemicals on the ize "I 'he- loud vacuoles in I 'orliccllti siinilis, lie frequently saw very small "needle-like" vacuoles form. FOOD-VACUOLE IN PER1TRICHA 197 water and the sac measured again, after which the distilled water was replaced by 0.006 M lactose in distilled water and the sac measured once more. The averages obtained for the minor axis under these three conditions were respectively 10.5/x, ll.S/i and 12.5/x. i— CV FIG. 5 FIGURE 5. Camera outline of a portion of a compressed specimen of Campanclla tincta, showing differences in the size of successively formed food-vacuoles and difference in their direction and extent of movement. CV, contractile vacuole ; P, pharynx; PR, pharyngeal ring; OS, oesophageal sac; OF, oesophageal fibers ; Ft - F0, five food-vacuoles formed in the order given ; arrow, direction of movement. Note that the first vacuole in this series was very much larger than the rest and moved only a short distance before it rounded up and that the four succeeding small vacuoles passed the large one and moved much further before they stopped and became spherical. T7, was drawn. very soon after it has been formed, F-,~F- immediately after F:, had been formed, i.e. after Fi and F~ had decreased considerably in size. 198 MAST AND BOWEN During one of the measurements of the oesophageal sac in the lactose solution a foocl-vacuole in the adjoining cytoplasm suddenly fused with the sac and caused a vet-}- marked increase in its size, immediately after which a portion of it was constricted off as an abnormally large food-vacuole. Immediately before the fusion took place the minor axis was 12.5^ long and during fusion it increased to 16.5M. The fact that the food-vacuoles in the cytoplasm can fuse with the oesophageal >ac. strongly supports the conclusions reached above, namely that there is nothing in the nature of an oesophageal tube in these organisms and that the rapid movement of the food-vacuoles after they leave the pharynx is clue to the action of a mobile structure which does not have a fixed position in the cytoplasm and does not prevent direct contact between the vacuoles and the cytoplasm. Discussion Greeff (1870) long ago observed that the food-vacuoles in Cauipanella itni- bellaria pass from the pharynx toward the posterior end of the body much more rapidly than the adjoining cytoplasm and he concluded consequently that they are not carried by the cytoplasm. He maintains, as stated above, that there is a long tube ("der Oesophagus") which extends into the cytoplasm from a spindle-shaped structure ("der Trichter" ) at one end of the pharynx. He asserts that in the "Trichter" the food which has been forced into it by the cilia in the pharynx, is formed into small spindle-shaped masses, which pass rapidly through the oesopha- gus" into the cytoplasm and that a membrane then forms at the surface of each mass and thus produces a food-vacuole. He accounts for the rapid movement of the food-vacuoles by assuming that they are forced through the "Oesophagus" by waves of contraction in it, i.e. by peristalsis. Kahl (1935, p. 652) confirms Greeff in reference to the "oesophageal" tube in Cmnpanclla and concludes that such a tube is present in all the peritricha. He says : "Der Osophagus ist bisher meist iibersehen worden ; er scheint aber nach eigenen Untersuchungen nie zu fehlen, ist aber nur bei grosseren Arten gut erkennenbar." The evidence presented above indicates that Greeff is correct in his contention that at the end of the pharynx there is a funnel-shaped structure to which is attached a long narrow structure which extends into the cytoplasm, but it indicates that the latter is a bundle of fibers instead of a tube and that the former is a sac, a portion of which is separated off to form a food-vacuole, rather than a funnel in which the food is formed into spindle-shaped masses which become food-vacuoles after they have been transported through the tube into the cytoplasm. It also indicates that the food-vacuoles arc propelled from the pharynx to the posterior end of the body by the action of fibers, not by contraction in the wall of a tube. Kitching (1938, p. 87) recently observed the rapid movement of the food- vacuoles referred to above, in a considerable number of species in several genera but he found no evidence of an oesophageal tube in any of them. He concludes that the rapid movement of the vacuoles is due to waves of contraction, in accord with Gr<-' mention, but that the contraction is in the cytoplasm, not in the wall of a tube in it. lie says: "It is concluded that the food-vacuoles are propelled over the determined course (i.e. from the pharynx to the posterior end of the body] by contractions in the surrounding protoplasm." FOOD-VACUOLE IN PERITRICHA 199 It is obvious, however, that to propel a vacuole by contraction in surrounding protoplasm which is not fixed as it is in a tube, the viscosity of the protoplasm would have to be continuously lower in front of the vacuole than back of the contracting region. There is no evidence indicating that this obtains. Kitching's hypothesis consequently has no objective support. Nirenstein (1905), Gelei (1934) and others maintain that in Paramecium the food-vacuole is separated from the pharynx by the pressure of protoplasmic currents. Biitschli (1889, p. 1405) and Bragg (1935, 1936) contend that con- traction of the distal end of the pharynx is also involved. Lund (1941) holds that neither is involved and that the vacuole is separated from the pharynx by the action of fibers which are attached to the pharynx and extend for a considerable distance into the cytoplasm. These views will be considered in a later paper. THE INITIATION OF THE CONSTRICTION OF THE FOOD-VACUOLE FROM THE PHARYNX It is generally assumed that the initiation of the constriction of the food-vacuoles from the pharynx is correlated with the size of the enlargement at the end of the pharynx. For example, Hall and Nigrelli (1930) referring to Vorticella say: "After the basal portion of the gullet reaches a certain size, it is rapidly constricted from the rest of the gullet and then separated completely as a food vacuole." The fact, however, that (as demonstrated above in observations on Campanella) successively formed food-vacuoles sometimes vary enormously in size, shows that the initiation of their separation from the pharynx is only very superficially corre- lated with their size, if at all. Bozler (1924) maintains that in Paramecium solid particles are necessary for the formation of food-vacuoles and that such particles must come in contact with the membrane at the end of the pharynx before a food-vacuole begins to form. Bragg (1935) maintains that while contact of a large particle with the inner surface of the "vacuolar membrane" always causes immediate separation of the food-vacuole from the pharynx, it is not necessary. We have, in observations on Campanella and Vorticella, repeatedly seen food-vacuoles form which contained no visible particles and we have seen some of these vacuoles disappear in the cytoplasm so rapidly that very little, if any, digestion could have occurred. These facts seem to show that these vacuoles contained no solid particles, and consequently that solid particles were not involved in their formation. Moreover, Schewiakoff (1891) and Wallengren (1901) assert that they observed food-vacuoles form in solutions which were free from solids. Kitching (1938) observed that if Pyxidinium asclli is mounted in "1/16 to 1/8% agar" food-vacuoles form without ciliary action on the "disc" or in the "gullet." We have confirmed this in observations on Campanella. We also observed that there is no change in the size of the pharynx during the separation of the food-vacuoles from it. This separation is therefore not correlated with changes in ciliary action in the pharynx or with contraction in it. It will be demonstrated presently that the size of the food-vacuoles depends upon the chemical composition of the surrounding medium. This seems to show that the chemical composition of the solution in the food-vacuoles has something to do with their separation from the pharynx, but it in no way accounts for the enormous variation in size referred to above, which occurred with no variation in the surrounding medium. 200 MAST AXD BOYVKX \Yhat is it, then, that sets off the process which separates the food-vacuoles from the pharynx ? It is highly probable that waves start at fairly regular intervals in the pharyngeal ring and pass simultaneously down all the oesophageal fibers and that each of these sets of waves initiates a constriction in the oesophageal sac, if it contains sufficient fluid to make a constriction possible. If this is true, the size of the vacuole is correlated with the rate1 at which fluid is forced into the oesophageal sac by the cilia in the pharynx and the rate at which it leaves this sac by osmosis. If these pro- cesses and the interval between successive waves depend upon the composition of the Mirrounding fluid, the temperature and the physiological state of the organism, it accounts for the observed variation in the size of the food-vacuoles and the intervals between their formation. If the food-vacuoles are separated from the pharynx by waves in the oesophageal fibers, one would, moreover, expect to find the observed correlation between the location of the constriction on the oesophageal sac and the size of the vacuole and also the observed absence of a constriction when the sac is very small. There would still remain, however, the problem of the origin of the periodic waves. FOOD AND FEEDING The observations considered in this and the following sections were made on / 'orticclla siinilis as follows : Several small pieces of substance with vorticellae attached were mounted in pond-water or culture-fluid between two parallel ridges of vaseline on a slide. A cover-glass was then added and pressed down until the pieces of substance were much flattened, but not enough to interfere with the activities of the vorticellae. In such preparations the fluid could readily be changed as desired by applying a strip of filter paper to one edge of the cover-glass, and if the flow of fluid was con- tinued so as to provide sufficient oxygen any selected vorticella could be studied under low or high magnification as long as desired and the effect of various sub- stances on its activities ascertained. Vorticella feeds almost exclusively on bacteria, but all sorts of particles in suspension in the surrounding fluid are carried into the vestibulum in the currents produced by the peristomal cilia. Many of these are, however, immediately carried out again in the outgoing current produced by the cilia in one region of the vesti- bulum. Nearly all the rest and some gelatinous substance secreted by the peristome or the walls of the feeding apparatus, are forced through the pharynx into the oesophageal sac by the pharyngeal cilia.- There is, however, great variation in the kind of particles that are selected and ingested by different individuals in the same preparation and by the same individual at different times. Yeast-cells, e.g. are, at any given time, freely ingested by some individuals and rigidly rejected by others, and ireely ingested by a given individual at one time and rigidly rejected at another. It is well known that when the food-vacuole leaves the pharynx the concen- tration of particles in the fluid in it is usually very much greater than it is in the fluid which enters the vestibulum. Greeff (1870) maintains that the cilia in the pharynx In I'orliccllii mounted in distilled water or in lactose (0.05 M) in tap-water or in X.i< 1, this .gelatinous substance is very evident. It gclatcs as the vacuoles :crease \» minimum in sixe (probably owinj; to the increase in acidity) and then solates as It is highly probable that it is formed under all conditions, as it appears Iliciiliiw, judging from the results of observations made by Andrews (1923). FOOD-VACUOLK IX PERITRICHA 201 come in direct contact with the particles in it and force them through the fluid into the oesophageal sac and that consequently only a relatively small amount of water is carried in with the particles. Xirenstein (1905) and Bozler (1924) referring to Paroincciuin maintain that the pharyngeal cilia force almost nothing hut fluid into the oesophageal sac until it has hecome nearly maximum in size and then almost nothing hut solid particles until it is well filled with them. Both of these views would account logically for the relatively great concentration of solid particles in the newly formed food-vacuoles. However, the results of our extensive and detailed observations do not confirm either of them. We found that when the oesophageal sac begins to enlarge the concentration of solid particles in it is usually only slightly greater than in the fluid which enters the vestihulum, hut that as the sac enlarges, the concentration of particles in it usually increases greatly. Selective action of the pharyngeal cilia would account for the concentration of particles in the newly formed food-vacuoles, hut it would not account for the observed gradual increase in concentration in the oesophageal sac except on the assumption of gradual increase in selective ciliary action. This is, however, not at all probable. How then can the gradual increase in concentration be explained ? It will be demonstrated presently that after the food-vacuole is formed fluid usually leaves it rapidly, owing to difference in osmotic concentration of the internal and external fluids. It is consequently practically certain that fluid passes continuously from the oesophageal sac out into the cytoplasm as it enlarges. The increase in the concentration of the particles in the fluid in the oesophageal sac is therefore, in all probability, due to this loss of fluid. Moreover, Frisch (1937), in observations on Paminccinin, has demonstrated that fluid passes from the pharynx into the adjoining cytoplasm. If this obtains in Vorticclla, it accounts for the probable increase in the concentration of solid particles as the fluid in which they are suspended passes through the pharynx. The junior author, in his measurements of the food-vacuoles in Vorticclla in different solutions, repeatedly saw the oesophageal sac suddenly decrease in size and at times, especially in distilled water, alternately decrease and increase like "the pumping of a heart." In one specimen in 0.014 M NaCl the oesophageal sac gradually increased to 11.56/A in diameter, then suddenly decreased to 8.84/x. in diameter, then remained without further measurable change in size for 20 seconds and then left the pharynx. The decrease in size observed under these conditions was, however, doubtless due to the forcing of fluid from the oesophageal sac back into the pharynx, probably by pressure on the surface of the sac by the action of the oesophageal fibers. THE SIZE OF THE FOOD-VACUOLES AND THE TIME REQUIRED FOR THEIR FORMATION Introduction No detailed measurements have heretofore been made on the size of the food- vacuoles in the peritricha or the time required for their formation. The results reported indicate, however, that while there is much variation in different indi- viduals under the same conditions and in the same individual under different conditions, consecutive vacuoles do not vary much either in size or in the time required for their formation. Hall and Nigrelli ( 1930) imply, e.g. that in 202 MAST AND i:<>\VK\ Vorticclla sp. the food-vacuoles are fairly uniform in size and the "intervals" between their formation rather constant for a given individual. We measured many food-vacuoles and the time required for them to form in Vorticclla siinilis in various solutions. Some of the results obtained will be considered in the following paragraphs. A more extended account of the work will be presented in a subsequent paper by the junior author. Methods Several vorticellae attached to a fragment of Leinna, or to a short hair, were mounted in a drop of water between two parallel ridges of vaseline on a slide and covered with a cover-glass. The slide was then put on the mechanical stage of the microscope and a narrow strip of filter paper, long enough to reach over the edge of the stage, placed at one edge of the cover-glass between the ridges of vaseline. Then some of the solution to be tested was placed on the slide at the other edge of the cover-glass between the ridges of vaseline and more added as, owing to the action of the filter paper, it flowed through under the cover-glass. A specimen which extended from its attachment well out into the current of solution was now observed. After the vorticella had been subjected to this current for ten minutes and thoroughly adapted to the new solution, measurements under an oil-immersion objective were made by means of a stopwatch and an ocular micrometer, on a series of successive food-vacuoles, in reference to the time required for their forma- tion and their maximum size, i.e. the length of the minor axis, as they were about to leave the pharynx. Another solution was then passed through under the cover-glass for ten minutes, after which measurements were made on another series of successive food-vacuoles in the same specimen or in a different specimen in the same solution. This was repeated with still other solutions. Then the whole process was repeated with other specimens. The results obtained are presented in Tables I and II. TABLE I Time required to form food-vacuoles in Vorticel'a similis Time in seconds required to form each of seven consecutive vacuoles in each of six individuals, selected at random In pond-water In distilled water . a b c d e f 40 48 52 69 79 42 38 56 51 70 83 58 51 56 29 50 83 58 62 50 39 67 86 46 31 50 39 75 56 54 39 49 38 43 58 44 42 46 39 60 ? 35 Aver 43.3 50.7 41 62 74.1 48.1 Total average 45 61.4 FOOD-VACUOLE IN PERITRICHA 203 Results Table I shows that there was marked variation in the time required for the formation of consecutive food-vacuoles in all six vorticellae studied, that the time required varied much with the individuals under both conditions and that it was on the average much longer in distilled than in pond-water. The results presented demonstrate, therefore, that the rate of formation of food-vacuoles is much higher in pond-water than in distilled water. Table II shows that the successive food-vacuoles in each of the five specimens tested varied greatly in size in all the solutions used, but that the food-vacuoles TABLE II Variation in the size of the food-vacuoles in Vorticella similis and the effect of various substances on its size A, a specimen subjected successively to distilled and pond-water; a, b, c and d, four speci- mens, each subjected successively to the solutions indicated. The lactose, NaCl and CaC^ solutions were made with redistilled water and they were equal in osmotic concentration. All the measurements for each specimen in a given solution were made on successive vacuoles. Length in micra of minor axis at maximum size Dis- tilled water Pond- water Redistilled water Lactose 0.026 M NaCl 0.014 M CaCh 0.01 M Designation of specimens A A a b c d a b c d a b c d a b No. of vacuoles measured 5 5 8 9 9 9 8 9 9 5 9 6 4 6 5 5 Minimum . . 12.7 13 11.56 7.14 7.46 8.5 13.6 7.48 6.12 8.84 6.8 5.44 3.4 4.76 6.8 4.76 Maximum. 14.1 16 17 10.2 9.18 HU 17 10.1 8.16 1(1. SS 14.16 7.48 4.76 5.44 8.8 6.12 Average 13.4 14.1 14.6 8.3 8.2 9.9 15.2 8.7 7.3 9.9 11 6.3 4 5.2 7.3 5.4 Total average for a and b. . . . 11.26 11.75 9.12 6.35 Total average .... 10.12 10.13 7.36 in different specimens in the same solution and in the same specimen in different solutions varied even more. It shows that in the four individuals measured the average length of the minor axis of the vacuoles ranged from 7.3 to 15.2^ in the solution of lactose, from 4 to llju, in the solution of NaCl and from 5.4 to 7.3/x, in the solution of CaCL. It indicates that the vacuoles were on an average slightly larger in pond-water than in distilled water, the same in size in redistilled water and the solution of lactose, much smaller in the solution of NaCl, and the smallest in the solution of CaCl2. Discussion The osmotic concentration of the solution of lactose used was obviously much higher than that of the redistilled water. The fact that the food-vacuoles formed 204 MAST AND HOWEN in these two fluids were practically the same in size indicates, therefore, that osmotic concentration is not involved in regulating the size of the vacuoles. The solutions of lactose, sodium chloride, and calcium chloride used were equal in osmotic concentration and in acidity. The differences in the size of the food- vacuoles formed in these solutions were therefore not correlated with either of these two factors. They consequently must have been correlated with the chemical properties of the substances in the solutions. The hydrogen-ion concentration of the distilled water used was pH 5.5 and that of the pond-water pH 8.2 : the osmotic concentration of the latter was much higher than that of the former and they differed greatly in chemical composition. The results referred to above indicate that the size of the food-vacuoles is not specifically correlated with the osmotic concentration or the acidity of their contents. The difference in the size of the vacuoles observed in pond-water and distilled water was therefore not due to either of these two factors. It consequently must have been due to difference in the chemical composition of their contents. The results in hand seem to show therefore that the size of the food-vacuoles in 1'orticclla is largely, if not entirely, dependent upon the nature of the chemicals they contain. As stated above the rate of formation of food-vacuoles is higher and the vacuoles are larger in pond-water than in distilled water. The rate of ingestion of fluid is therefore higher in the former than in the latter, but since these fluids differ greatly in acidity, osmotic concentration, and chemical composition, the difference in the rate of ingestion may be due to any one or any combination of these factors. We are well aware that some of the results presented in this section are equi- vocal, and that more results are needed before valid conclusions concerning the regulation of the size of the food-vacuoles and the rate of ingestion can be reached. We had intended to extend the observations made and to investigate the effects of other chemicals in various concentrations, but other duties interfered and we see no prospect of continuing the work in the near future. We are therefore presenting these inadequate results with the hope of encouraging further work. CHANGES IN THE SIZE AND THE FORM OF THE FOOD-VACUOLES Introduction After the food-vacuoles have been separated from the pharynx thev move rapidly to the posterior end of the body on a definite course, as previously stated, then slowly on a very indefinite course to the vestibulum. They are spindle-shaped until they reach the posterior end of the body then they usually become spherical and gradually decrease in size to a minimum, remain so for about two minutes and then rapidly increase in size again (Fig. 6).3 Numerous measurements were tt-ith a stopwatch and an ocular micrometer on the time required for these changes and their extent. The following results were obtained: During the dec-reuse in size the particles in suspension frequently, but not always, Igregate near the center of the vacuole, leaving a clear space at the surface (Fig. 6), which n disappears, hut usually forms again when the vacuole begins to enlarge, after which the particles soon become equally distributed. FOOD-VACUOLE IN PERITRICHA 205 Change in form. — The time required for the change in the form of the food-vacuoles from spindle-shaped to spherical varies enormously. Under some conditions it occurs almost immediately after the vacuoles have reached the posterior end of the body. Under others it requires an hour or more and under still others it probably does not occur at all. The rate of change in form seems to be closely correlated with the viscosity of the content of the vacuoles. The particles in suspension in the fluid in the vacuoles which changed rapidly in form were invariably in violent Brownian movement, indicating low viscosity, whereas those in the vacuoles which changed slowly were often practically stationary, indicating high viscosity. In specimens which had ingested carmine granules or lactose (0.025-0.05 M) the change was consistently very slow. FIGURE 6. Camera outlines showing the separation of a food-vacuole from the pharynx in Vorticclla siinilis and subsequent changes in its size and form. P, pharynx; PR, pharyngeal ring; OS, oesophageal sac; /•",, a food-vacuole in the process of separation from the oesophageal sac ; I'--Ff,, subsequent stages in the food-vacuole. Decrease in size. — The decrease in size is clue to loss of fluid. This ordinarily continues until there is no perceptible fluid left in the vacuoles and the surface membrane is in close contact with the mass of particles. Consequently, if the vacuoles contain relatively few particles, they decrease much more than if they contain many, and if the particles are large, yeast-cells, e.g. the surface becomes very irregular. Many of the vacuoles which were measured decreased three-fourths in diameter, i.e. to one sixty-fourth in volume. Under some conditions there is, however, still considerable fluid in the vacuoles when they have become minimum in size and under these conditions the shrinkage is obviously less. The reduction in size requires from one to three minutes. The loss of fluid, resulting in the decrease in the size, is probably due in part to difference in the osmotic concentration of fluids in the vacuoles and the cytoplasm 206 MAST AND BOWEN (that of the latter being higher than that of the former)4 and in part to inward pressure of the elastic membrane on the surface of the vacuoles, which was stretched by the pressure of fluid forced into them by the pharyngeal cilia. If the decrease in the size of the food-vacuoles is correlated with excess external osmotic concentration, the extent of change in size should decrease if the internal osmotic concentration is increased. This can readily be accomplished by adding physiologically neutral osmotic substance to the surrounding medium. Observations were consequently made on vorticellae in pond-water containing lactose in various concentrations, and the vacuoles formed in each concentration measured at short intervals after they left the pharynx. It was found that no vacuoles were formed in concentrations higher than 0.05 M, that in concentrations of 0.05 M and lower the vacuoles decreased in size and that they decreased least in the highest of these, i.e. 0.05 M, but that the decrease in this concentration was definitely less than in pond-water, the maximum being not more than one-third in diameter in place of three-fourths or more. It was also found that after the vacuoles had reached minimum size, those formed in the lactose solutions contained much more fluid than those formed in pond-water and were never irregular in shape, like many of those formed in pond-water. These results show that difference between internal and external osmotic con- centration is involved in the observed decrease in the size of the vacuole and they indicate that the inward pressure of the membrane around them is also involved. If this is true, it is obvious that fluid leaves the vacuoles, not only from the time they reach the posterior end of the body until they have become minimum in size, but continuously from the very beginning of their formation, for these two factors function in the oesophageal sac as well as in the vacuole, since the vacuole is, as previously demonstrated, merely a portion of the sac. In vorticellae in the lactose-pond-water solution, it was repeatedly observed that the food-vacuoles often coalesce with each other after the sudden increase in size and that owing to this and lack of elimination of undigested substance, the body became well filled with huge vacuoles. Coalescence of vacuoles was not observed in pond-water or culture fluid. The lactose must consequently produce changes in the vacuolar membrane which make it possible. Increase in size. — The vacuoles usually remain minimum in size for nearly two minutes, then very rapidly increase until they are nearly, if not quite, as large as they were originally, after which they remain fairly constant in size until their content is discharged into the vestibulum (Fig. 3). The increase in size requires on an average, a little less than three seconds. I hiring this time the vacuole is literally flooded with fluid from the cytoplasm. This fluid usually first appears as a well defined layer between the membrane and the viscous central mass, then this mass disintegrates and the solid particles in it soon become uniformly dispersed with violent Brownian movement throughout. Digestive enzymes are doubtless carried from the cytoplasm into the vacuoles with the fluid that enters, for digestion begins soon after the vacuoles have enlarged. In Parameciutn and some other ciliates numerous so-called neutral red bodies aggregate on the surface of the food-vacuoles. It is maintained by some that these * The osmotic concentration of the fluid in the cytoplasm, as will be demonstrated in the last section of this paper, is approximately 0.3 atmospheres higher than that of the surrounding medium. FOOD-VACUOLE IX PERITRICHA 207 bodies contain digestive enzymes, that they enter the vacuoles and are therefore involved in digestion. In the Peritricha there is no aggregation of such bodies on the food-vacuoles and there is no indication that any enter. They are consequently in all probability not involved in digestion in these organisms. The inflow of fluid, resulting in increase in size, is probably entirely due to greater osmotic concentration within the vacuole than without. If this is true, the internal osmotic concentration must increase greatly during the time that the vacuole remains minimum in size. This could readily be brought about by transformation in the vacuole of osmotically inactive to osmotically active substance, for example, starch to sugar. In food-vacuoles which contain lactose, the gelatinous substance in them, referred to above, increases greatly in viscosity as the vacuoles decrease in size (as indicated by observations on Brownian movement) and then decreases greatly as they increase in size. The increase in viscosity is correlated with increase in acidity (as will be demonstrated presently). It may well be that this increase in acidity causes chemical changes in the gelatinous substance which result in increase in osmotic concentration and that this in turn causes the rapid inflow of fluid from the cytoplasm which in turn, owing to decrease in acidity, causes the observed decrease in viscosity. CHANGES IN THE HYDROGEN-ION CONCENTRATION IN THE FOOD-VACUOLES IN VORTICELLA Introduction Numerous observations have been made by several investigators on the hydrogen-ion concentration in the food-vacuoles in various protozoa. Nearly all the results obtained indicate that as the food-vacuoles pass through the body the hydrogen-ion concentration first increases, then decreases and then remains nearly- constant. However, only a fewr of the observations concern the extent of these changes. Shapiro (1927) on the basis of changes in the color of indicator dyes, concludes that in the food-vacuoles in Paraincciitni the hydrogen-ion concentration increases to pH 4, then decreases to pH 7, that in Vorticdla it increases to pH 4.5. then decreases to pH 7 and that in Stylonlchia it increases to pH 4.8, then decreases to pH 7. Claff et al. (1941) using essentially the same methods conclude that in Bresslaua it increases to between pH 4.2 and 3 and then decreases (extent not given). Mast (1942) using similar methods and others found that in Amoeba it increases to pH 5.6. then decreases to pH 7.3. And Howland (1928) on the basis of results obtained by injecting dyes into the food-vacuoles, concludes that in Actinospliaeritdii it increases to pH 4.3 ±0.1 and then decreases to between pH 5.4 and 7. These conclusions indicate that the change in hydrogen-ion concentration in the food-vacuoles differs greatly in the protozoa. The validity of some of them is, however, so equivocal that further investigations are highly desirable. Detailed observations were therefore made on the changes in the hydrogen-ion concentration in the food-vacuoles in Vorticdla. Two methods were used : one consisted of observations on the solubility of crystals in the vacuoles ; the other of observations on changes in the color of ingested yeast-cells which had been stained with various indicator dyes. 208 MAST AND BOWEN Ingested crystals indicating acidity Neutral red was added to pond-water (pH 8.2) and left for several hours. During this time numerous long needle-like yellowish brown crystals formed. Some of these crystals were broken up by mounting a little of the solution containing them under a cover-glass on a slide and vigorously tapping the cover-glass. Some of the broken crystals were then drawn under the cover-glass on a preparation containing several vorticellae in pond-water. The vorticellae occasionally ingested pieces of the crystals, some minute, others as long as the diameter of the vacuoles. A considerable number of vacuoles containing such pieces were carefully observed. No changes were seen in any of the pieces of the crystals until after the vacuoles which contained them had left the pharynx and had decreased considerably in size (but not to a minimum) then they suddenly dissolved. The relation between the solubility of these crystals and the acidity of the solu- tion surrounding them was ascertained by adding some to Clark buffer solutions differing in hydrogen-ion concentration and to Hahnert culture solutions containing different quantities of HC1. It was found that their solubility is closely correlated with the acidity of the solutions and that the lowest acidity in which they dissolve readily is approximately pH 5 in the buffer solutions and approximately pH 4 in the HC1 solutions (Mast, 1942). The results obtained in the observations on the crystals in the food-vacuoles indicate, therefore, that the hydrogen-ion concentration of the fluid in the food- vacuoles increased from pH 8.2 to about pH 5 as the size of the vacuole decreased. But since the crystals dissolved before the vacuoles had (as stated above) reached their minimum size, the maximum acidity of the fluid in them must have been higher than pH 5. The results obtained in the following observations confirm this contention. Ingested indicator dyes sliowing uia.riinmn and minimum acidity in the food-vacuoles Methods. — Yeast-cells were boiled in distilled water containing respec- tively the following indicator dyes: meta cresol purple (range pH 1.2-2.8 and 7.4-9), thymol blue (range pH 1.2-2.7 and 8-9.6) metanil yellow (range pH 1.2-2.8), benzopurpurin (range pH 1.2-4), dimethyl yellow (range pH 2.8-4.4), brom phenol blue (range pH 3-4.6), methyl orange (range pH 3.2-4.4), methyl red (range pH 4.2-6.3), brom cresol purple (range pH 5.2-6.8), congo red (range pH 3-5), brom thymol blue (range pH 6-7.6), neutral red (range pH 6.8-8), phenol red (range pH 6.8-8.4), nile blue (range pH 7.2-8.6), and cresol red (range pH 7.2-8.8). The yeast-cells stained well in benzopurpurin, brom phenol blue, congo red, brom thymol blue, neutral red and nile blue but not in any of the others. Some yeast-cells stained with each of these six different dyes were put respec- tively into pond-water (pH 8.2) and presented to vorticellae in pond- water under rover-glasses. In nearly all the preparations the vorticellae ingested some of the stained yeast cells, and in these they sometimes ingested them so freely that the food- vacuoles became well filled with them. The number in the vacuole could, however, be controlled by regulating the number in suspension in the surrounding medium. FOOD-VACUOLE IN PERITRICHA 209 Observations were made on numerous vacuoles, containing various numbers of yeast-cells, from the time the cells entered the oesophageal sac until they were discharged. A series of Clark buffers 5 in small test-tubes was arranged in a row in a test- tube rack for each dye and an appropriate amount of the dye added to each buffer in the series. The acidity of adjoining buffers differed by 0.2 pH. The color of the stained yeast-cells in the vacuoles was, as the vacuoles formed and circulated in the body, continuously compared with those of the buffers in the series containing the dye under consideration and the hydrogen-ion concentration of that which it most nearly matched noted. It is assumed that this was the hydrogen-ion concen- tration of the substance in the vacuole at the time the comparison was made. Congo red (pH j, orangc-pH 5, blue) The results obtained with yeast-cells stained with congo red are more clear-cut than those obtained with any of the other dyes used. This is due to the brilliance and density of the color of the yeast-cells stained with this dye and to the striking change in color correlated with changes in hydrogen-ion concentration. The yeast-cells in the pond-water in which the vorticellae were mounted were dense brilliant orange in color. Those which were ingested retained this color for an average of 75 seconds after the food-vacuole had left the pharynx, then, as the vacuoles decreased in size, they gradually became purple, then more and more bluish until the vacuoles had become minimum in size and the cells, if there were but a few in a vacuole, sky-blue in color (about pH 3). This color they now retained for an average of nearly 2 minutes, i.e. until the vacuoles very rapidly increased in size, then the cells suddenly became orange of the same shade as that which they had when they entered the vacuoles. This color was retained until the content of the vacuoles was discharged which usually occurred within half an hour. There was no indication of digestion in the discharged yeast-cells. A typical record taken from our notes reads as follows : (2:10 p.m.) A yeast-cell entered a vacuole; (40 sec. later) the vacuole, con- taining only one yeast-cell, left the pharynx, spindle-shaped, 12 p, long and 8/t wide, yeast-cell still orange; (75 sec. later) yeast-cell slightly purple,6 vacuole spherical, 8/i in diameter; (75 sec. later) yeast-cell sky-blue (about pH 3), vacuole 3/i in diameter, slightly irregular in form; (2 min. later) yeast-cell orange, vacuole spherical, 8/t in diameter; (15 min. later) yeast-cell orange, no change in structure, vacuole same in size. The results presented indicate, therefore, that the acidity of the fluid in the food-vacuoles in Vorticella increases nearly, if not quite, to pH 3 and that this is closely correlated with decrease in the size of the vacuoles. The conclusion that increase in the acidity of the content of the food-vacuoles is closely correlated with decrease in size is strongly supported by results obtained in observations on congo red-stained yeast-cells ingested in 0.05 M lactose in pond- 5 The following buffers were used : phthalate, pH 2.6-3.4 ; acetate, pH 3.6-5.6 ; phosphate, pH 5.8-8; borate, pH 7.8-10. 6 Under high power (oil-immersion objective), it could be seen clearly that the central portion of the cells was still orange and that the purple was confined to a thin layer at the surface. 210 MAST AM) !'.( (WEN water. In these observations it was found that the foocl-vacuoles do not decrease as much in size as they do in pond-water without lactose, there always being considerable fluid left in them and that the color of the yeast-cells usually changes from orange to purple hut never to blue, indicating that the acidity of the content of the vacuoles increases only to pH 5 in place of nearly to pH 3. \Ye repeatedly observed that if the food- vacuoles contained many congo red- stained yeast-cells, the cells did not become blue. We consequently made extensive observations on the relation between the number of yeast-cells in a vacuole and the extent of change of color in them and found the following: In the vacuoles which contained five cells or fewer there usually was a change in color from orange to blue, but it required considerable longer in those which contained five than in those which contained only one or two cells. In those which contained six to nine cells, there usually was a change from orange to purple but not to blue, and the decrease in the size of the vacuoles was much less than in those which contained only a few cells. In the vacuoles which contained ten or more yeast-cells no change in color was observed and there was but little if any decrease in size. Two typical records from our notes follow: (10:05 a.m.) A vacuole containing five yeast-cells left the pharynx; (2 min. later) yeast-cells getting purple, vacuole but little larger than the five cells; (1 min. later) cells bluish; (15 sec. later) cells blue, very little fluid in vacuole, irregular in form; (30 sec. later) cells turning orange, vacuole clearly larger, nearly spherical, hyaline layer at surface; (45 sec. later) cells orange, vacuole spherical, original size. (10:30 a.m.) A vacuole containing about ten yeast-cells left the pharynx; observed continuously for six min. ; no perceptible change in the color of the yeast-cells or the size of the vacuole. These results show that the extent of change in color from orange toward blue in congo red-stained yeast-cells in the food-vacuoles and the extent of decrease in size of the vacuoles vary inversely with the number of yeast-cells in the vacuoles. They consequently support the conclusion that increase in the acidity of the content of the vacuoles is correlated with decrease in their size. The question now arises as to whether or not the extent of change in color depends upon the time that the yeast-cells are in the vacuoles. Information con- cerning this question was obtained by making observations on cells which entered the vacuole at different times. Since the formation of the vacuole required from 30 to 60 seconds, this sometimes differs by nearly 60 seconds. It was repeatedly observed, however, that if a yeast-cell enters immediately after the vacuole begins to form it takes just as long for it to become purple after the vacuole has left the pharynx as it does if the cell enters just before the vacuole leaves it. Moreover, several vacuoles were studied in which one cell had entered at the beginning of formation and another just before the end of formation, and it was found that in all these vacuoles the two cells became purple and blue at the same time, although one of them had been in the vacuole nearly 60 seconds longer than the other. These results show that the acidity of the solution in the forming vacuoles is not high enough to have any perceptible effect on the color of yeast-cells and that the observed changes in color in them is not specifically correlated with the time they have been in the vacuole>. FOOD-VACUOLE IN PERITRICHA 211 Brom phenol blue (pH 3, yelloiv-pH 4.6, blue; benzopurpurin pit 1.2, violet-pH 4, red) The results obtained with brom phenol blue confirm in general those obtained with congo red. With vorticellae in pond-water the stained yeast-cells were dense sky-blue when they entered the food-vacuoles and they became distinctly greenish yellow when the vacuoles had reached minimum size, but their color corresponded more nearly with buffer, pH 3.2 than pH 3. When the vacuoles increased in size the yeast-cells rapidly became blue again. The results obtained with brom phenol blue therefore indicate that the maximum acidity reached by the substance in the vacuoles is pH 3.2, i.e. not quite so high as is indicated by those obtained with congo red. The yeast-cells stained with benzopurpurin were deep red when they entered the food-vacuoles and no appreciable change occurred as the vacuoles passed through the body. If the maximum acidity in the food-vacuoles is actually pH 3 as the results obtained with congo red indicate, one might expect some evidence of change in the color of the cells stained with this dye, for its range extends from pH 1.2 to 4. The difference in color between buffer pH 3 and pH 4 was however so inconspicuous that it would be extremely difficult to distinguish in yeast-cells in food-vacuoles. The fact then that no change in color was observed in the food- vacuoles containing yeast-cells stained with benzopurpurin, does not seriously militate against the results obtained with congo red and brom phenol blue. The results presented, therefore, seem to prove that the acidity of the substance in the food-vacuole in Vorticella increases from somewhat less than pH 5 to a maximum of pH 3.2 as the size of the vacuole decreases to a minimum and that the acidity very rapidly decreases as the size of the vacuole suddenly increases, and they show that this decrease extends beyond the highest limit of the ranges for the dyes used, namely pH 6.8, but they do not show how far beyond this range it extends. The results obtained with brom thymol blue and neutral red concern this, and also the acidity of the content of the oesophageal sac. Brom thymol blue (pH 6, yclloii'-pH 7.6, blue); neutral red (pH 6.8. rcd-pH 8, amber) ; nilc blue (pH 7.2, blue-pH 8.6. purple) The yeast-cells stained with brom thymol blue were deep blue when they entered the vorticellae in pond-water. In the oesophageal sac, they became dis- tinctly yellowish, pH 6.4, if there were but few present. After the vacuoles had formed and left the pharynx and began to decrease in size they soon became bright lemon yellow, pH 6 ; then when they suddenly increased in size they rapidly became yellowish blue, like buffer pH 6.8, possibly pH 7, but positively not so blue as pH 7.2 and not nearly so blue as they were when they entered the vacuoles. The results obtained with brom thymol blue consequently indicate that the minimum acidity reached is approximately pH 6.9. The yeast-cells stained with neutral red were brownish yellow (pH 8.2) when they entered the vacuoles. They very soon became reddish pink after the vacuoles had left the pharynx and began to decrease in size, but there was no appreciable change in color when the vacuoles later suddenly increased in size. The color of the buffers in the prepared series was essentially the same from pH 5 to pH 7, but at pH 7.2 it was distinctly yellowish. There was no indication of this color in the 212 MAST AND BOWEN yeast-cells in the old vacuoles. The acidity in these cells therefore did not decrease to pH 7.2. These results therefore support the conclusion reached on the basis of those obtained with brom thymol blue, namely, that the minimum hydrogen-ion concentration reached in the substance in the food-vacuoles in Vorticclla during the process of digestion is between pH 6.8 and 7, i.e. that the substance in the food- vacuole decreases greatly in acidity but does not actually become alkaline. The yeast-cells stained with nile blue in pond-water were sky-blue and there was no change in color in those which were ingested. These results therefore have no bearing on the problem under consideration. It can be concluded, then, that in Vorticclla after the food-vacuole leaves the pharynx, the acidity of its content increases from approximately pH 6.4 nearly to pH 3 in about two minutes, with a decrease in size during this time to about 1/27 of its original volume, that it then remains nearly constant in acidity and in size for nearly two minutes, after which it very rapidly increases in size with a ver\v rapid decrease in acidity to about pH 6.9. The problem concerning the processes involved in the changes in size has been considered in a preceding section ; that con- cerning those involved in the changes in acidity will be considered in the following section. Discussion Shipley and DeGaris (1925) maintain that in Paramccinm the fluid in the food-vacuole first becomes alkaline, then acid, then alkaline again. We obtained no evidence whatever indicating a preliminary alkaline phase in the food-vacuoles of Vorticclla. Shapiro (1927) also failed to find any indication of it in this genus, but he maintains that he found a preliminary alkaline phase in Paramccinm if the culture fluid is neutral but not if it is alkaline. It would seem, however, that in alkaline solutions, as Rowland (1928) has well said, "it obviously should have been more prominent than in neutral solutions." The contention of Shipley and DeGaris is consequently equivocal. Moreover, evidence will be presented in a sub- sequent paper which indicates that it is not valid. The food-vacuoles in Actinospherium into which Rowland (1928) injected dyes contained active ingested organisms. These, owing to metabolism, un- doubtedly caused increase in the acidity of the fluid in the vacuoles and the mechanical injury produced by the pipet used in the process of injection also augmented the acidity. The maximum acidity she observed, namely pH 4.3 ±0.1, is therefore higher than that which obtains under normal conditions in vacuoles which do not contain living organisms. Gaff et al. (1941 ) maintain that in culture fluid containing neutral red, the fluid in the iood-vacuole in Bresslaua becomes pink and also the organisms in it after they die, indicating increase in acidity. They hold that this increase in acidity is due to "a sudden release of an acid into the newly-formed food-vacuole" from the sur- rounding cytoplasm. But they also maintain that there are numerous "cherry red granules" in the cytoplasm and that many of them aggregate on the surface of the vacuole. \Ve have made many observations which strongly indicate that the pink color observed by ClaiT et al. in the fluid was due to the effect of the "cherry red granules" on the transmitted light, not to dye in the fluid, and that the pink color in the dead organisms was due to the acid produced in them as they died, not to acid FOOD-VACUOLE IN PERITRICHA 213 in the fluid around them, for similar changes in color occur in organisms which die in neutral red solutions which are not in the food-vacoules. To obtain accurate results with dyes concerning the hydrogen-ion concentration of the content of the food-vacuoles in protozoa, it is therefore necessary to avoid injuring the cytoplasm around the vacuoles and to consider the effect of colored granules in the cytoplasm on the light transmitted through it, and the acid produced by metabolism and death of. organisms in the vacuoles. In the methods used in the observations on changes in acidity in the food- vacuoles in Vorticclla considered above all these sources of error were avoided. The results obtained must therefore be fairly accurate. FACTORS INVOLVED IN THE CHANGES IN ACIDITY IN THE FOOD-VACUOLES It is widely held that change in acidity observed in the food-vacuoles in the protozoa is due to secretion of acid or base by the cytoplasm adjoining the vacuoles (Greenwood and Saunders, 1894; Nirenstein, 1905; Lund, 1914; Rowland, 1928; Claff et al., 1941). Mast (1942) maintains, however, that this does not obtain in Amoeba. He says that in this organism "the cytoplasm secretes neither acid nor base" and he concludes (p. 203) : "The increase in the acidity of the fluid in the food-vacuoles probably is due to respiration in the ingested organisms, chemical changes associated with their death, disintegration of the ingested plasmalemma, impermeability to acids of the membrane around the vacuoles and diffusion of fluid from the vacuoles. The decrease in acidity is due to diffusion of alkaline fluid from the cytoplasm into the vacuoles. The cytoplasm secretes neither acid nor base." Let us consider these views in reference (1) to the increase and (2) to the de- crease in acidity observed in the food-vacuoles in Vorticclla. (1) Increase in acidity If the increase in acidity in the food-vacuole is due to secretion of acid by the surrounding cytoplasm, the acid must pass from the cytoplasm either into the oesophageal sac or the food-vacuoles. We have demonstrated that fluid passes continuously out of the food-vacuoles from the time they begin to form until they have become minimum in size, i.e. during the time that the acidity in them increases to maximum. Consequently, if the increase in acidity is due to secretion of acid by the cytoplasm, it must pass into the pharynx or the vacuole against an outward current of fluid. This is highly improbable. Moreover, since the acidity of the content of the food-vacuoles reaches pH 3.2 and that of the adjoining cytoplasm is, as will be demonstrated presently, approximately pH 7.4 the acid in the vacuoles could come from the cytoplasm only by active secretion. There is, however, no indication whatever of a structure by means of which this could be accomplished. Secretion of acid by the cytoplasm into the vacuole is therefore not at all probable. If living organisms in the food-vacuoles are involved in the increase in acidity in them, there obviously should be no change in acidity in food-vacuoles which do not contain living organisms. The following observations concern this : Vorticellae were mounted in normal pond-water, then this was replaced by sterile pond-water by letting it flow continuously through the preparation for at least 214 MAST AND BOWEN live minutes, then yeast-cells stained with eongo red in sterile pond-water were added. The vorticellae ingested some of the yeast-cells and the color of those ingested changed as the vacuoles proceeded on their course. No difference in these changes and those which occur in normal pond-water, either in time or shade, could he detected. This experiment was repeated several times with sterile pond- water and also with distilled water. The results obtained agree with those presented above, with the exception that in distilled water it required a little less time for the change from orange to purple (increase in acidity) and the vacuoles were not quite so small when it occurred. This is doubtless due to the fact that the distilled water used was pH 5.5 and contained no buffers, whereas the pond-water was pH 8,2 and contained buffers, and therefore required more acid to produce the observed increase in acidity. It can consequently be concluded that if metabolism in living organisms in the food-vacuoles in Vorticella is a factor in the production of the observed increase in acidity, it is of minor importance. Yeast-cells which have been stained with congo red are, as previously stated, not digested. In the food-vacuoles formed by vorticellae in distilled water, con- taining these cells, there is consequently very little if any digestion. It was found, however, that the increase in acidity in these vacuoles is just as great as it is in those which contain an abundance of digestible substance. It is therefore obvious that digestion is not extensively, if at all, involved in the production of acid in the food- vacuoles. What, then, causes the observed increase in acidity in the food-vacuoles ? Lund (1914, p. 14) demonstrated that in Bitrsaria the acidity of the substance which enters the vestibulum increases as it passes thru the pharynx and he con- cluded that this shows that the cytoplasm secretes acid and pours it into the pharynx. There is, however, a more likely cause of the increase in acidity observed by Lund. In the protozoa the cilia in the feeding apparatus are very active during the process of feeding and they perform a considerable amount of work in forcing fluid into the vestibulum and through the pharynx into the oesophageal sac. Metabolism in them and in the cytoplasm associated with them is, therefore, high. This, owing to the production of carbonic, lactic, and other acids, causes increases in the hydrogen-ion concentration 7 of the fluid as it passes through the feeding apparatus. After the fluid has entered the oesophageal sac, some of it passes out through the limiting membrane into the cytoplasm and still more after the food- vacuole has been formed and has left the pharynx, as shown by its rapid decrease in size. Moreover, Bo/.ler (1924), Fortner (1924, 1926), Eisenberg (1925). Miiller (1932) and especially Frisch (1937) have demonstrated fairly conclusively that fluid passes continuously from the pharynx into the cytoplasm during the process of feeding. This would further increase the acidity of the substance in the pharynx as it passes through, if the wall of the pharynx is impermeable to the acids produced by metabolism but permeable to bases, as it may well be. It is therefore highly probable that the increase in acidity in the pharynx observed by Lund is due to the end products of metabolism rather than to secretion by the cytoplasm. The acid in the pharynx obviously passes into the food-vacuole, and if the 7 By adding brom thymol blue to weakly buffered culture fluid containing protozoa, it can readily be demonstrated that they produce acid in the process of metabolism. FOOD-VACUOLE IN PERITRICHA 215 membrane at the surface of the vacuole is impermeable to acids but permeable to bases, the acids will remain in the vacuole as fluids and bases pass out, and the acidity of its content will increase. Mast (1942) accounted for the increase in the acidity of the food-vacuole in Amoeba by means of similar assumptions. The source of the acid appears however to differ greatly in the two organisms. The question now arises as to whether the loss of fluid from the food-vacuole in Vorticella is great enough to produce the observed increase in acidity in it. We do not know precisely what the hydrogen-ion concentration of the content of the food-vacuole is when it leaves the pharynx, but the results obtained in observations on ingested yeast-cells stained with brom thymol blue indicate, as stated above, that it is about pH 6.4. However, Lund (1914) found in observa- tions on Bursaria that ingested vitellin and yolk granules in an alkaline solution containing litmus, change from blue to red in the pharynx, before they enter the food-vacuole. This shows that the content of the forming food-vacuole in Bursaria is distinctly acid. Lund has reproduced the color assumed by the litmus-stained granules in the pharynx. By comparing this color with that of litmus paper in each of a series of buffers, ranging from pH 5.2 to pH 6.6, it was found that it is more nearly like the litmus paper in buffers pH 5.8 (and lower) than that in any of the other buffers in the series. This indicates that the solution ingested by Bursaria changed from distinctly alkaline approximately to pH 5.8. These results support the conclusion reached above, namely, that the hydrogen-ion concentration of fluid ingested by Vorticella increases considerably before the food-vacuole leaves the pharynx, and they indicate that it probably increases to pH 6. If this is true, the acidity of the food-vacuoles in Vorticella increases approximately from pH 6 to a maximum of pH 3.2 as the vacuoles decrease in size. As previously stated, the decrease in the size of the food-vacuoles and the in- crease in the acidity of their content varies greatly, the one being roughly propor- tional to the other. Let us therefore consider the results obtained in actual measure- ments of the changes in size and acidity observed in a typical vacuole. These results show that the vacuole selected decreased in size from an ellipsoid 8 ) ' 12/j. to a sphere 3^ in diameter and that the acidity of its content increased approxi- mately from pH 6 to pH 3.2. If it had been full of fluid at pH 6 when it was maximum in size, it would have contained 6 X 10~19 moles of H+, and if it had been full of fluid at pH 3.2 when it was minimum in size, it would have contained 89 X 10~19 moles of H+, i.e., it would have contained nearly 15 times as much H+ when it was minimum in size as it would have when it was maximum in size. According to the postulated hypothesis it should contain the same amount. The loss of fluid during the reduction in size would therefore not have been sufficient to account for the observed increase in acidity on the basis of this hypothesis. The vacuole was however not full of fluid. It contained approximately three percent of solids when it was maximum in size and 97 percent when it was mini- mum. It therefore contained about three percent less than 6 X 10 ut moles of H+ or 5.82 ) ; 10~H) moles at maximum size, and 97 percent less than 89 X 10"19 moles of H+ or 2.67 ; ' 10~1!) moles at minimum size, i.e., less than half as much as at maximum. The increase in the concentration of hydrogen-ions, owing to loss of water during the decrease in the size of the vacuole, would therefore seem to be ample to account for the observed increase in acidity if there is, in accord with our hypothesis, no loss in hydrogen-inns. 216 MAST AND BOWEN (2) Decrease in acidity The decrease in acidity in the food-vacuoles is, as previously stated, accom- panied by a very rapid and extensive inflow of fluid from the cytoplasm. The fact that this inflow requires only about three seconds and is many times as great in volume as the fluid already in the vacuole, indicates very strongly that the decrease in acidity is due to low acidity of the fluid which enters from the cytoplasm, and that nothing in the nature of secretion is involved. THE HYDROGEN-ION CONCENTRATION OF THE CYTOPLASM IN VORTICELLA No one has previously investigated the hydrogen-ion concentration of the cytoplasm in any of the ci Hates but several have investigated it in the rhizopods. Pantin (1923) maintains that the hydrogen-ion concentration of the cytoplasm in a small marine amoeba is pH 7.6-7.8 in the plasmasol, pH 7.2 in the plasmagel and pH 6.8 in the protruding pseudopods. Needham and Needham (1925) con- clude that in Amoeba protcus it is pH 7.6 throughout, and Chambers, Pollack and Hiller (1927) contend that in Amoeba protcus and Amoeba ditbia it is pH 6.9 ±0.1. Mast (1942) has considered these contentions critically. He contends that the methods used are not reliable and comes to the conclusion on the basis of his own observations that the hydrogen-ion concentration of the cytoplasm in Amoeba proteus is approximately pH 7.4. The results presented above show that in Vorticclla the flooding of the food- vacuole with fluid from the cytoplasm usually causes the vacuole to increase about 25 times in volume and the acidity of their content to decrease approximately from pH 3.2 to pH 6.9. But since the vacuole contains approximately 97 percent solids at minimum size and only some five percent at maximum, the fluid in it increases more than 500 times. Since the two fluids mixed in the vacuole are buffered, and their relative amounts and the hydrogen-ion concentration of one of them and that of the mixture are known approximately, that of the other (the fluid in the cytoplasm) can be ascertained approximately, by mixing appropriate buffers in proper proportions and measuring the hydrogen-ion concentration of the mixture. This was done, and it was found that if one part of a pH 3.2 buffer is added to 500 parts of a pH 7 buffer (the approximate proportion of the two fluids mixed in the vacuole), the hydrogen-ion concentration of the mixture is pH 6.98. This indicates that the hydrogen-ion concentration of the cytoplasm in Vorticella is slightly higher than pH 7. i.e., considerably higher than that of the cytoplasm in Amoeba protcus. THE FUNCTION OF THE CHANGES IN THE HYDROGEN-ION CONCENTRATION IN THE FOOD-VACUOLES Hemmeter (1896), Howland (1928) and Claff et al. (1941) maintain that the increase in acidity in the food-vacuoles in protozoa serves to kill the ingested organ- isms. Xirenstein (1905) concludes, however, that in Paramccium the acidity of the content of the food-vacuoles does not become high enough to kill the ingested nrganisms and Mast (1942) comes to the same conclusion in reference to Amoeba. It is consequently doubtful whether the increase in acidity functions as a killing FOOD-VACUOLE IN PERITRICHA 217 agent in any protozoa, and according to Greenwood and Saunders (1894), Niren- stein (1905) and Mast (1942) it does not function directly in digestion in Amoeba and Paramecium, for digestion does not begin in these organisms until after the acidity in the food-vacuoles has decreased to a minimum. Vorticclla, as previously stated, feeds almost exclusively on bacteria. After the bacteria have been carried into the forming food-vacuole by the action of the cilia in the feeding apparatus, they swim actively about in the fluid in it and they continue swimming until a few moments after the vacuole has left the pharynx and has de- creased somewhat in size, then they stop abruptly (all coming to rest at practically the same instant) and usually soon aggregate in a dense mass in the central region of the vacuole. They have doubtless been killed, for they do not become active again when the vacuole enlarges and the mass breaks up. This also occurs if lactose (0.05 M) is added to the culture fluid. The hydrogen-ion concentration of the fluid in the vacuoles when the bacteria became inactive, could not be accurately measured, but the results obtained in obser- vations on ingested yeast-cells stained with congo red indicate that it is not higher than pH 5. Moreover, in culture fluid containing 0.05 M lactose the acidity, as stated above, increases only to approximately pH 5. The bacteria in it are, there- fore, not subjected to higher concentration of acid than this. The lethal concentration of acid for the bacteria was ascertained by adding to given quantities of culture fluid different quantities of hydrochloric acid and measur- ing the time the bacteria in the culture fluid lived. It was found that they lived in- definitely in the culture fluid at pH 5 and more than 30 seconds in the culture fluid at pH 4. It is consequently obvious that death of the bacteria in the food-vacuoles is certainly not entirely due to the increase in acidity. The time between the separation of the vacuole from the pharynx and the cessa- tion of movement of the bacteria in it, was measured with a stopwatch. It was found that this varies considerably in consecutive vacuoles in the same individual, but that the average for different individuals is fairly uniform. The variation for ten consecutive vacuoles in a typical individual was 14 to 18 seconds with an average of 16.3 seconds. It requires, as stated above, about 50 seconds to form a food-vacuole. The bac- teria which enter when it begins to form are, therefore, in it about 66 seconds before they are killed, whereas those which enter just before it leaves the pharynx are in it only about 16 seconds. The cause of death must therefore be due largely, if not entirely, to changes in the content of the vacuole after it leaves the pharynx. There are, as previously stated, two very prominent changes during this time, increase in acidity and decrease in fluid. It was demonstrated above that the increase in acidity is not fatal. Death in the food-vacuoles is therefore probably due to the loss of fluid. Mast (1942) comes to the same conclusion in reference to the cause of death in the food-vacuoles in Amoeba. He contends that the loss of fluid augments the decrease in oxygen in the vacuoles due to respiration in the bacteria, to such an extent that it is fatal. There is no visible indication of digestion of the bacteria in the food-vacuoles until after the acidity in them has decreased to minimum. This seems to show that the increase in acidity does not function in digestion. There are however profound changes in the vacuole while the acidity in it is maximum for, as previously stated, the osmotic concentration of the fluid in it during this time increases greatly. It 218 MAST AND BOWK.X may well be, therefore, that the increase in acidity functions in the production of this increase in osmotic concentration, e.g. by hydrolizing complex molecules, which in turn functions in the inflow of fluid-carrying enzymes which facilitate digestion. The decrease in acidity in the food-vacuoles is clearly correlated with digestion, but since it is merely the result of rapid inflow of fluid from the cytoplasm it is obviously not the result of anything in the nature of secretion by the cytoplasm. THE OSMOTIC CONCENTRATION OF THE CYTOPLASM IN VORTICELLA If the decrease in the size of the food-vacuoles in Vorticella were, entirely due to difference between internal and external osmotic concentration, and if the mem- brane at the surface of the food-vacuoles were permeable to water only, and if no osmotically active substance passes into the feeding apparatus from the cytoplasm, the osmotic concentration of the cytoplasm could be measured by changing that of the ingested fluid until there is no decrease in the size of the vacuole after it leaves the pharynx. It was however demonstrated above that inward pressure of the stretched membrane around the vacuole is functional in the decrease in its size, and it is highly probable that some osmotically active substance enters the feeding appa- ratus from the adjoining cytoplasm. The decrease in the size of the food-vacuole is therefore probably not closely correlated with the relation between the osmotic con- centration of the fluid in the food-vacuole and that of the fluid in the cytoplasm. It was found however that the size of the entire body varies consistently with the osmotic concentration of the surrounding medium and that this relation can be fairly accurately measured. Observations on it were therefore made as follows : A vorticella attached to a fragment of Lcinna was mounted in pond-water or tap-water under a cover-glass supported on two parallel ridges of vaseline and the length and width of the body measured by means of an ocular micrometer. Then the water was replaced with a solution of lactose in pond-water or tap-water, left ten minutes and the vorticella again measured. This was now repeated with dif- ferent concentrations of lactose and with different individuals. The results ob- tained in reference to length are presented in Table III. The width varied directly with the length, but it also varied with the surface viewed. It was therefore not recorded in the table. Table III shows that the vorticellae decreased in size in the higher concentra- tions of lactose used, but not in the lower, and that the decrease varied directly with the concentration, but that it was greater in the vorticellae which had been adapted to pond-water than in those which had been adapted to tap-water. It shows that nl the seven individuals adapted to pond-water, five became slightly smaller in 0.0125 M lactose in pond-water and two did not change in size, but that in 0.025 M lactose all became definitely smaller; whereas in the nine individuals adapted to tap- water only two became smaller in 0.0125 M lactose in tap-water and only seven became smaller in 0.025 M lactose. The lowest osmotic concentration which causes any decrease in si/.e, is, therefore, a little lower than that of 0.0125 M lactose in pond-water for vorticellae adapted to pond-water, and a little higher than 0.0125 M lactose in tap-water for vorticellae adapted to tap-water. If, then, the decrease in si/e in the lactose solutions is due to the difference between internal and external osmotic concentration this difference must be slightly less than the osmotic concen- tration of 0.0125 M lactose for the vorticellae which have been adapted to pond- FOOD-VACUOLE IN PERITRICHA 219 TABLE III Relation between the size of Vorticella and the osmotic concentration of the surrounding medium Each specimen used was measured successively in the four concentrations; specimen a, three times in each; b, c and d, twice in each; and the rest once in each. The lengths of a, b, c and d given, are averages. Length of body in micra Designation Concentration of lactose in pond-water of specimens 0.05 M 0.025 M 0.0125 M OM a 77.50 81.66 86.66 90.83 b 52.50 57.50 63.25 65.00 c 43.75 46.75 48.75 50.00 d 60.00 65.00 71.25 72.50 e 75.00 75.00 80.00 82.50 f 50.00 62.50 70.00 70.00 g 57.50 62.50 70.00 70.00 Total average 59.46 64.41 69.98 71.26 • Concentration of lactose in tap-water 0.05 M 0.025 M 0.0125 M OM ai 70.00 77.00 80.50 80.50 b, 59.50 66.50 66.50 66.50 C] 56.00 70.00 70.00 70.00 d, 52.50 66.50 70.00 73.50 e, 60.75 73.50 77.00 77.00 f, 52.50 59.50 63.00 63.00 gi 63.00 66.50 70.00 70.00 hi 85.75 91.00 92.75 87.50 ii 53.25 63.00 63.00 64.75 Total average 61.45 70.38 72.52 72.52 water and slightly more than that of 0.0125 M lactose for those which have been adapted to tap-water. The osmotic concentration of the pond-water used, calculated from the depres- sion of the freezing point, is at 22° C, equivalent to 0.79 atmospheres and that of the tap-water practically zero. The results presented indicate, therefore, that the lower the external osmotic concentration is, the greater the difference between in- ternal and external osmotic concentration becomes. The osmotic concentration of 0.0125 M lactose at 22° C is equivalent to 0.3282 atmospheres (International Critical Tables). That of the fluid in the cytoplasm must, therefore, be approximately equivalent to 0.3282 plus 0.79 atmospheres or 1.1 atmospheres in vorticellae adapted to pond- water, but only slightly higher than 0.3282 atmospheres in those adapted to tap-water. Kitching (1938) concludes that in Zoothamnium sp., a freshwater peritrich, the excess of internal over external osmotic concentration is equivalent to that of 0.05 M 220 MAST AND BOWEN sucrose, that is, four times as large as the results \ve obtained in our observations on Vorticclla. This difference is much larger than would be expected in organisms which are so nearly alike in structure and habitat. Kitching's conclusion was based ( m results obtained with specimens treated with cyanide. It may well be, therefore, that it is not valid for specimens under normal conditions. Mast and Fowler ( 1935) found that 0.005 M lactose in culture fluid is the lowest concentration which produces a consistent measurable decrease in the volume of .•linocbd protcus. This indicates that the difference between internal and external osmotic concentration is much smaller in Amoeba than it is in V orticclla. \ SUMMARY 1. The feeding apparatus in the Peritricha consists of a ciliated tube (the outer portion of which is called the vestibulum and the inner the pharynx) and about ten fibers (oesophageal fibers) which are attached to the distal end of the pharynx and extend as a bundle through the cytoplasm nearly to the posterior end of the body. There is no oesophageal tube. 2. The Peritricha feed largely on bacteria but various inanimate particles are also ingested. 3. At the end of the pharynx surrounded by the oesophageal fibers there is a cone-shaped sac (the oesophageal sac) which consists of a membrane -probably pro- duced by the interaction between the fluid in it and the cytoplasm around it. 4. The pharyngeal cilia force into the pharyngeal sac culture fluid with particles in suspension and usually gelatinous substance secreted by the peristome, the ves- tibulum and the pharynx. 5. The sac enlarges and becomes spindle-shaped. Then a portion of it is con- stricted off to form a food-vacuole. 6. The constriction is probably due to local simultaneous inward pressure of the oesophageal fibers. 7. The food-vacuoles vary greatly in size. 8. Initiation of the constriction in the sac and the size of the food-vacuole formed by it are not specifically correlated with the size of the sac or particles in suspension in the fluid in it or the chemical composition of this fluid, but they are to some extent dependent upon these factors, especially the chemical composition of the fluid. 9. The concentration of particles in suspension in the fluid in the oesophageal sac increases as the sac increases in size. This is largely due to the passage of fluid through the oesophageal membrane into the cytoplasm. After the food-vacuole is formed, it passes rapidly on a fixed course through the cytoplasm to the posterior end of the body and then slowly on a varied course to the anal spot in the wall of the vestibulum. The rapid movement is probably due to waves passing synchronously along the fibers. The slow movement is due to cyclosis. After the vacuole has reached the posterior end of the body it usually be- comes spherical in form and gradually decreases greatly in size; as it decreases in M/e the acidity of its content increases to a maximum of pH 3.2, then it increases very rapidly in si/.e and the acidity of its content decreases to pH 6.9. 'he hydrogen-ion concentration of the fluid in the cytoplasm is approxi- mately pi I 7. FOOD-VACUOLE IN PERITRICHA 221 13. The decrease in size requires about two minutes. It is due in part to exces- sive external osmotic concentration and in part to inward pressure of the stretched membrane at the surface. The increase in size requires about three seconds. It is due to excessive internal osmotic concentration probably caused by chemical changes produced by the increase in the acidity of its content. 14. The increase in the acidity of the content of the vacuole is probably largely due to the production of acid, owing to metabolism in the peristome, the vestibulum and the pharynx and impermeability of the vacuolar membrane to organic acid, resulting in its retention and consequent concentration as the vacuole decreases in size. 15. The decrease in acidity is due to the flooding of the vacuole with fluid from the cytoplasm. 16. The osmotic concentration of the fluid in the cytoplasm of Vorticella varies directly with that of the surrounding medium. The former is higher than the latter approximately by an equivalent of 0.0125 M lactose or 0.3282 atmospheres. LITERATURE CITED ANDREWS, E. A., 1923. Folliculina : case making, anatomy and transformation. Jour. Morph., 38: 207-278. BOLZER, EMIL, 1924. IJber die Morphologic der Ernahrungsorganellen und die Physiologic der Nahrungsaufnahme von Paramaecium caudatum Ehrb. Arch. f. Protistk., Bd. 49, S. 163-215. BRAGG, A. N., 1935. The initial movements of the food-vacuoles of Parameciwm trichinm. Arch. f. Protostk., Bd. 85, S. 421-425. BRAGG, A. N., 1936. Observations on the initial movements of the food-vacuoles of Paramecium multimicronucleata Powers and Mitchell with comments on conditions in other species of the genus. Arch. f. Protistk., Bd. 88, S. 76-84. BUTSCHLI, O., 1889. Protozoa, Bronn's Klassen und Ordnungen des Thier-Reichs., Bd. 1, S. 1-2035. CHAMBERS, R., H. POLLACK, AND S. HILLER, 1927. The protoplasmic pH of living cells. Proc. Soc. Exp. Biol. and Med., 24: 760-761. CLAFF, C. L., VIRGINIA C. DEWEY, AND G. W. KIDDER, 1941. Feeding mechanisms and nutrition in three species of Bresslaua. Biol. Bull., 81 : 221-234. EHRENBERG, C. G., 1838. Die Injusionsthierchen als vollkommene Organismen. Leipzig. 508 S. EISENBERG, E., 1925. Recherches sur le fonctionnement de la vesicule pulsatile des infusoires dans les conditions normales et sous 1'action de certains agents experimentaux : pression osmotique et electrolites. Arch, de Biol., T. 35, pp. 441-464. FORTNER, H., 1924. tiber die physiologisch differente Bedeutung der kontraktilen Vakuolen bei Paramaecium caudatum. Zool. Ans., Bd. 60, S. 217-230. FORTNER, H., 1926. Zur Frage der diskontinuierlichen Exkretion bei Protisten. Arch. f. Protistk., Bd. 56, S. 295-319. FRISCH, JOHN A., 1937. The rate of pulsation and the function of the contractile vacuole in Paramecium multimicronucleatum. Arch. f. Protistk., Bd. 90, S. 123-161. GELEI, J. v., 1934. Der feinere Bau des Cytopharynx von Paramaecium und seine systematische Bedeutung. Arch. f. Protistk., Bd. 82, S. 331-362. GREEFF, R., 1870-71. Untersuchungen iiber den Bau und die Nahrungsgeschichte der Vorti- cellen. Arch. f. Naturgesch., Bd. 36, S. 353-384; Bd. 37, S. 185-221. GREENWOOD, M., 1894. On the constitution and mode of formation of "food vacuoles" in infusoria, as illustrated by the history of the processes of digestion in Carchesium polypinum. Trans. Roy. Soc. London (B), 185 : 355-383. GREENWOOD, M., AND E. R. SAUNDERS, 1894. On the role of acid in protozoan digestion. Jour. Physiol, 14 : 441^67. HALL, R. P., AND R. F. NIGRELLI, 1930. Relation between mitochondria and food vacuoles in the ciliate Vorticella. Trans. Am. Micr. Soc., 49: 54-57. 222 MAST AND BOWEX HEMMETER, J. C., 1896. On the role of acid in the digestion of certain rhizopods. Amer. Nat., 30: 619-625. ROWLAND, RUTH B., 1928. The pH of gastric vacuoles. Protoplasma, Bd. 5, S. 127-134. KAHL, A., 1935. Urtiere oder Protozoa. 1. Wimpertiere oder Ciliata (Infusoria). Jena, 886 S. KITCHING, J. A., 1938. The physiology of the contractile vacuoles. III. The water balance of fresh-water peritrichs. /. E.rp. Biol., 15: 143-151. KITCHING, J. A., 1938a. On the mechanism of movement of food vacuoles in peritrich ciliates. Arch. f. Protistk., Bd. 91, S. 78-88. KOEHRING, VERA, 1930. The neutral-red reaction. Jour. M 'or ph. ', 49:. 45-137. LUND, E. E., 1941. The feeding mechanisms of various ciliate protozoa. Jour. Morph.f 69: 563-573. I.IXD, E. J., 1914. The relations of Bursaria to food. I. Selection in feeding and in ex- trusion. II. Digestion and resorption in the food vacuole, and further analysis of the process of extrusion. Jour. Exp. Zool., 16: 1-52; 17: 1-43. MAST, S. O., 1942. The hydrogen-ion concentration of the content of the food vacuoles and the cytoplasm in Amoeba and other phenomena concerning the food vacuoles. Biol. Bull., 83 : 173-204. MAST, S. O., 1944. A new peritrich belonging to the genus Ophrydium. Trans. Am. Mic. Soc.. 64: 181-186. MAST, S. O., AND COLEEN FOWLER, 1935. Permeability of Amoeba protcus to water. /. Cell. and Comp. Physiol., 6: 151-167. MULLER, W., 1932. Cytologische und vergleichend-physiologische Untersuchungen iiber Para- macciuin multimicronucleatum and Paraniaccium caudatum, zugleich ein Versuch zur Kreuzung beider Arten. Arch. f. Protistk., Bd. 78, S. 361-462. NEEDHAM, J., AND DOROTHY M. NEEDHAM, 1925. The hydrogen ion concentration and the oxidation-reduction potential of the cell-interior. Proc. Roy. Soc. L., Ser. B, 98 : 259-286. NIRENSTEIN, E., 1905. Beitrage zur Ernahrungsphysiologic der Protisten. Zcit. f. ally. Physiol., Bd. 5, S. 435-510. PANTIN, C. F. A., 1923. On the physiology of amoeboid movement. I. Jour. Mar. Biol. Assoc., 13, 24-69. SCHEWIAKOFF, W., 1891. Uber die Natur der sog. Exkretkorner der Infusorien. Zcit. f. iviss. Zool., Bd. 57, S. 32-56. SCHRODER, O., 1906. Beitrage zur Kenntnis von Campanclla umbcllaria L. sp. Arch. f. Protistk., Bd. 7, S. 75-105. SCHRODER, O., 1906. Beitrage zur Kenntnis von Epist\lis plicatilis (Ehrbg.). Arch. I. Protistk., Bd. 7, S. 173-196. SCHRODER, O., 1906. Beitrage zur Kenntnis von VorticeUa monilata Tatem. Arch. f. Protistk., Bd. 7, S. 395-410. SCHUBERG, A., 1890. Zur Kenntnis des Stcntor cocruleus. Zool. Jahrb., Bd. 4, S. 197-238. SHAPIRO, N. N., 1927. The cycle of hydrogen-ion concentration in the food vacuoles of Paramecium, VorticeUa and Stylonychia. Trans. Am. Micr. Soc., 46: 45-53. SHARP, R. G., 1914. Diplodinium ccaitdatum, with an account of its neuromotor apparatus. Univ. of Calif. Pub. Zool., 13: 43-122. SHIPLEY, P. G., AND C. F. DEGARIS, 1925. The third stage of digestion in Paramecia. Science, 62 : 266-267. WALLENGREN, H., 1901. Inanitionserscheinungen der Zelle. Zcit. f. allg. Physiol., Bd. 1, S. 67-128. HYDROGEN-ION CONCENTRATION OF ALBUMEN AND YOLK OF THE DEVELOPING AVIAN EGG ALEXIS L. ROMANOFF Agricultural Experiment Station, Cornell University, Ithaca, New York The changes in hydrogen-ion concentration of the albumen and yolk in the avian egg have been considered as indicative of the character of the metabolic processes occurring within the egg. These changes are, therefore, of importance in studies of embryonic development (Needham, 1931 ; Romanoff and Hay ward, 1943). A review of the literature (Needham, 1931 ; Romanoff and Romanoff, 1929) indicates that considerable work has been done on the hydrogen-ion concentration of the hen's egg and very little on the eggs of other species (Shklyer, 1937). It is of interest to know whether or not the changes in this physical property of eggs are similar in different species of birds. METHODS AND MATERIALS To obtain these data the present study includes the eggs of the Leghorn chicken (Callus gallus) , Ring-necked pheasant (Phasianus torquatus), Bobwhite-quail (Co- linus virginianus) , White Holland turkey (Mcleagris gallopavo}, Pekin duck (Anas platyrhynchos} and of domestic goose (Anscr anscr}. Particular effort was made to obtain eggs as fresh as possible. On an average the chicken eggs were one or two hours old, while the age of the eggs of other species varied from 24 to 36 hours. The pH value was determined electrometrically, using an hydrogen electrode. The observations were carried out : ( 1 ) on the albumen until it was merged with the yolk sac, thus losing its physical entity, and (2) on the yolk until hatching time. EXPERIMENTAL RESULTS Egg albumen The data for hydrogen-ion changes in albumen (Fig. 1 A) show a striking simi- larity in all the curves. The initial rise in pH, from as low as 7.6 to as high as 9.5, at the beginning of incubation is followed first by a rapid, then by a more gradual decrease to approximate neutrality. All values obtained were for the middle dense layer of albumen, for it has been shown (Romanoff, 1943a) that the pH values of the different layers do not vary to any great extent even in fresh eggs. The results presented here agree with those of Shklyer (1937) for hens, turkeys, ducks and geese in all essentials except for the initial pH values which were higher than ours. Evidently the eggs used by Shklyer in his experiments were of more advanced age before their setting for incubation. Egg yolk Previous observations show that at certain stages of incubation there is a mor- phological differentiation of egg yolk into two fractions — dense and liquefied (Roma- noff, 1943b), with quite distinct electrical conductivities (Romanoff and Grover, 223 224 A. L. ROMANOFF 1936). For that reason, as was anticipated, the dense egg yolk undergoes an en- tirely different change in hydrogen-ion concentration. There is a gradual rise from slight acidity, of about pH 6.0, to an alkalinity of about pH 7.8 near the end of the incubation period (Fig. 1 B). Then the pH values decrease slightly before the time of hatching. The data for all species follow the same general trend of change. This again is in close agreement with data published by Shklyer (1937). pH 95 9.0 85 8.0 Z ui I 75 CD 70 65 - 60 - 5.5 PH 95 9.0 8.5 * _l O LJ 8.0 75 £70 6.5 6.0 20 40 60 80 5.5 B CHICKEN PHEASANT QUAIL TURKEY DUCK GOOSE PH 9.5 9.0 8.5 6.5 6.0 0 INCUBATION 20 40 60 80 100 PERIOD CIN PER CENT) 5.5 20 40 60 FIGURE 1. Changes in pH of avian eggs during embryonic development : A, albumen, B, dense yolk, and C , liquefied yolk. The data based on observations of over 600 eggs, daily averaging from 3 to 15 eggs for each species. The liquefied yolk of avian eggs is consistently more alkaline in reaction than the .semi-solid or dense portion (Fig. 1 C). According to Shklyer (1937) the liquefied yolk in the egg of the domestic fowl maintains an average pH value of about 7.7 throughout its period of existence. DISCUSSION It is now recognized that to obtain comparable results, certain variables must be controlled in ascertaining changes in the hydrogen-ion concentration of the develop- ing egg; this is especially true of the albumen. Two of the most important of these variables are the age of the egg at the time of setting for incubation, and the condi- tions under which they have been kept in storage. HYDROGEN-ION CONCENTRATION IN AVIAN EGGS The pH value of the alhumen increases rapidly in an egg with aging ( Romanoff and Romanoff, 1929). Unless eggs are set for incubation immediately after laying, the initial rise in pH of albumen during the development may not be fully observed. For this reason the values for the incubated eggs given by other investigators (see reviews by Romanoff and Romanoff, 1929; Needham, 1931; and Shklyer, 1937) frequently show high initial pH value, which afterwards has only a steady decrease towards acidity. The older the egg at the beginning of the incubation the nearer the pH value will be to the peak of alkalinity in the initial stages. In eggs kept under ordinary environmental conditions (temperature about 12-13° C.) for only about seven days, the portion of the curve showing a rapid rise in pH would be almost completely eliminated. Consequently without adequate control of the age of the egg, the results of many former studies demonstrate either only a very slight rise in the curve (Gueylard and Portier, 1925; Penionschkevitch, 1934; Berenstein and Penionschkevich, 1935) or none at all (Aggazzotti, 1913; Buytendijk and Woerdeman, 1927; Shklyer, 1937). The initial rapid rise in the pH of the albumen during early incubation, as well as in storage, has been shown to be caused by the loss of carbon dioxide ( Sharp and Powell, 1931; Brooks and Pace, 1938). It has been determined experimentally that the pH of the egg albumen is in direct relationship to the concentration of car- bon dioxide in the incubator (Romanoff and Romanoff, 1930, 1933). With 10 per cent of carbon dioxide in the air the pH value of albumen does not rise at all— the curve flattens out, and the normal peak of high alkalinity is not observed. Simi- larly in storage at low temperature (0° C.), the rise in pH may be prevented by high carbon dioxide pressure (Moran, 1937). It is the author's experience that with the eggs of the same preincubation age, the incubating temperature, within the range of embryonic survival, 35.5-39.5° C. (Romanoff, Smith and Sullivan, 1938), has a very insignificant effect on the vari- ation in hydrogen-ion concentration of albumen and yolk. Also, negative results were obtained with the changes of relative humidity in the incubator (Penionsch- kevitch, 1934). However, according to Sharp and Powell (1931), the rise in pH of egg albumen prior to incubation is hastened by a higher temperature. SUMMARY The observations on incubated eggs of chicken, pheasant, quail, turkey, duck, and goose clearly indicate that changes in hydrogen-ion concentration of albumen, and of dense and liquefied portions of yolk are similar for all species studied, and suggest a pattern which may be characteristic of all ayian eggs. In the albumen there is a rapid rise in pH, then a fall ; in the dense yolk, a gradual rise with a slight fall at hatching ; and in the liquefied yolk during its existence, high pH value without change. The initial pH value of egg albumen during embryonic development depends chiefly upon the preincubation age of the egg — fresh eggs would give low, while older eggs would give high pH values of albumen at the beginning of incubation. LITERATURE CITED AGGAZZOTTI, A., 1913. Influenza dell'aria rarefatta sull'ontogenesi ; Nota II. La reazione del liquidi dell'ovo durante lo sviluppo. Arch. f. Entivmcch. Ore}., 37: 1-28. 22(> A. L. ROMANOFF BERENSTEIN, F. J., AND 1C. 1C. PEXIONSC HKEVITCH, 1935. Ueber die aktive Reaktion des Inhalts der Hulinrr- und Enteneicr wahrend dor Inkuhationsperiodc. Ph\siol. Jour. ( U.S.S.R.), 18: (.54-659. BROOKS, ]., AND PACK, 1938. Distribution of carbon dioxide in the hen egg. Proc. Ro\. Soc. London. B, 126: 196-210. Hi VTK.XDI.IK, F. J. J., AND M. \\". WoKRDE.M A N , 1927. Die physico-cliemisclien Erscheinungen wahrend des Eientwicklung. I. Die Messung der Wasserstoffionenkonzentration. Arch. f. Entwmech. Org., 112: 387-410. (irr.vi.AKD, F., AND P. POKTIEK, 1925. Reaction ionique des differents constituants de 1'oeuf de la poule. Ses modifications an cours de 1'incubation. Compt. Rend. Acad. Sci., 180: 1962-1963. MOHAN. T., 1937. Gas storage of eggs. Jour. Soc. L'licin. hid.. 56 (T.) : 96-101. XEEDIIA.M, J. 1931. Chemical cinl>ryolo. ZooL, 56: 451-457. ROMANOFF, A. L., AND A. J. ROMANOFF, 1933. Biochemistry and biophysics of the developing hen's egg. II. Influence of composition of air. Cornell Univ. Agr. Exp. Sta. Memoir 150: 1-36. ROMANOFF, A. L., L. L. SMITH, AND R. A. SULLIVAN, 1938. Biochemistry and biophysics of the developing egg. III. Influence of temperature. Cornell Univ. Agr. Exp. Sta. Memoir 216: 1-42. SHARP, P. F., AND C. K. POWELL, 1931. Increase in the pH of the white and yolk of hens' eggs. hid. Eng. Chcm., 23 : 196-199. SHKLYER, N. M., 1937. A study of physico-chemical changes in the egg during embryonic de- velopment of birds. I. Changes in the concentration of hydrogen-ion in relation to embryonic development in the eggs of domestic fowl (hens, turkeys, ducks, geese). Ukrainian Bioclicin. Jour., 2 : 379^406. ON THE INTERPRETATION OF RATES OF REGENERATION IN TUBULARIA, AND THE SIGNIFICANCE OF THE INDEPENDENCE OF MASS AND TIME1 S. SPIEGELMAN AND FLORENCE MOOG Department of Zoology, Washington University, Saint Louis, Department of Zoology, Columbia University, New York, and The Marine Biological Laboratory, Woods Hole In his investigations into the metabolic basis of dominance in Tubularia stems and into the factors involved in hydranth regeneration, Barth (1938a, b) proposed as a measure of regeneration rate L/t, in which L is the length of the regenerating primordium and / the time in hours from the removal of the old hydranth to the appearance of a constriction between the primordium of the new hydranth and the rest of the stem. In some cases Barth also used irr-L/t, where r is the radius of the stem ; but with uniform short stems r is virtually constant and L becomes an adequate estimate of the mass or volume of tissue involved. Although ad- mitting that variation in primordium length occurs, Child (1940) criticized Earth's definition on the grounds that since growth or increase in cell number is not involved in the reconstitution, the inclusion of mass or volume in the measure- ment of rate is of doubtful validity. Child therefore maintained that I// gives a better indication of the kinetics of the process. Miller (1942), contending that Earth's definition is ". . . based on the implied assumption that length and time are inversely related to one another," also questioned the validity of the "implied assumption" on the basis of experiments in which length varied although time did not. Needless to say, Earth's definition of regeneration rate no more implies an inverse relation between the components of the ratio than the usual definition of velocity implies an inverse relation between distance and time. It is important to note that I//, commonly used as a measure of rate, is not free of ambiguity. Since the "one" in the numerator is a dimensionless quantity, it is clear that the magnitude defined by \/t is not a rate in the generally accepted sense. Ordinarily other dimensions such as mass or length are involved, either alone or in combination, in determination of rate.2 This is particularly true when such dimensions enter into the process being studied. Regeneration and differ- entiation, for example, while not usually accompanied by a net mass increase (growth), do imply the occurrence of transformation of mass from one type into another. Presumably these transformations can ultimately be referred to the formation of certain types of compounds at the expense of others. On this basis, then, a rational definition of rate of development would consider the mass of tissue transformed, and might be formulated in terms of mass per unit time. It is admitted that in many cases the difficulty of obtaining mass or volume measure- 1 Aided by grants from a Rockefeller Foundation fund administered by Dr. H. B. Steinbach; and from the Dyckman Fund of the Department of Zoology, Columbia University. 2 The familiar exception is angular velocity, which has the form \/t. Even here however the "one" in the numerator is a consequence of the definition of an angle, which is the ratio of two lengths and therefore dimensionless. 227 SPIEGELMAN AND MOOG merits of the differentiated tissue has made the use of the 1/7 definition of rate necessary; nevertheless it may be of value to recognize that 1/7 does differ from other rates, and to attempt a more precise interpretation of its significance. An analysis of this kind becomes of paramount importance in those investigations which seek to correlate metabolic and developmental rates. Such physiological rates as Q02, <2co2. or <2o2 involve a measurement of mass as well as time. Un- critical comparisons between these rates and I//, which neglect the mass trans- formed, may well lead to erroneous conclusions. In a previous study of the effects of various respiratory inhibitors on regenera- tion and respiration rates, the present authors (Moog and Spiegelman, 1942) used Earth's definition of regeneration rate in establishing that with certain in- hibitors (e.g., urethanes, azide) rather drastic decreases in regeneration rates can occur without any concomitant measureable effects on Q0,. But subsequent experiments by Moog (1942) and Spiegelman (1942) indicated clearly that length and time may be independently affected, and so brought out the inadequacy of L/t expression of regeneration rate as a means of comparing directly the results obtained with a variety of agents. A similar situation was noted by Miller (1942), and also by Spiegelman and Goldin (1944) in interpreting the parallel effects of pH variation on regeneration and respiration rates. Thus it seemed desirable to re-examine the extent of the independence of L and t under different experimental conditions in a more systematic way than had previously been attempted. On the basis of the data obtained and presented here, the significance of L/t as a rate measure will be examined, and the inde- pendent variation of L and / will be interpreted in terms of synthetic reactions in open systems approaching the steady state. MATERIALS AND METHODS The solutions used were made up fresh each week in filtered sea water, and when necessary were adjusted to pH 8.2 with hydrochloric acid. Young un- branched stems uniform in translucence, length, and diameter were selected from colonies freshly gathered from the waters of Vineyard Sound or Cape Cod Bay during the months of July and August. Stem segments 6 mm. in length were cut from regions about five mm. proximal to the hydranth. Groups of 25 stem segments were kept in 100 ml. of the appropriate solution in partly filled, tightly stoppered flasks which were shaken at intervals to redistribute the oxygen. Solu- tions were changed daily, but the stems were kept in the flasks until they re- constituted or were finally transferred to fresh sea water, after four or five days. They were counted as totally inhibited, with rate of regeneration zero, if after being transferred they developed hydranths. In the temperature experiments, the desired temperatures, held constant to 0.5° or better, were obtained with water baths or incubators placed in cold rooms. Solutions in which stem segments were placed were brought to temperature before use. RESULTS A. Narcotics Table I summarizes the results obtained with different narcotics at various concentrations. In the case of ethyl urethane decreases are not observed in re- REGENERATION AS MASS AND TIME TABLE I The effects of narcotics on regeneration of Tubular in 229 Concentration (moles/liter) Number of stems Time (hours) Rate Length Length/time I// % of con- trol Micra % of con- trol Lit % of con- trol Ethyl urethane 1. Control . 23 24 23 19 22 22 25 17 15 18 17 11 10 17 18 18 17 19 18 14 9 17 25 20 16 20 15 16 16 20 17 18 21 15 14 14 24 14 17 18 24 11 6 11 38.1 37.0 38.8 39.6 49.7 56.9 60.7 39.8 38.8 38.6 52.8 59.3 71.5 43.1 48.4 44.8 45.4 49.5 44.5 49.1 75.0 59.6 46.7 48.9 57.7 63.4 67.6 68.3 68.6 34.8 30.3 33.8 37.7 51.4 53.8 66.2 26.6 53.5 43.0 46.0 48.3 45.7 112.0 100.0 0.0262 0.0271 0.0257 0.0253 0.0201 0.0170 0.0165 0.0251 0.0257 0.0259 0.0189 0.0168 0.0140 0.0232 0.0207 0.0225 0.0221 0.0202 0.0226 0.0204 0.0137 0.0168 0.0214 0.0205 0.0173 0.0158 0.0148 0.0147 0.0146 0.0288 0.0333 0.0296 0.0266 0.0195 0.0186 0.0151 0.0376 0.0187 0.0233 0.0217 0.0207 0.0219 0.0089 0.0100 100.0 104.0 98.0 96.5 76.7 65.0 64.0 100.0 102.0 103.0 75.2 67.0 55.7 100.0 89.0 96.9 95.1 87.0 97.3 87.8 59.0 100.0 127.0 122.0 103.0 94.1 88.0 87.5 87.0 100.0 116.0 103.0 92.4 67.6 64.6 52.4 100.0 49.6 61.9 57.6 55.0 58.1 23.6 26.5 908 958 957 957 928 989 1019 963 1009 989' 922 1016 981 949 1038 972 986 974 1045 888 764 812 773 914 884 858 767 800 757 1009 1100 963 822 851 536 380 1150 920 1002 1006 878 935 560 590 100.0 104.0 104.0 104.0 102.0 109.0 112.0 100.0 105.0 103.0 95.6 106.0 102.0 100.0 110.0 102.0 103.0 102.0 110.0 93.5 80.5 100.0 94.1 111.0 107.0 105.0 93.5 97.5 92.1 100.0 109.0 95.4 81.5 84.2 53.1 37.6 100.0 80.0 88.0 88.0 76.3 81.3 48.7 51.2 23.8 25.9 24.7 24.2 18.6 17.4 16.8 24.2 26.0 25.6 17.5 17.1 13.8 22.0 21.4 21.6 21.7 19.7 23.6 18.1 10.2 13.8 16.5 18.7 15.3 13.5 11.3 11.7 11.0 28.9 36.3 28.2 21.8 16.6 10.0 5.7 43.2 17.2 23.1 21.7 18.1 20.4 5.0 5.9 100.0 108.0 104.0 102.0 77.8 72.8 70.0 100.0 107.0 103.0 72.0 70.5 57.0 100.0 97.3 98.0 98.9 89.9 108.0 82.1 46.5 100.0 120.0 133.0 111.0 98.0 82.0 84.9 79.8 100.0 122.0 98.0 75.0 57.3 34.5 19.7 100.0 39.9 53.4 50.2 41.9 47.1 11.6 13.7 1 X ID"3. 5 X 1C-3 8 X 10~3. . 1 X ID"2 2 X 10-2 8 X 1C-2 2. Control 5 X ID-3. . 1 X 1C-2. . 2 X 10-2. 3 X ID"2 4 X ID"2 3. Control . . . 1 X 10~7 1 X 10~6 1 X 10~5 1 X 10~4 1 X 10-3 1 X 10~2 5 X 10"2 Phenyl urethane 4. Control 1 X 10~4 1 X 10-3. . 1 X JO"2 2 X ID"2 3 X ID"2 4 X ID"2 5 X 10-2. 5. Control 1 X ID"3 1 X 10-2 2.5 X 10~2.. 4 X 10~2 5 X ID"2 6.5 X 10~2 Chloretone 6. Control 1 X 10-2. 1.25 X 10-2.... 1.50 X ID"2. 1.75 X ID-2.... 2 X10-2.... 2.25 X ID-2.... 2.50 X 10-2. . . . 230 SPIEGELMAN AND M( >< Ml generation rate (L/t) until the concentration reaches about 0.1 molar. Although the sensitivity does vary from group to group it is evident that the major effect is on the time to constriction. This is made strikingly apparent by figure 1, which is a plot of both primordium length and time to constriction, expressed as per cent of control, against the logarithm of the molar concentration multi- plied by 106. Here over a concentration range which produces a 44 per cent decrease in the 1/7 factor, the lengths of the regenerating primordia remain unaffected. 120 100 80 CL O 60 40 20 ETHYL URETHANE 3.0 3.2 3.4 3.6 3.8 4.0 4.2 ln(C- I06) 4.4 4.6 4.8 5.0 FiGi'Ki: 1. The effect of ethyl urethane on rate (1/0 of regeneration (open circles) and on length of the regenerating primordium (half-closed circles). Data from experiment 2, Table I. With both phenyl urethane and chloretone the results are quite different. It is evident from Table J that within the concentration range in which decreases in regeneration rate are obtained, the inhibition involves comparable decreases in both the L and \/t factors. The difference between the data obtained with these two narcotics and that obtained with ethyl urethane is illustrated by figure 2, which represents the data of experiment 6, with chloretone. In comparison with the effects of ethyl urethane, the parallelism of effects here is clear. /I Cyanide, uziile, and oxygen tension. Tables II and III summarize the data obtained with these reagents. The regeneration rate L/t is extremely sensitive to even relatively low concentrations of cyanide. Thus 6 X 10~(1 molar cyanide caused a 17 per cent reduction in L/t, and 5 ; ; 10~5 molar a 61 per cent reduction. However it will be noted that these reductions were due almost entirely to diminishing \/t values. This is illustrated by figure 3, which is a plot of the data of experiment 8; there it may be seen that REGENERATION AS MASS AND TIME 231 in a concentration range which yielded a 61 per cent decrease in the 1/t factor the length was affected only to the extent of six per cent. The use of higher concentrations however led quickly to drastic reductions in the amount of tissue transformed. In the case of azide in the range from 1 X 10~6 molar to 2 X 10~3 molar, the differential effect on the factors of the rate was not as clear-cut as in the case of cyanide. There was again however a tendency for the length to be less sensitive to lower concentrations than \/t. Thus in experiment 10, 7 X 10~4 molar azide decreased 1/t by 43 per cent and length by eight per cent. In experiment 12, the same concentration resulted in a 38 per cent decrease in 1/t and a 12 per cent decrease in length. 20 00 ,3 3 CHLORETONE 4.0 4.2 4.3 ln(OI06) 4.4 4.5 FIGURE 2. The effect of chloretone on rate (\/f) of regeneration (open circles), and on length of the regenerating primordium (half-closed circles). Data from experiment 6, Table I. Both cyanide and azicle presumably act by poisoning the cytochrome-cyto- chrome oxidase system. A study of their effects on both respiration and recon- stitution (Moog and Spiegelman, 1942) has however indicated that they follow different pathways in depressing the regeneration process, for cyanide inhibition of reconstitution was always accompanied by a strong depression of the respira- tory rate, whereas azide, in a concentration which invariably cuts the reconstitu- tion rate by at least 80 per cent, scarcely altered the rate of oxygen uptake at all. The data presented in Table II further indicate that the ultimate effect of azide differs from that of cyanide. In the case of cyanide a 20 per cent decrease in length is accompanied by a 70 per cent increase in time to constriction, but a 232 SPIEGELMAN AM) MOOG TABLE II The effects of cyanide and azide on regeneration of Tubular ia Concentration (moles/liter) Number of stems Time (hours) Rate Length Length/time Ht % of con- trol Micra % of con- trol Lit % of con- trol Sodium cyaniik- * 7. Control . . 20 36.4 0.0275 100.0 1175 100.0 32.1 100.0 5 X 10-*.... 16 62.9 0.0159 57.7 1025 87.1 16.3 50.7 6.5 X 10-5. . . . 18 76.0 0.0132 48.0 1039 88.2 13.7 42.6 8 X 10-6. . . . 19 88.6 0.0113 40.4 1105 94.0 12.5 39.0 9.5 X 10~5. . . . 18 117.6 0.0085 30.9 953 81.0 8.1 25.2 1.2 X ID-4.... 24 128.0 0.0078 28.4 567 48.3 4.4 13.7 X. Control 19 26.2 0.0382 100.0 1176 100.0 44.9 100.0 1 X lO-6.... 19 27.1 0.0370 97.0 1193 102.0 44.1 98.3 6 X ID-6.... 18 30.8 0.0325 85.1 1146 97.4 37.2 82.3 2 X 10-5. . . . 16 37.2 0.0269 70.5 1170 99.4 31.4 70.0 5 XlO-5.... 20 63.9 0.0157 41.1 1103 93.9 17.3 38.5 7.5 X 10-5... . 17 123.6 0.0081 21.3 889 76.1 7.3 16.2 9 X 10-*. . . . 13 168.4 0.0060 15.7 445 37.8 2.6 5.8 <).5 X 10~5. . . . 12 172.2 0.0058 15.2 380 32.2 2.1 4.7 i xio-4.... 20 209.5 0.0048 12.6 141 12.0 0.7 1.6 Sodium azide 9. Control 19 46.4 0.0215 100.0 986 100.0 21.3 100.0 2 X lO"6 18 37.4 0.0268 125.0 1003 102.0 26.8 126.0 7 X 10~6 17 41.6 0.0241 1170 1013 115.0 24.3 114.0 2 X 10~5 13 43.7 0.0229 1 1 •- 1 • \J 106 0 1013 102.0 23.0 108.0 7 X ID"5 18 48.5 0.0206 1 \J\s.\J %i 948 96.0 19.5 91.5 2 X 10-" 14 71.9 0.0139 . 1 As A 803 81.9 11.2 52.6 7 X 10~4 15 92.6 0.0109 \J v ' . \f SO R 710 72.0 7.7 36.2 10. Control 18 29.6 0.0338 J\ 1 , 0 100 0 1349 100.0 45.2 100.0 1 X lO-6. . . . 17 34.0 0.0294 1 \J\Jt\J 75.8 1200 89.0 35.3 78.0 1 X 10-5. . . . 17 49.5 0.0201 51.8 1162 86.4 23.5 51.9 1 X 10-4. ... 17 31.8 0.0314 81.0 1280 95.0 40.2 88.9 4 X 10-4.... 18 36.5 0.0274 70.5 1278 94.6 34.8 76.8 7 X 10-4. ... 17 45.1 0.0222 57.1 1244 92.3 24.4 53.9 9 X10~4.... 13 51.2 0.0195 50.1 1244 85.0 18.8 41.5 9.5 X 10-'.... 18 55.9 0.0179 46.0 1029 76.2 15.3 33.8 1 X 10-3.... 11 56.7 0.0177 46.0 1031 76.5 12.1 26.7 1.3 X 10-3... . 10 62.1 0.0161 41.5 989 73.4 9.1 20.1 11. Control 19 36.6 0.0274 100.0 1138 100.0 31.8 100.0 5 X 10-5. . . . 18 33.7 0.0298 108.0 1079 94.6 31.5 99.2 2 X 10~4. . . . 18 34.8 0.0288 105.0 994 87.2 28.2 88.7 6 X 10~4... . 18 40.6 0.0246 90.0 1035 91.0 25.7 80.9 9 X 10 4... . 19 55.6 0.0180 65.5 909 79.7 16.3 51.3 1.5 X 10-3.... 13 76.1 0.0132 48.1 669 58.7 11.4 35.8 12. Control 18 30.6 0.0328 100.0 1162 100.0 38.0 100.0 2.5 X 10-4. ... 17 38.6 0.0259 79.0 1121 96.1 29.8 80.4 4 X 10-4.... 19 37.6 0.0266 81.0 1059 90.8 28.1 78.4 5.5 X 10- '.. .. 20 41.4 0.0242 74.0 1042 89.5 25.2 70.4 7 X 10-4. .. . 19 49.4 0.0203 62.1 1024 88.0 20.7 57.8 8.5 X 10-'.... 19 44.6 0.0224 68.1 886 76.1 19.9 55.5 1 X 10 17 45.7 0.0219 66.9 897 77.0 19.8 55.3 2 X Hi 12 62.9 0.0159 48.5 314 27.0 5.0 13.9 REGENERATION AS MASS AND TIME 233 TABLE III The effect of varying oxygen tensions on regeneration of Tubularia (Data from Earth, 1938V) Oxygen tension (cc./liter) Number of stems Time (hours) Rate Length Length/time \lt % of con- trol Micra % of con- trol Lit % of con- trol 13. Control*. . 2.4 20 20 20 20 20 20 20 20 20 26.2 36.1 28.1 26.8 26.3 24.5 24.6 24.1 23.7 0.0382 0.0277 0.0263 0.0374 0.0381 0.0408 0.0407 0.0415 0.0422 100 72 93 98 100 107 106 106 111 1370 1072 1284 1370 1365 1640 1809 1840 1846 100 78 94 100 99 120 132 134 135 52.3 29.9 45.8 51.1 52.0 67.0 73.5 76.5 77.9 100 52 83 93 99 128 140 146 148 3.2 4.1 . . . 4.8 8 2 11.3 ... 14.3 16.5 * Dish open to air. comparable decrease in length by azide poisoning yields only about a 35 per cent increase in time. In considering the role the oxygen-utilizing system plays in regeneration, it is of interest to examine, from the point of view of this paper, Earth's (1938b) < 80 rr O z 60 40 20 POTASSIUM CYANIDE 0.0 0.4 0.8 1.2 1.6 ln( I06) 2.0 2.4 FIGURE 3. The effect of cyanide on rate (\/t) of regeneration (open circles), and on length of the regenerating primordium (half-closed circles). Data from experiment 8, Table II. 234 SPIEGELMAN AND MOOG data on regeneration rates at various oxygen tensions. Table III gives the calcu- lations made from Earth's experiment 7 on young stems comparable to the ma- terial used in the present study. There are not enough data on the effect of low oxygen tensions to determine definitely whether unavailability of oxygen acts in the same way as cyanide. The indication is however that cyanide involves other factors, since at 2.4 cc. of oxygen per liter a 22 per cent decrease in length is accompanied by only a 28 per cent increase in time to constriction. The interesting fact to emerge from Table III, in any case, is that at high oxygen TABLE IV The influence of temperature on regeneration of Tubularia Temperature °C. Number of stems Time (hours) Rate Length Length/time lit % of con- trol Micra % of con- trol Lit % of con- trol 14. 20.5 20 25 21 19 19 21 24 23 17 21 22 23 24 18 18 21 25 26 20 15 17 30.8 35.1 47.7 57.2 95.2 39.4 43.4 50.0 64.5 83.4 40.0 47.6 57.2 66.6 100.0 37.8 45.6 48.8 55.5 71.5 100.0 0.0325 0.0285 0.0210 0.0175 0.0105 0.0250 0.0230 0.0200 0.0155 0.0120 0.0250 0.0210 0.0175 0.0150 0.0100 0.0265 0.0238 0.0205 0.0180 0.0140 0.0100 114.0 100.0 73.8 61.4 36.8 109.0 100.0 86.9 67.4 42.2 119.0 100.0 83.5 71.5 47.6 111.0 100.0 86.1 75.6 58.9 42.0 1100 1120 1280 1400 1460 1020 1031 1180 1290 1380 1010 1174 1210 1260 1320 1010 1082 1120 1180 1380 1410 98.5 100.0 114.0 125.0 130.0 99.0 100.0 115.0 125.0 134.0 86.2 100.0 103.0 107.2 112.3 93.5 100.0 103.3 109.0 127.0 130.2 35.8 32.0 26.9 24.5 15.3 25.3 23.8 23.6 20.0 16.6 25.5 24.6 21.2 18.9 13.2 26.8 23.2 23.0 21.2 19.3 14.1 111.9 100.0 84.0 76.7 48.8 106.0 100.0 99.1 84.0 69.6 103.4 100.0 86.4 77.8 53.6 115.5 100.0 99.0 91.5 83.3 60.8 18.7* 13.5 10.8 7.0 15. 20.5 18.7 13.5 10.8 7.0 16. 20.5 18.7 13.5 10.8 7.0 17. 20.5 18.7 160 13.5. 10.8. 7.0 * 18.7° was chosen as the control temperature since it is closest to the natural optimum of the material. tensions the increased regeneration rate, which can go as high as 148 per cent of normal, results in major part from increases in the mass of tissue transformed. Thus at 14.3 cc. /liter the L factor is 34 per cent above normal, whereas the I// factor is increased only six per cent. This is in sharp contrast to the effect of cyanide, which over a wide range influences the regeneration rate by changing the I// factor while leaving the L factor relatively unaffected. C. Temperature The most striking exhibition of the independence of the length and time factors emerges from the data on the influence of temperature on regeneration, REGENERATION AS MASS AND TIME 235 40 20 <100 cc o Z 80 60 40 8 0 8 20 22 12 14 16 TEMPERATURE °C FIGURE 4. The effect of temperature on rate (1/0 of regeneration (half-closed circles), and on length of the regenerating primordiurn (open circles). Data from experiment 17, Table IV. as summarized in Table IV; the actually opposite effects of temperature on the two rate components are illustrated in figure 4, which is a graph of experiment 17. It is of interest for later discussion to note that, in addition to moving in different directions, the two factors are independently sensitive to temperature changes. Table V shows the <2io values calculated from 7° to 22° C. in five degree intervals. Each value in the "average" column was obtained from the results of four separate determinations. The average value over the entire range is also noted. The high values in each set of four experiments are included in the table to give an estimate of the upper limit of sensitivity for the two factors. TABLE V Temperature coefficients (Qio) for length and rate in the regeneration of Tubularia Temperature Length Rate (1/0 Average High Average High 7-12 1.20 1.28 2.14 2.63 12.1-17 1.33 1.37 1.82 1.97 17.1-22 1.16 1.25 1.57 1.76 Average 1.23 1.30 1.88 2.12 236 SPIl-XiKLMAX AXD MOOG It is evident that the Qio values for l/t are consistently higher both in the average and in the highest limits attained than those for L. It will also be noted that Qio for length scarcely changes from low to high temperatures, while on the other hand the coefficient of the \/t factor drops 0.57 from the lowest five degree interval to the highest. Attempts have been made to explain such differences by invoking two separate processes unlike in their temperature sensitivities, one controlling the mass of tissue transformed and the other the time to constriction. Yet fundamentally length and time may be merely measurements of separate aspects of the same process. From tliis point of view it is more likely that the true explanation of the different responses to temperature is to be found in the purely numerical character of the terms which determine the two magnitudes, and in the way these magnitudes depend on parameters which vary with temperature. We hope to show in the discussion that this view is quite plausible. For the moment, how- ever, it is sufficient to point out that the independent responses to temperature of L and / serve to emphasize further the independence displayed by the temperature coefficients. ^ DISCUSSION That the mass of tissue involved in Tubularia reconstitution is independent of the time to constriction of the new primordium has been demonstrated with a variety of effective agents. It is evident then that L/t cannot be used as a measure of rate under different conditions unless it is accompanied by separate analyses of the behavior of the two components of the ratio. Thus for example the L/t ratio might be found constant over a range of temperature because of inversely proportionate effects on length and 1/7. It is equally evident that the solution of the problem will not be reached by ignoring one or the other of the factors; for example, I// would not constitute an adequate description of the effect of high oxygen tension on the regenerative process. From another point of view L/t, despite its correct dimensionality, may be expected to prove to be a relatively inaccurate measure of transformation rate. For the / in the definition is unlike similar factors in ordinary rate formulae, but is unique in the sense that it is determined by a stage in the development of the system. Its use implies the possibility of measuring rate by taking only two points (zero time and the time to the stage chosen) on the transformation-time curve and using the slope of the line connecting the two points as the rate of the process. Accuracy under such circumstances would be obtained only if the transformation-time curve were perfectly linear, i.e., if the rate up to the stage chosen were constant. In the case of a non-linear curve the approximation would become more and more crude with increasing deviation from linearity as well as with increasing distance between the selected points. Thus the approximate nature of L/t as a rate measure may be easily recog- nized; the reasons for the relative independence of the length and time com- ponents, however, are not so evident. An insight into some of the factors in- volved may perhaps be gained by examining an attempt to measure the course of a simple chemical reaction by the same method, i.e., selecting only two points for observation, one at / ---- 0 and the other close to the end of the process, when the system becomes time-independent. So let us assume the following REGENERATION AS MASS AND TIME 237 transformation S-II, (1) the forward and backward reactions having velocity constants of k and k' re- spectively. Let 5 and h represent the initial concentrations of S and H, and x the number of moles of 5 transformed into H in / minutes. Then at the end of t minutes, (s -- x) is the concentration of 5 and (h + x) is the concentration of H. After suitable rearrangements the transformation rate at any moment is given by dx •^=(k1-- k'h) -- (k + k'}x. (2) Equation (2) may be integrated to yield the complete time course of the trans- formation, which takes the form x ••-- A •- Ae-w*', (3) where - = ^ It is evident from equation (3) that as / increases the exponential term becomes smaller and x approaches A, which represents its equilibrium value. Generally in measuring regeneration rates, and particularly in hydranth reconstitution, the stage chosen is one close to the end of the process, beyond which no further significant transformation occurs. A comparable /, in the simplified system being examined, would be one sufficiently large to make the exponential term nu- merically negligible. Let such a particular / be represented by T. At such time the amount of H present, a measure which would be comparable to the L in the regeneration rate formula, would be given by (h -\- A}. The "rate" analagous to L/t then takes the form h + A rate = - -jr~ • (5) Examination of equations (3) and (4) reveals why the components of a rate so determined may be independent. The magnitude of / which will reduce the value of Ae~(k+k>)t sufficiently to make x time-independent obviously depends on the magnitudes of A and (k + k'). Any experimental procedure which either increases or decreases k and k' proportionately will leave A, and consequently (h + A), undisturbed, but will change the / (i.e., T) necessary to reduce the second term to insignificance. Under such conditions, the rate as defined by equation (5) would vary solely because of a changing denominator, the mass factor in the numerator remaining constant. On the other hand, an experimental procedure that varied s, the initial concentration of S, might from an observa- tional point of view affect only the mass, since A is a function of 5. Strictly speaking, any change in A also influences the T value, since A is included as a factor in the time-determining term. However, in any combination of an or- dinary algebraic and an exponential factor, the latter quickly predominates in determining the numerical value of the product. Thus, unless the variation in 5 produced a very marked change in A, the T values before and after the change 238 SPIEGELMAN AND MO0G might not be experimentally distinguishable, even though the difference in the A values were easily detected. It is hardly conceivable that the complex of processes leading to the reconsti- tution of a hydranth can have much resemblance to the simple reaction repre- sented by equation (1). About the only properties the two have in common are that they both involve molecular transformation and that they both take time to arrive at a time-independent state. The fact that the "rate" of the chemical reaction as defined by equation (5) exhibits many of the characteristics experi- mentally found for L/t is most likely inherent in the basic similarities underlying the two definitions. The approximate nature of both rate formulations would tend to conceal any differences in the processes they are used to measure. In any case, the above analysis strongly suggests the possibility that some of the peculiarities of the L/t rate found experimentally with the various reagents may in part be characteristic of the definition rather than of the mechanism of regeneration. It is worthy of note that the decrease in L observed at higher temperatures is not shown by the mass factor of the rate described by equation (5). The numerical reason for the constancy of the numerator with temperature variation is that a temperature change can only yield proportionate changes in the forward and backward velocity constants; since these appear to the same powers in both the numerator and the denominator of A, the net result is that A remains con- stant. Underlying this behavior is the fundamentally important fact that equa- tion (1) represents a reaction occurring in a closed system and as such its equi- librium point, as far as the concentration of reactants is concerned, is independent of temperature. If the temperature variation of L is to be examined, therefore, it is necessary to study an open system whose time-independence is maintained by a constant How of material or energy through it. As has already been pointed out by Burton (1939), open systems are far more likely than closed systems to possess kinetic characteristics typical of living organisms, simply because the latter are themselves open systems, and approach steady states rather than true thermo- dynamic equilibria. If instead of equation (1) we introduce a source 0 for S, and a sink P for H, the system becomes an open one, since the concentrations of 5" and H now become dependent on parameters external to the transformation, namely the levels of O and P. For purposes of simplicity we shall assume that the back reaction k' is either zero or at least negligibly small as compared with the forward reaction. This is plausible whether we consider the S to H trans- formation itself the energy-yielding reaction which leads to hydranth synthesis, or whether we regard the transformation as being driven by some other energy- yirlding reaction. In the first case the transformation would tend to be relatively irreversible, in the second the coupled energy-yielding reaction would tend to make the reverse reaction from // to 5 relatively insignificant. Instead of (1) then we may write k0 k kp O^S-^H^P. (6) The velocity constants, k0 and kp, connecting 0 to 5 and H to P respectively, are taken to represent both the forward and the backward velocities of the two REGENERATION AS MASS AND TIME 239 reactions they govern. This assumption of equal forward and backward ve- locities, while not necessary for the analysis, avoids the undue complication that would result from too many constants. In addition, if O represents a type of source in which 5-substrate diffuses from 0 to the site of the reaction, and P represents a type of sink toward which the produced //"diffuses, then the assump- tion of equality would exactly describe the situation. Letting C0, Cs, Ch, and Cp represent the concentration levels of 0, S, H, and P, the following equation may be written for the kinetics of the transformation of S into H i /"* ~ = k0Ca -- Cs(ku + k}. (7) Integrating equation (7) yields the time variation of the concentration of S, which is given by C = B + Ge-^+v, (8) where R = T^I <9> ko + R and G-.-(CSo--B), (10) Cs<> being the concentration of 5 at zero time. Equation (8) is formally identical with equation (3). It is readily seen from equation (8) that B represents the time-independent or steady state value of C, and is analagous to A of the previous case. Setting up for the present system a rate similar to L/t and equation (5), we may write Rate = j. (11) in which T has the same significance as in (5). Thus again, by the same arguments used in the analysis of a reaction going to equilibrium, which need not be repeated here, it becomes evident that differ- ential effects on either mass or T factor in (11) may be obtained. Thus a suitable experimental procedure which affected only C0 would manifest itself by marked changes in B and weak changes in T. Such may well be the explanation for the results of Barth's experiments on the effects of high oxygen tension, in which the length was increased strongly and the T factor very little. The implication that high oxygen tension raises the level of the source for 5 (i.e., C0) fits into the hypothesis proposed by Barth (1940) that oxygen is directly involved in the synthesis of a substance 5 whose transformation yields hydranth. Again as in the previous case, treatments which affect the velocity constants k0 and k would result in marked variations in T as compared with B. On the basis of this analysis, one would then interpret the results with ethyl urethane, sodium cyanide, and sodium azide in terms of decrease of the values of velocity constants by poisoning of the enzymes involved in the transformation. Phenyl urethane and chloretone on the other hand, in addition to decreasing the velocity constants, also appear to lower the C0 value by interfering with the synthesis of the substance 51 or its immediate precursors. 240 SPIEGELMAX AND MOOG Since increases in temperature raise the values of velocity constants, it is evident why lower T values are found at higher temperatures. A possible reason for the decrease of mass transformed with increasing temperatures may be found in an examination of the effects of variations in k0 and k\ on B, the steady state value of C. The change in B for variations in the velocity constants is given by Ct(kdk0 - k0dk) dB = - ,, , N, (ko -\- ki)~ \Ye are concerned here with temperature increases; consequently both dk0 and dk will be increments, i.e., positive quantities. All other factors of the right-hand member of (12) being positive, it is clear that dB will be either positive or negative according as (kdka - - k0dk) is either positive or negative. Thus the mass factor will decrease with increasing temperature if kdko -- k0dk < o. (13) Inequality (13) can be satisfied in several ways. Thus for example, if k0 were of the nature of a diffusion coefficient and k the velocity of a chemical reaction, then for a given rise in temperature the increment in k0 would be about half that realized by k. Inequality (13) then becomes — k0 } < 0 and is satisfied if k < 2k0. If on the other hand the two processes have the same temperature coefficients, so that dk were equal to dk0, inequality (13) could be satisfied if k0 were greater than k. Whatever the intimate details of the situa- tion may actually be, it is evident that opposing responses of mass and time found in regeneration may plausibly be explained as an expression of the opera- tion of an open system approaching a steady state. Decreases of size with in- creasing temperature are not confined to regeneration in Tubularia, but are a general phenomenon of development and have been studied in other forms in- cluding the trout (Gray, 1928; Merriman, 1935), the whitefish (Price, 1940), and the frog (Chambers, 1908). The different sensitivities to temperature of length and time apparent in the <2io values given in Table V find their most plausible explanation in the algebraic composition of the two terms that determine them. We have already noted that in the special case in which the <2io values of k0 and k are equal, the Qio for the mass would be unity since proportional increases in these constants would cancel out and leave B unchanged for any temperature rise. On the other hand 7", since it is determined by the sum of k0 and k, would be affected more or less strongly according to the magnitude of the temperature change. This same difference in response will be carried over to the more general case where the i emperature coefficients of k0 and k differ. By the very nature of the dependence of B on these constants, increases or decreases in the constants cannot result in as marked changes in B as they would in T, which is exponentially dependent on their sum. REGENERATION AS MASS AND TIME 241 It is apparent from this discussion that neither \/t nor L/t can be treated as ordinary rates. However the L/t definition of regeneration rate, if supplemented by a further analysis of the separate behavior of L and 1/t, can yield interpretable information on the regeneration process. The omission of the mass factor is surely not justifiable on the grounds of "correcting" the L/t definition. The latter will in many cases yield information which the "corrected" rate would miss entirely. SUMMARY 1. Data are presented which show that various agents produce differential effects on the length of the regenerating primordium of a Tubularia hydranth and on the time to the constriction of the primordium from the rest of the stem. 2. The significance of this independence of length and time for the L/t formulation of regeneration rate is discussed. 3. The differential effects are interpreted in terms of a reaction approaching a steady state in an open system. 4. The criticisms of the L/t definition and the proposed substitution of l/t are discussed in terms of the above analysis. It is concluded that the L/t defini- tion, if supplemented by a further analysis of the independent behavior of L and l/t, provides a useful and informative measurement of regeneration. LITERATURE CITED BARTH, L. G., 1938u. Quantitative studies on the factors governing the rate of regeneration in Tubularia. Biol. Bull., 74: 155-177. BARTH, L. G., 1938b. Oxygen as a controlling factor in the regeneration of Tubularia. Physiol. Zool., 11: 179-186. BARTH, L. G., 1940. The process of regeneration in hydroids. Biol. Rev., 15: 405-420. BURTON, A. C., 1939. The properties of the steady state compared to those of equilibrium as shown in characteristic biological behavior. Jour. Cell. Comp. Physiol., 14: 327-349. CHAMBERS, R., 1908. Einfluss der Eigrosse und der Temperatur auf das Wachstum und die Grosse des Frosches und dessen Zellen. Arch. mikr. Anal., 72: 607^ CHILD, C. M., 1940. Patterns and Problems of Development. University of Chicago Press. GRAY, J., 1928. The growth of fish. III. The effect of temperature on the development of the eggs of Salmo fario. Brit. Jour. Exp. Biol., 6: 125-130. MERRIMAX, D., 1935. The effect of temperature on the development of the eggs and larvae of the cut-throat trout (Salmo clarkii clarkii Richardson). Jour. Exp. Biol., 12: 297-305. MILLER, J. A., 1942. Some effects of covering the perisarc upon Tubularian regeneration. Biol. Bull., 83: 416-427. MOOG, F., 1942. Some effects of temperature in the regeneration of Tubularia (abstract). Biol. Bull., 83: 291. (Also Collecting Net, Woods Hole, 17: 72-73.) MOOG, F., AND S. SPIEGELMAN, 1942. Effects of some respiratory inhibitors on respiration and reconstitution in Tubularia. Proc. Soc. Exp. Biol. Med., 49: 392-395. PRICE, J. W., 1940. Time- temperature relations in the incubation of the white fish, Coregonus clupeaformis (Mitchell). Jour. Gen. Physiol., 23: 449-468. SPIEGELMAN, S., 1942. Mass and time relationships in the regeneration of Tubularia (abstract). Biol. Bull., 83: 291-292. (Also Collecting Net, Woods Hole, 17: 73-74.) SPIEGELMAN, S., AND A. GOLDIN, 1944. A comparison of regeneration and respiration rates of Tubularia. Proc. Soc. Exp. Biol. Med., 55: 252-253. NEUROSECRETION VI. A COMPARISON BETWEEN THE INTERCEREBRALIS- CARDIACUM-ALLATUM SYSTEM OF THE INSECTS AND THE HYPOTHALAMO-HYPOPHYSEAL SYSTEM OF THE VERTEBRATES x BERTA SCHARRER AND ERNST SCHARRER The Department of Anatomy, Western Reserve University, and the Marine Biological Laboratory, Woods Hole, Massachusetts A comparison of data on the secretory activity of nerve cells in invertebrates with those obtained from corresponding studies in vertebrates revealed an interest- ing parallelism between the intercerebralis-cardiacum-allatum system of insects and the hypothalamo-hypophyseal system of vertebrates. The functional mechanism in- volved cannot be fully explained at present ; but the observations are in themselves intriguing and offer a point of departure for the discussion of certain neuro- endocrine relationships. INSECTS In the larvae of muscoid Diptera the ring-gland, an endocrine organ concerned with development (Haclorn, 1937; Hadorn and Neel, 1938; Burtt, 1938; Becker and Plagge, 1939; Vogt, 1942a; Gloor, 1943; Bodenstein, 1943a, 1943b, 1944), contains the elements of two glands, the corpus cardiacum and the corpus allatum (Scharrer and Hadorn, 1938; Vogt, 1942b ; Day, 1943; Poulson, 1944). In other insects these two components form more or less individual organs. In Leucophaea tnadcrae, a species used in the present study, the corpora cardiaca and allata are paired organs which, as in other representatives of the Orthoptera (De Lerma, 1937; Hanstrom, 1940), lie dorsal to the esophagus behind the brain. The ante- rior portions of the elongate corpora cardiaca form part of the wall of the dorsal blood vessel. The posterior ends of the corpora cardiaca are in contact with the corpora allata, which lie more laterally than the former. In Leucophaea the two glands are not, as in other species, separated by a nervus corporis allati but constitute an almost continuous mass of glandular tissue. Histologically the corpus cardiacum can be easily differentiated from the corpus allatum. Cardiacum tissue contains nervous as well as glandular elements, whereas there is no indication of a nervous component in the corpus allatum. In Leucophaea the corpus cardiacum is to a varying degree filled with deeply staining colloid masses ; in older specimens the gland may be replete with such acidophil substances. By comparison little material that can be interpreted as a secretory product is, as a rule, found in the corpus allatum. The physiological significance of the variations of the histological appearance of both the corpora cardiaca and corpora allata is not clear at present ; but it is evident that they are both glands. 1 This research was ,-iided by a grant made to Western Reserve University by the Rockefeller Foundation. 242 NEUROSECRETION 243 The corpora cardiaca receive a well defined fiber bundle (nervus corporis car- diaci, Pflugfelder, 1937; nervus corporis cardiaci I, Hanstrom, 1940; nervus occipi- talis, Nesbitt, 1941) from the pars intercerebralis of the protocerebrum. In Leu- cophaea the fibers turn from their origin antero-medially and downward. Most of them, perhaps all, cross in the midline and continue toward the base of the brain. Thence the fiber bundle turns backward and shortly after leaving the brain enters the corpus cardiacum. The bundle can be followed all the way through the gland, which it innervates ( Fig. 1 ) . It seems that some of the fiber components enter the corpus allatum of the same side where they are distributed. COQPUS ALLATUM . PARS INTZQCCQEBBALIS CAQDIACUM CORPO&IS CAKDIACI FIGURE 1. Diagram of the intercerebralis-cardiacum-allatum system of an insect. The pars intercerebralis of the insect brain is distinguished by the occurrence of secreting nerve cells. Such cells have been found in Hymenoptera (Weyer, 1935; Scharrer, 1937), Hemiptera (Hanstrom, 1938; Wiggles worth, 1940), Lepidoptera (Day, 1940a), Coleoptera, Neuroptera, Trichoptera, and Diptera (Day, 1940b; Vogt, 1942a).- In the Orthopteran Leucophaea madcrac, for instance, the medium sized nerve cells of the pars intercerebralis contain distinctly staining inclusions varying in size and number. There may be only two to three granules present in one cell, or they may be so numerous that they fill the entire cell body. Not only the number of granules in different cells, but also the number of secreting elements varies in different specimens. There may be numerous cells on either side of the midsagittal plane containing granules or there may be only a few such cells. The secretory material is in some cases concentrated near the axon hillock, and may continue for a certain distance into the axis cylinder, which in this case appears wider than in nerve cells of comparable size without secretory granules. - R. P. Holdsworth found neurosecretory cells in the pars intercerebralis of Pteronarcys, a representative of the Plecoptera (personal communication). 244 srilARRER AND SCHARRER The colloid granules in the pars intercerebralis resemble in size and stainability those found and previously described in the neurosecretory cells of the subesophageal ganglion ( Scharrcr, 1941a). The neuroglandular cells are larger in the subesopha- geal ganglion than in the pars intercerebralis. Along the fibers of the nervus corporis cardiaci colloid masses are found in varying, sometimes very great amount, particularly in older specimens of Leu- cophaea. The colloid content of the fiber bundle permits its tracing and differenti- ation from other tracts (Fig. 1). The concept of an anatomical system formed by the pars intercerebralis and the corpora cardiaca and allata facilitates the understanding of the hormonal regulation (if postembryonic insect development. Several hormones derived from different sources have been demonstrated to control growth and differentiation in various groups of insects. The corpus allatum is known to furnish an "inhibitory hor- mone" in Hemiptera (Wigglesworth, 1934, 1936, 1940), Orthoptera (Pflugfclder, 1937; Pfeiffer, 1942; Scharrer, 1944), Lepidoptera (Bounhiol, 1939; Piepho, 1943), and Coleoptera (Radtke, quoted from Piepho, 1943). A substance originating in the brain brings about molting in Hemiptera (Wigglesworth, 1940), and pupation in Lepidoptera (Kopec*, 1922; Caspari and Plagge, 1935; Kiihn and Piepho, 1936) and Hymenoptera (Schmieder, 1942). Additional centers in the thorax (or pos- sibly upper abdomen) have been claimed to play a role in pupation and imaginal differentiation of Lepidoptera (Hachlow, 1932; Bounhiol, 1938; Bodenstein, 1938; Fukuda, 1940; Piepho, 1943) and Neuroptera (Ochse, 1944). In the highly spe- cialized muscoid Diptera the ring-gland containing both corpus cardiacum and al- latum controls growth, molting (Bodenstein, 1944), pupation (Hadorn, 1937; Hadorn and Neel, 1938; Becker and Plagge, 1939; Vogt, 1942a), and imaginal differentiation (Bodenstein, 1943b), whereas the brain is said to have no influence on these processes. There is indirect evidence that at least the substance causing puparium formation is produced by the cardiacum component of the ring-gland (see Scharrer, 1941b). Finally, removal of the corpora cardiaca in Orthoptera causes a retardation of molting (Pfeiffer, 1939). In an attempt to reconcile some of the seemingly divergent data it may be useful to discuss first the various hormones named, and second their source in the organism. Concerning the hormones controlling postembryonic insect development two interpretations are possible: (a) Each developmental step is brought about by one or several specific hormones. Accordingly there would exist molting, pupation, and metamorphosis hormones, (b) There are two types of hormones interacting during development. The one type activates the imaginal potencies in a measure regulated by the responsiveness of the developing tissue and thus brings about periodic growth and differentiation. Factors of this type are called in this paper "growth and dif- ferentiation hormones." The other type, juvenile or inhibitory hormone, activates the "juvenile," i.e. larval potencies of the cells, and in this way prevents the onset of metamorphosis. According to this concept, first formulated by Wigglesworth 34. 1936, 1940), the presence of both factors in adequate proportion causes larval (nymphal) molting, whereas in the absence of the juvenile factor metamorpho- sis takes place. There is strong evidence that this "dualistic" mode of regulation exists nol only in hemimetabolous (Hemiptera, Wigglesworth, 1934, 1940; Orthop- tera, Pflugfclder, 1(^7, 1940; Pfeiffer, 1942; Scharrer, 1944), but also in holo- NEUROSECRETION 245 ' metabolous forms (Lepidoptera, Bounhiol, 1939; Coleoptera, Radtke, quoted from Piepho, 1943). In holometabolous insects, then, the effect ascribed by certain authors to a "molt- ing hormone" would actually result from the combined action of two hormonal factors, a juvenile hormone and a growth and differentiation hormone. Pupation and metamorphosis would take place in the presence of one or more differentiation factors alone. As sources of the hormones controlling insect development three organs in the head region of insects are known at present: (1) the glandular corpora allata, (2) the corpora cardiaca, consisting of nervous and glandular elements, (3) the pars intercerebralis of the brain containing glandlike nerve cells. In contrast to the known action of the corpus allatum which is the source of the juvenile hormone (see p. 244), the role of the two remaining centers has been less well understood. In one group of insects the brain seems to produce a hor- mone (or hormones) which in another group is provided by the corpus cardiacum. These two sources do not need to be treated as two separate centers of glandular activity, different as they may seem at first sight. On the basis of their unusual morphological relationship it is proposed to consider them as components of one neuro-endocrine complex whose role in the developing insect is the regulation of growth and differentiation. As to the mechanism of this glandular complex there are two possibilities. Either both the brain and the corpus cardiacum cooperate in the elaboration of growth and differentiation factors, or in different animals the one or the other component has become the predominant hormone source. Considering the vari- ability in the development of neuroglandular organs in the insect head one may expect to find examples for either alternative among the various groups of insects. The first possibility has to be considered, if extirpation of one of the two glandu- lar centers leads to disturbances but not to a complete interruption of the endocrine mechanism. For instance, it is known that cardiacectomy in Melanoplus (Orthop- tera, Pfeiffer, 1939) delays but does not entirely prevent molting. There are data that indicate the second possibility, i.e. an autonomous action of either the brain or the corpus cardiacum. In nymphs of Rhodnius (Hemiptera, Wigglesworth, 1940) brain implants cause molting in the absence of the corpus cardiacum. In Drosophila and Calliphora the ring-gland (in all probability its cardiacum component) furnishes growth and differentiation hormones, whereas the brain alone has little or no effect. Most of these data in Diptera (Hadorn, 1937; Hadorn and Neel, 1938; Burtt, 1938; Becker and Plagge, 1939; Vogt, 1942a; Gloor, 1943 ; Bodenstein, 1943a, 1943b, 1944), as well as observations made in other groups of insects, do not preclude, although they do not prove a collaboration be- tween pars intercerebralis and corpus cardiacum in the production of growth and differentiation hormones.3 All these data and considerations concern the developing insect. In the adult the corpora allata control egg development (Wigglesworth, 1936; Pfeiffer, 1939; Thomsen, 1940; Vogt, 1940; Scharrer, 1943) and color change (Pflugfelder, 1939) ; 3 Further information will be necessary about the identity and mode of action of certain thoracic centers, mentioned on p. 244, before they fit into the present concept of the hormonal control of insect development (see also Richards, 1937). _>4() S< IIARRKR AND St IIARRKR the functional significance of the intercerebralis-cardiacum complex is still unknown. However, the production of physiologically active substances by the intercerebralis- ranliacum complex also in the imago is indicated by the fact that extracts from either component in Periplaneta yield chromatophorotropic responses in crustaceans ( Brown and Mcglitsch. 1940). In summary, the intercerebralis-cardiacum-allatum system furnishes two types of hormonal factors which by their interaction control the rate of insect development. The one type. i.e. growth and differentiation hormone (or hormones), originates in the neuroglandular intercerebralis-cardiacum complex, the other (juvenile, in- hibitory hormone )in the corpus allatum. This seems to be the most satisfactory interpretation of the numerous existing observations. As has been indicated (p. 245) it is not the only one possible. VERTEBRATES The pituitary gland consists of two components, the pars buccalis (anterior lobe), and the pars nervosa (posterior lobe). The latter receives its innervation from nuclei in the hypothalamus ; how many of the nerve fibers also end in the anterior lobe, is not definitely known, and it seems to vary in different groups of vertebrates. In the fishes, the nucleus preopticus and the nucleus lateralis tuberis send their fibers to the hypophysis. The nucleus preopticus alone innervates the gland in the am- phibians. In the reptiles the nucleus preopticus is divided into two nuclei, the nuclei supraopticus and paraventricularis (Meyer, 1935). The cells of these two nuclei send their axons to the pituitary gland in the reptiles and the mammals. The significance of these hypothalamic nuclei as nervous centers controlling pituitary activity has been studied carefully in some species such as the cat, particularly by Ranson and his collaborators (Ranson and Magoun, 1939). The same cells which through their axons innervate part or all of the pituitary gland have been shown in a number of vertebrates to exhibit characteristics of gland cells (Scharrer and Scharrer, 1940). This means that the cells pass through cycles of secretory activity during which they produce granules and colloid droplets. The question arises whether this secretory activity of the nerve cells is connected with the control of the pituitary gland or has no relation to this function. The latter would be difficult to understand. The nervous control of the neural lobe is known to be of great importance for the normal function of this organ. It seems inconceivable that the cells which innervate the gland could themselves change into gland cells to the extent observed in some animals if the secretory activity of these nerve cells would serve a purpose unrelated to the activity of the pituitary gland. Such an independent glandular function would probably interfere with the task of the cells of innervating the hypophysis. The alternative, that the secretory activity of the neurons is part of the mecha- nism through which the hypothalamic nuclei exert control over the pituitary gland, would appear to be more acceptable. Evidence to support this view may be seen in two kinds of observations: ( 1 ) The secreted material can be traced along the axons to the hypophysis. (2) There is a seasonal cycle in certain cases of neurosecretory cells which may have a significance with regard to seasonal cycles in hypophyseal activity. NEUROSECRETION 247 Granules and droplets, discharged by secreting nerve cells, have been traced along the axons in a number of species (Fig. 2). In the catfishes, Notunts flavus and Amciurus ncbulosus, the nucleus preopticus and the nucleus lateralis tuberis both consist of secreting nerve cells (Palay, 1943). Acidophil granules and drop- lets of the same kind as produced by these cells are found all along the fibers from the preoptic nucleus to the pituitary gland. This tract can actually be differentiated from other fiber connections by the granules which in Masson preparations stain red and mark the bundle as clearly as the black granules would in a successful Marchi preparation (Palay, unpublished). PRCOPTIC NUCLEUS COLLO/D HrPOPHYSIS FIGURE 2. Diagram of the hypothalamo-hypophyseal system of a vertebrate. The granules along the axons of the preoptic nucleus can be seen also in other fishes, such as Tinea (Scharrer, 1936), Fundulus, and Centropristes. Likewise in amphibians, for instance in the toad, the granules are attached to the axons in bead- like arrangement for a long distance from the cell of origin. In reptiles, particu- larly in snakes, the fiber tract between the supraoptic nucleus and the hypophysis may be filled with colloid droplets as in the catfish. Also in the human supraoptic nucleus cells have been seen with acidophil granules along the axon some distance from the cell body (Gaupp and Scharrer, 1935). In the vertebrates a considerable amount of experimental work has been done, but the physiological role of the material secreted by the nerve cells is still un- known. All that can be said at present is that if the acidophil material discharged by the nerve cells contains an active principle, it appears to be directed toward the hypophysis. 248 SCHAKRKR AND SCHARRER It should be mentioned here that a number of investigators have suggested a migration in the opposite direction, i.e. of hypophyseal colloid from the pituitary gland to the hypothalamus (Edinger, 1911 ; Gushing, 1925; Collin, 1928; Popjak, 1940). There is no doubt that this actually takes place. The hypophyseal colloid can be differentiated from the colloid of the nerve cells in that it stains slightly dif- ferent and appears in the shape of irregular masses instead of sharply defined gran- ules. This hypophyseal colloid cannot be traced very far, and it is questionable whether it reaches the hypothalamus ; but the possibility that colloid may be ex- changed in both directions must be acknowledged. The significance of such an ex- change is largelv obscure, but a close interrelation between the activity of the hypo- physis and that of neurosecretory cells in the hypothalamus is suggested. A seasonal cycle of the secretory activity of neurosecretory cells has been found so far only in one species of teleosts. The cells of the nucleus lateralis tuberis of the tench (Tinea ruh/aris], a close relative of the carp, show no secretory activity during the winter months. It is very conspicuous during the summer months with gradual increase in spring and decrease in fall (Scharrer, 1936). In catfishes (Amcinnis ncbulosus and Noturus flavus), collected during the past three years, no corresponding cycle was found (Scharrer and Palay, unpublished). The pituitary gland of fishes is also subject to seasonal changes (Bock, 1928; Matthews, 1939; Evans, 1940). Whether and in which way the cyclic hypophyseal phenomena are related to those taking place in the nucleus lateralis tuberis is not known at present. The data available require closer investigation. Consequently it is proposed to follow the suggestion of physiologists and patholo- gists who have been considering the hypothalamic nuclei of higher animals together with the neural lobe as an interdependent system. Such a hypothalamo-hypophyseal system could be assumed to have originated from a hypothetical situation in which from one neuroglandular area in the brain the secreting hypothalamic nuclei and the pars nervosa of the hypophysis have been derived. The exchange of colloid, what- ever its functional meaning may be, could be considered as a remnant of the original connection. Charlton (1932) has presented evidence that in phytogeny the nucleus preopticus of the fishes has migrated rostrally. The cauda of the nucleus preopticus in fact points toward the hypophysis, and the irregularly occurring nucleus lateralis tuberis is still in very close proximity to the pituitary gland. DISCUSSION A comparison of the hypothalamo-hypophyseal system of vertebrates with the intercerebralis-cardiacum-allaturn system of insects reveals a parallelism which is the more striking because insect and vertebrate organs differ so greatly that no true organ homology can exist between these phyla. The hypothalamic nuclei of the vertebrates have their equivalent in the pars inter- cerebralis of the insects. In both centers neurosecretory cells are found, and both ^ M-nd nerve fibers to innervate complex endocrine organs, i.e. the pituitary gland and the corpora cardiaca and allata. In both the vertebrates and the invertebrates these nerve fibers contain colloid which can be traced from the nerve cells all the way to the glands innervated by them. The endocrine glands too are comparable as Hanstrom (1941) has pointed out. The hypophysis produces a number of well-known hormones influencing growth, NEUROSECRETION 249 gonadal development, chromatophores, etc. The corpora cardiaca and allata con- trol processes of equivalent importance in the life of insects, such as growth and metamorphosis, egg development, color change, etc. Evidently the corpora cardiaca and allata play a role in insects similar to that of the pituitary gland in vertebrates. Both glands are composite structures. The pituitary consists of a neural portion which forms the posterior lobe, and a glandular portion which forms the anterior lobe. These two components become associated to a varying extent ; they are most closely connected in the teleosts where the pars nervosa penetrates the pars glanclu- laris. In the insects the corpora cardiaca are comparable to the neural portion, the corpora allata to the glandular portion of the hypophysis. The corpora cardiaca and allata become associated in most insects ; in the muscoid Diptera they form an organ (ring -gland) in which the two components can be differentiated only histologically (Scharrer and Hadorn, 1938; Vogt, 1942b; Day, 1943; Poulson, 1944). The parallelism in the organization of the two systems here compared could be merely a coincidence. However, it seems more likely that the comparison is signifi- cant in that it indicates a fundamentally similar relationship between the "master glands" and the central nervous system in invertebrates and vertebrates. SUMMARY The hypothalamo-hypophyseal system in vertebrates is in many respects similar to the intercerebralis-cardiacum-allatum system in insects. (1) In vertebrates the hypothalamic nuclei innervating the pars nervosa of the pituitary gland contain secreting nerve cells. In a number of species colloid droplets can be traced along the axons from the neurosecretory cells of the hypothalamus to the hypophysis. (2) In insects the pars intercerebralis of the protocerebrum contains neuro- secretory cells. A bundle (nervus corporis cardiaci) innervating the corpus cardi- acum and probably also the corpus allatum originates in the pars intercerebralis. In Leucophaea (Orthoptera) as in the vertebrates, colloid can be traced from the secret- ing nerve cells of the pars intercerebralis to the corpora cardiaca all along the nervus corporis cardiaci. (3) On the basis of these morphological relationships the hypothalamic nuclei (nucleus preopticus and its homologues) and the pars nervosa of the hypophysis appear as one closely interconnected system. Likewise the pars intercerebralis and the corpus cardiacum of insects may be viewed as one neuro-endocrine complex rather than as two separate sources of hormones. In this way certain seemingly inconsistent data concerning the endocrine control of development in insects can be better understood (see p. 244). LITERATURE CITED BECKER, E., AND E. PLAGGE, 1939. Ueber das die Pupariumbildung auslosende Hormon der Fliegen. Biol. Zbl, 59: 326-341. BOCK, F., 1928. Die Hypophyse des Stichlings (Gasterosteus aculeatus L.) unter besonderer Berikksichtigung der jahrescyklischen Veranderungen. Z. zeiss. Zool., 131 : 645-710. BODEN STEIN, D., 1938. Untersuchungen zum Metamorphoseproblem. II. Entwicklungsrela- tionen in verschmolzenen Puppenteilen. Arch. Entw. Mcch., 137 : 636-660. BODENSTEIN, D., 1943a. Factors influencing growth and metamorphosis of the salivary gland in Drosophila. Biol. Bull., 84: 13-33. 250 SCHARRER AND SCHARRKR BODENSTEIN, D., 1943h. Hormones and tissue competence in the development of Drosophila. B'wl Hull.. 84: 34-58. BODEXSTEIX, D., 1944. The induction of larval molts in Drosophila. Biol. Bull., 86: 113-124. BOVXHIOL, J. J., 1938. Recherches experimentales sur le determinisme de la metamorphose chez les lepidopteres. Bull. Biol. France ct Bel;/., suppl. 24 : 1-199. BOUXHIOL, J. J., 1939. Recentcs recherches experimentales sur les insectes : les fonctions des corps allates (corpora allata). Arch. zool. c.i'pcr. gen.. 81: 54—64. BROWN, F. A., Jr., AXD A. MEGLITSCII, 1940. Comparison of the chromatophorotropic activity of insect corpora cardiaca with that of crustacean sinus glands. Biol. Bull.. 79: 409-418. BURTT, E. T., 1938. On the corpora allata of dipterous insects. II. Proc. Rov. Soc. London, Ser. B, 126: 210-223. CASPARI, E., AND E. PLAGGE, 1935. Versuche zur Physiologic der Verpuppung von Schmet- terlingsraupen. Naturwiss., 23: 751-752. I'M. \RI.TON, H. H., 1932. Comparative studies on the nucleus preopticus pars magnocellularis and the nucleus lateralis tuberis in fishes. /. coinp. Ncur., 54: 237-275. COLLIN, R., 1928. La neurocrinie hypophysaire. Etude histophysiologique du complexe tubero- infundibulo-pituitairc. Arch. Morph. (/en. c.rp., Paris, 28: 1-102. GUSHING, H., 1925. The pituitary gland as now known. Lancet, 209: 899-906. DAY, M. F., 1940a. Neurosecretory cells in the ganglia of Lepidoptera. Nature, 145 : 264. DAY, M. F., 1940b. Possible sources of internal secretions in the heads of some holometabolous insects. Anat. Rcc., 78, suppl. : 150. DAY, M. F., 1943. The homologies of the ring gland of Diptera Brachycera. A tin. Ent. Soc. Am, 36: 1-10. DE LERMA, B., 1937. Osservazioni sul sistema endocrine degli Insetti (corpora allata e corpi faringei). Arch. zool. ital., 24: 339-368. EDIXGER, L., 1911. Die Ausfuhrwege der Hypophyse. Arch, inikr. Anat., 78: 496-505. EVANS, H. M., 1940. On some seasonal changes in the pituitary gland of the eel. Brit. Mcd. Journ., serial no. 4135: 565-567. FUKUDA, S., 1940. Induction of pupation in silkworm by transplanting the prothoracic gland. Proc. Imp. Acad. Tokyo, 16 : 414-416. GAUPP, R., Jr., AND E. SCHARRER, 1935. Die Zwischenhirnsekretion bei Mensch und Tier. Z. flcs. Nenr. Psychiat., 153: 327-355. di.noR, H., 1943. Entwicklungsphysiologische Untersuchung an den Gonaden einer Letalrasse (Igl) von Drosophila melanogaster. Rev. Suissc Zool., 50: 339-394. HACHLOW, V., 1932. Zur Entwicklungsmechanik der Schmetterlinge. Arch. Entiv. Mcch., 125: 26-49. HADORN, E., 1937. An accelerating effect of normal "ring-glands" on puparium-formation in lethal larvae of Drosophila melanogaster. Proc. Nat. Acad. Sci., 23: 478-484. H ADORN, E., AND J. NEE!,, 1938. Der hormonale Einfluss der Ringdriise (corpus allatum) auf die Pupariumbildung bei Fliegen. Arch. Entiv. Mcch., 138: 281-304. HANSTROM, B., 1938. Untersuchungen aus dem Oeresund XXVI. Zwei Probleme betreffs der hormonalen Lokalisation im Insektenkopf. Lands Univ. Arsskr., N. F., Avd. 2, 34 : no. 16, 1-17. HANSTROM, B., 1940. Inkretorische Organe, Sinnesorgane und Nervensystem des Kopfes einiger niederer Insektenordnungen. Kgl. Svctisk. J'etcnsk. Hand!., 18: no. 8, 1-265. HANSTROM, B., 1941. Einige Parallelen im Bau und in der Herkunft der inkretorischen Organe der Arthropoden und der Vertebraten. Lunds Univ. Arsskr., N. F., Avd. 2, 37: no. 4, 1-19. Kni'Kr, S., 1922. Studies on the necessity of the brain for the inception of insect metamorphosis. Bio!. Bull., 42 : 323-342. KCIIX, A., AND H. I'lKi-HO, 1936. Uebcr hormonale Wirkungen bei der Verpuppung der Schmetterlinge. Ccs. ll'iss. (iottitujcti, Nachr. a.d. Biol., 2: 141-154. MATTHEWS, S. A., 1939. The relationship between the pituitary gland and the gonads in Fundulus. Biol. Bull., 76: 241-250. MEYER, W. C, 1935. Phylogenetische Ableitung des Nucleus supraopticus vom Nucleus para- vcntricularis. Dtsch. Z. Ncrvcnhcilk., 138: 65-74. NESBITT, H. H. J., 1941. A comparative morphological study of the nervous system of the Orthoptera and related orders. Ann. Enl. Soc. Am., 34: 51-81. NEUROSECRETION 251 OCHSE, W., 1944. Experimentelle und histologische Beitrage zur innern Metamorphose von Sialis lutaria L. Rev. Suisse Zool., 51 : 1-82. PALAY, S. L., 1943. Neurosecretion. V. The origin of neurosecretory granules from the nuclei of nerve cells in fishes. /. comp. Neur., 79 : 247-275. PFEIFFER, I. W., 1939. Experimental study of the function of the corpora allata in the grass- hopper, Melanoplus differentialis. /. cxpcr. Zool., 82: 439-461. PFEIFFER, I. W., 1942. Suppression of metamorphosis in the grasshopper, Melanoplus differ- entialis. Anat. Rec., 84: 486. PFLUGFELDEK, O., 1937. Bau, Entwicklung und Funktion der Corpora allata und cardiaca von Dixippus morosus Br. Z. ^ik K> ever, to the central ends ot" the branches of the melanophores. The outflow of pig- ment granules continues to reach the distal ends of all branches, hut the inflow re- verses itself before the granules have fully reached the central protoplasmic mass. Thus at the maximum of each concentration the melanophore displays a number of short stumplike excrescences which correspond to the number of branches which are fully developed at maximum expansion. At pressures between 3000 and 6000 Ibs. in.2, the amplitude of the pulsations is reduced more and more, entirely at the central end. The outflow still reaches the tips of the branches, but the inflow, at the higher pressures, is so restricted that almost the full length of each branch remains when the concentration of pigment is complete. Finally, at pressures between 7000 and 8000 lbs./in.-, the pulsations cease altogether. Xow all the melanophores remain in their completely expanded form. If the pressure is suddenly released, an immediate concentration phase alwavs sets in. This "release contraction" is very rapid, and it endures somewhat longer than the "contraction" which occurs in the ordinary rhythm of an uncompressed specimen. Effects on contracted - melanophores Since pressure inhibits and finally abolishes the contraction phase of the pulsa- tion cycle, it was of interest to determine how the melanophores might react in the presence of reagents which induce the pigment to remain in the concentrated state. Most of these experiments were done with scales immersed in X 10 KG solutions, although the same results were also obtained with adrenalin (1 : 1000). The me- lanophores of uncompressed control scales in these solutions remained in a fully contracted state for a period far in excess of the time required for the experiments. The results of these experiments are indicated in figure 1. The several photo- graphs are of the same melanophore successively exposed to different degrees of pressure. The response to increased pressure is always an immediate expansion, and the degree of this expansion bears a direct relation to the intensity of the pres- sure up to 7000-8000 Ibs. in.-. At this level a maximum dispersal of the pigment is always observed. The degree of expansion depends directly upon the intensity of the pressure. The same configuration is reached at a certain pressure regardless of whether the melanophore expands to this point, as a result of increasing the pressure, or con- tracts as a result of decompressing. Also when a fixed pressure is maintained, the characteristic degree of expansion is maintained for manv minutes. But whenever the pressure is completely released, the melanophores return immediately to a fully contracted state. Experiments witii denervated melanophores Since Parker ( 1934) has shown that the terminal branches of the nerves which supply the melanophores retain some influence on the contractile state even after 2 The term cUGL \S A. MARS! \ND all connections with tin- central nervous system have heen severed, it was necessary to detenniiie whether or not the action of pressure is mediated through the activity of the surviving nerve remnants. Therefore the foregoing experiments were re- peated, using melanophorcs in which complete denervation was assured. Kor this ]>urpose the method of Parker was used. Rroad dark bauds were estab- lished iu the caudal fins of ten medium sized Fundulus, by making dorso-ventral cuts completely through the fms about two mm. caudal to the origin. After ten days the hands were not quite as wide' as originally, but they were still clearly discernible as darker areas in light-adapted tish and as lighter areas in dark-adapted specimens. In accordance with Marker's conclusion, these observations indicate that the degen- eration of the residual nerves is complete within ten days, and that longer periods must elapse before secondary i'nnervation can occur. »5 /,»« ^ * m 4000 60OO 7000 ORE; SCALE OF FUNCUIUS iso In, i RE 1. Progressive inhibition of contraction hy hydrostatic pressure (courtesy of Iowa State l/nllr^e l're>-.. Ames. lo\\a). On the eleventh dav a small square section ol the tail-lm, so selected that about half the area was derived from the denervated region, was excised, washed in three changes of N 10 KC1, and placed in the pressure chamber, which also contained the same solution. All the melanophores of the excised piece were fully contracted and there was no discernible difference between the melanophores of the innervated and denervated areas. This proves that the contracting action of the KO solution does not depend upon the survival of the terminal nerves. l)itterent pressures, ranging up to 8000 lbs./in., gave exactly the same results as were described previous! v. Roth groups of melanophores, regardless of the pres- or absence of surviving nerves, gave the same' degree of expansion with each increment of pressure, and in both groups a full expansion was readied at 7000- Tliis proves that the pressure acts directly upon the melanophores per se, r than iudirectlv, via surviving nerve elements. Centrijuging experiments going experiments indicate that pressure exerts an inhibiting effect upon >omc protoplasmic reaction which determines the contractile state of the melano- PIGMENT FLOW IN l'HU< )M ATOPHOK TS phore. Apparently the equilibrium of this reaction can he shifted by each i or decrease of pressure in the range up to 8000 Ibs. in.2. In this respect the activity of melanophores closely resembles amoeboid movement, cleavage and cycl<»si>. And since these physiological activities are known to be determined by sol -gel changes occurring in the protoplasm, centriiuging experiments were undertaken, on the hypothesis that similar factors are involved in the present case. It soon became apparent that melanophores are very unfavorable for measuring gelational changes in the protoplasm. Very high centrifugal forces must be em- ployed before any sign of pigment displacement can be obtained and. due to the highly irregular form of the melanophore, quantitative measurements ot the degree of displacement are quite impossible. However, certain qualitative indications were obtained when the isolated scales were subjected to a centrifugal torce of 70.000 gravity in an air turbine ultracentrifuge.3 In the centrifuging experiments it was necessary to find a method lor holding the centrifuged scale in a position such that the anterior-posterior axis of the scale (and consequently the plane occupied by each melanophore and its processes) was parallel to the centrifugal axis. This was accomplished by rolling the scale into the form of a cylinder and inserting it into a short length of pyrex capillary tubing. sealed at the centrifugal end. The diameter of the lumen of the tubing was about half the width of the scale. Consequently when the elastic scale begins to unroll. its outer surface becomes firmly pressed against the inner surface of the tubing. The external diameter of the pyrex tubule was only slightly less than the internal diameter of the metal jacket in the head of the ultracentrifuge. and consequently the tubule axis and the axis of the centrifuge were approximately identical. Both the metal jacket and the pyrex tubule were filled with the immersion solution, and this arrangement tended to reduce the force of the impingement of the glass tubule upon the bottom of the metal jacket. Assuming that the resistance to the displacement of the melanin granules through the protoplasm of the melanophore provides an index of gelation, the experiments support the view that the protoplasm is set more firmly in the contracted than in the expanded melanophore. Using a standard force of 70,000 gravity and a fixed period of three minutes, no displacement of pigment was ever observed for con- tracted melanophores.* But an easily discernible pigment displacement (see figure 2) was always obtained under the same conditions with expanded specimens. Moreover, essentially the same results were obtained regardless of the agency u.sed to induce contraction (KC1 and adrenalin solutions), or expansion ( XaCl, acetyl choline,5 and physostigmine G solutions). Watching the melanin granules redistribute themselves after they have been dis- placed centrifugally. also provides an indication that the protoplasm is in a relative state of sol when the melanophores are expanded. A complete redistribution of the granules, after they are displaced as in figure 2, appears to depend upon I'.rown- ian movement. At any rate an exceedingly active Brownian movement can clearly 3 Cordial thanks arc extended to Dr. E. Newton Harvey and to Dr. Kthel Hn>\vn Harvey for their kindness in permitting the use of this equipment at the Marine Biological Laboratory. 4 In contracted melanophores no sign of pigment displacement was obtained with cvm greater forces (up to 125,000 gravity) employed for periods up to 12 mitiv n Acetyl choline chloride, Merck, 1X10"4, in N/10 NaC'l. c Physostigmine sulfate. Merck, 3 X W^, in N/10 Nad. 256 IK >UG1 \S A. M \KS1.A.\D be seen as tin- melanin granules begin to invade the part of tin- protoplasm which previously was cleared of the pigment. This movement continues for about an hour, at which time the redistribution is almost complete. Fna'KE 2. (./ ) contracted, and ( /•> ) expanded melanophoro, both centrifuged for 6 minutes at a force of 70,0(1(1 gravity. \o pi.unient displacement was ever obtained in contracted spcci- ii-iiiL' forces up to 125,000 gravity. (( ') snin-cxpanded specimen, shouin.u tin- renlral hyaline ])lasnia-iil re»ioii. I Jt-spite the drastic centrifugal treatment, the melanophores do not ajijiear to he' After the melanin granules have been redistributed and the normal 1 form has been regained, the melanophores are susceptible to further con- traction.-, and expansions, if the scales are successively immersed in solutions <>t KC1 and NaCl. PIGMENT FLOW IN CHROMATOPHORES 257 Temperature experiments Since it is known that the magnitude and even the sign of the pressure effect upon certain physiological processes (bioluminescence. Brown, Johnson and Marslancl. 1942; and muscular contraction. Brown, 1934) depend upon temperature, some of the experiments were repeated at low (6° C.) and high (30° C.) temperatures. These latter experiments demonstrate that high temperature fosters a contrac- tion of the melanophores ; whereas low temperature tends to induce expansion. At room temperature (20-22° C.), melanophores immersed in N/10 NaCl are all ex- panded ; but at 30° C. they all reach a full state of contraction. Moreover, at room temperature melanophores in N/10 KC1 are fully constricted; but at 6° C., a ma- jority on a given scale are about two-thirds expanded, and the others are fully ex- panded. Also at 30°, when fully contracted KCl-immersed specimens are subjected to hydrostatic compression, it is very difficult to initiate expansion. In fact, the first sign of this effect does not appear until the pressure reaches about 6000/in.2, in contrast to the room temperature level, namely 1000-2000 lbs./in.2. At 30° C., however, the viability of the melanophores is limited to about two hrs., at least in solutions containing only NaCl or KC1. After this time irritability is lost rapidly, and no further contractions or expansions can be elicited. DISCUSSION The action of pressure is localized in the melanophore itself, rather than in the surviving nerve supply, as is clearly established by the work on denervatecl speci- mens. Thus it seems likely that pressure has its main effect upon some intrinsic component in the protoplasmic system. Recent studies on the biological effects of pressure indicate that pressure exerts at least two main types of action in protoplasmic systems. The first type of action appears to be primarily chemical, in that the pressure modifies the velocity or equi- librium of one or more of the metabolic reactions which energize the physiological process ; but the second effect is physical, in that pressure appears to change the viscous and tensile properties of specific gel structures in the cell. No doubt both types of pressure effects are present in any given system, but the first kind of action seems to dominate in the studies on muscle, nerve, and lumines- cence; whereas the second is dominant in amoeboid movement, cleavage, and cyclosis (Marsland, 1942). The present evidence indicates that the action of pressure upon melanophores is mainly of the second type, and that sol-gel changes are definitely concerned with the development of the forces which cause the granules to flow back and forth in the protoplasmic branches of the pigment cells. However, quantitative measurements to substantiate this view could not be obtained, and consequently the qualitative evidence must be considered very carefully. Pressure determines the dispersion and concentration of the melanophore pig- ment in a very regular and decisive fashion, and this action is clearly parallel to the pressure effects upon amoeboid movement, cleavage, and cyclosis. This leads to the conclusion that sol-gel changes are likewise of critical importance in melanophore activity. According to this view, contraction of the melanophore is determined by a gelation of the protoplasm; whereas expansion depends upon solation. 258 DOUGLAS A. MAINLAND This hypothesis permits a logical interpretation of the observed effects of both temperature and pressure. The gel system of the melanophore clearly conforms to that of the several other protoplasmic gels which have been studied, in that it he- haves like a type III gel (Freundlich, 1937). This type of gel sets more finnlv K'itli inereasiin/ temperature and, in setting, undergoes a small but definite increase of roluuie. Type III gels invariably undergo solation when the pressure is in- creased, or when the temperature is lowered (Marsland and Brown, 1942). Conse- quently the facts : that both high pressure and low temperature regularly cause ex- pansion in the melanophore ; and conversely, that decreasing pressure and increas- ing temperature regularly bring about contraction, may be considered as strong evi- dence in favor of the hypothesis. IOC-' 90 • - Arbitrary point, all other values are relative to this. ' = Gel va ue. Amoeba. j. - Gel va ue unfcrt Arbacia *gg 80 X - Gel va ue 4- =. Rate 0 cl cleaving. Arbacn egg \» — Gel va ue Elodea cells. 70 0 = Rate of streaming. Elodea cells M = Gel value pH — 65 myosm gel (rabbit) temp 23~-24 C 60 B A 50 T \ 40 \i '\ JO 20 10 0 \ |\ \ «* N^, j," ^*-> ^^X_ » 1 2345678 9 10 II PRESSURE -LBS./ 1 N2x I03 KHJURE 3. .Inhibition of protoplasmic movements in relation to the solating effects of hydrostatic pressure. A substantiation of this view must await more data, but the existing evidence ins very significant. The contraction phase of the pigmentary response is limited by pressure in a manner that parallels, at least roughly, the inhibition of gelation diich has been demonstrated in plasmagel systems generally. The "half-expanded"' state of the chromatophorc (Fig. 1 ) occurs at about 2500 lbs./in.2 (at 20°-22° C), v'hich corresponds to a gelation value of approximately 50 per cent (Fig. 3) ; and it seems probable that the other values, should thev become available, will likewise lall upon the general curve. Also the centrifnging experiments provide at least a qualitative demonstration ih.il -el.ilioii and .-.olation are of critical significance in melanophore activity. Regardless of the agent ( XaCl, physostiginine and acetyl- PIGMENT FLOW IX CHROMATOPHORES 259 choline solutions) which was used to bring about the pigment dispersal, expanded melanophores always showed a clear displacement of the melanin granules at a centrifugal force of 70,000 Xg. ; whereas this force never gave any sign of pigment displacement in contracted melanophores. It seems worthwhile to speculate briefly as to how gelation may instrument the movement of the pigment from branches into the body of the chromatophore. The simplest concept of the mechanism comes from the views of Lewis, 1942, and Mars- land. 1942, which were derived from studies on amoeboid movement and cleavage. According to this view every gel, by virtue of its intrinsic structure, tends to con- tract spontaneously. The gel is conceived of as a colloidal network of interconnected protein units, the interstices of which are filled with a fluid residuum (if the original sol. Conditions wrhich foster gelation lead to a strengthening of the interconnecting bonds and to a folding of the extended protein units which constitute the framework of the gel. These factors account for the contractile tendency, which is accompanied by an exudation of sol, expressed from the shrinking interstitial spaces. In the melanophore, the pigment granules are probably enmeshed in, or at- tached to, the gel framework which, in the expanded specimen, extends out into the outlying branches. During contraction, conditions favoring a firmer setting of the gel are imposed upon the system and consequently the gel framework begins to shrink and pull the pigment granules inward from the extended branches. Simul- taneously, however, these must be an outflow of sol into the branches, in sufficient quantity to compensate for the volume of the retreating gel and pigment. Direct observations of the melanophores under oil immersion tends to support the foregoing hypothesis. In the partially and fully expanded specimen the central region of the body of the melanophore is clearly hyaline, and the pigment granules are located only in the peripheral parts of the body and in the extended branches (Fig. 2). This differentiation indicates that the plasmasol and plasmagel of the melanophore occupy the same relative positions in amoeboid cells generally. Conse- quently it is proposed that the hyaline protoplasm of the body of the melanophore be considered as the plasmasol, which is surrounded by the pigmented outlying proto- plasm, the plasmagel. During contraction the central hyaline plasmasol region of the melanophore be- comes obliterated. No doubt this results from the encroachment of pigment gran- ules which come in from the branches. But in the process of this obliteration, prob- ably, the framework of the plasmagel must shrink, exerting a pressure upon the enclosed sol. Thus part at least of the sol must seep out through the meshes of the surrounding gel and escape into the branches to replace the material which is retreating from these parts. Such an exudation of hyaline sol through the mesh- work of the surrounding gel would be homologous to the escape of sol which occurs through the plasmagel sheet at the advancing tip of a pseudopodium in Amoeba (Mast, 1926). The outlying branches of the melanophore persist during contraction (Matthews, 1931). But no one has been able to observe the outflow of hyaline sol, which must inevitably occur while the pigment is retreating from the branches. The difficulties of such observations are, no doubt, first that the sol is completely hyaline, and second, that the branches tend to be obscured by other tissues occupying a more superficial position on the scale (Matthews, 1931). 260 DOUGLAS A. MARSLAND Indirect evidence of the outflow of sol may be obtained by observing (oil immer- sion) the inflow of pigment granules from the peripheral tips of the branches of the melanophore during the early stages of contraction. These granules behave as if they were being dragged, so to speak, against the stream. They exhibit a peculiar bobbing movement which is distinctly different from Brownian movement. The individual granules tend to arrange themselves on a linear series and do not change their relative positions despite the irregularity of their movement. Later, when contraction nears completion and the pigment reaches the stouter trunk-like origin of each branch, the linear arrangement of the granules is even more accentu- ated, but the bobbing movements have practically subsided. These observations appear to reinforce two main points in the sol-gel hypothesis : first that the pigment granules are definitely affixed in the contracting gel framework and consequently tend to display a definite pattern of arrangement ; and second, that the contraction of the gel framework generates an outward flow of the hyaline sol, derived partly from the central fund of plasmasol and partly from the interstices of the gel itself. No evidence can be offered as to the mechanism of relaxation, which re- distributes the pigment granules after a contraction has abated, except that the protoplasm of the pigment cells always shows a definite degree of solation when relaxation occurs. Compared to the very firm gelation of the contracted state, this solation may be just as great as the solations which have been demonstrated in such relatively loose gels as the plasmagel of the amoeba. But the whole system of the melanophore is pitched at a higher level of gelation. This is indicated by the great centrifugal force necessary to displace the pigment even in the relaxed cells. In the amoeba, a force of less than 7000 gravity is adequate to displace all granules even when the plasmagel is set to its maximum firmness, but in the melanophore a ten times greater force (70,000) is needed, even when the gel system is at minimum "solidity." Apparently there is a very definite residuum of gel struc- ture in the melanophore protoplasm even under conditions of maximum solation. Consequently it is possible that relaxation results from an unfolding of this per- sistant gel mesh-work, by a reversal of the same processes which determine its contractile folding. In any event it is plain that redistribution of the pigment does not depend on Brownian movement. In drastically centrifuged melanophores, in which presumably the pigment granules have been torn loose from their connection with the gel structure, more than an hour elapses before the displaced granules reach the periphery of the cleared protoplasm in the body and branches of the pigment cells. SUMMARY (1) Increasing hydrostatic pressure progressively inhibits the concentration of melanophore pigment, at least roughly in proportion to the magnitude of the pressure, in the range up to 7000 pounds per square inch. At each higher pressure the capacity to contract is further reduced, not only in the case of pulsating melano- pl lores (Spaeth method), but also in the case of steady contractions induced by various chemical agents. (2) This action of pressure is entirely independent of the nerve supply of the melanophores, since denervation does not in any way alter the pressure responses of the pigment cells. PIGMENT FLOW IN CHROMATOPHORES 261 (3) Low temperature (6° C.) reinforces the pressure inhibition of contraction, but high temperature (30° C.) has a counteracting effect. (4) Both the pressure and the temperature effects indicate that contraction depends upon the capacity of the protoplasm of the pigment cells to undergo gelation ; whereas expansion involves solation. This hypothesis, which is borne out by a number of microscopic observations, brings melanophore activity into line with several other types of protoplasmic movement. LITERATURE CITED BROWN, D. E. S., 1934. The pressure-tension-temperature relation in cardiac muscle. Amcr. Jour. PhysioL, 109 : 16. BROWN, D. E. S., F. H. JOHNSON, AND D. A. MARSLAND, 1942. The pressure, temperature relations of bacterial luminescence. /. Cell, and Comp. PhysioL, 20: 151-168. BROWN, D. E. S., AND D. A. MARSLAND, 1936. The viscosity of Amoeba at high hydrostatic pressure. Jour. Cell and Comp. PhysioL, 8 : 159-165. FREUNDLICH, H., 1937. Some recent work on gels. Jour. Phys. Cliem., 41 : 901-915. HOOKER, D., 1914. Amoeboid movement in the corial melanophores of frogs. Anat. Rec., 8: 103. LEWIS, W. H., 1942. The relation of the viscosity changes of protoplasm to amoeboid loco- motion and cell division. In The structure of protoplasm, W. Seifriz (Ed.), Iowa State College Press, Ames, Iowa. MARSLAND, D. A., 1938. The effects of high hydrostatic pressure upon cell division in Arbacia eggs. Jour. Cell, and Comp. PhysioL, 12 : 57-70. MARSLAND, D. A., 1939a. The mechanism of cell division. Hydrostatic pressure effects upon dividing egg cells. Jour. Cell, and Comp. PhysioL, 13 : 15-22. MARSLAND, D. A., 1939b. The mechanism of protoplasmic streaming. The effects of high hydrostatic pressure upon cyclosis in Elodea canadensis. Jour. Cell, and Comp. PhysioL, 13 : 23-30. MARSLAND, D. A., 1942. Protoplasmic streaming in relation to gel structure in the cytoplasm. In The structure of protoplasm, William Seifriz (Ed.), Iowa State College Press, Ames, Iowa. MARSLAND, D. A., AND D. E. S. BROWN, 1936. Amoeboid movement at high hydrostatic pressure. /. Cell, and Comp. PhysioL, 8 : 167-178. MARSLAND, D. A., AND D. E. S. BROWN, 1942. The effects of pressure on sol-gel equilibria, with special reference to myosin and other protoplasmic gels. /. Cell, and Comp. PhysioL, 20 : 295-305. MAST, S. O., 1926. Structure, movement, locomotion, and stimulation in Amoeba. Journ. Morph. and PhysioL, 41 : 347-425. MATTHEWS, S. A., 1931. Observations on pigment migration within the fish melanophore. Jour. E.vp. Zool., 58 : 471-485. PARKER, G. H., 1934. The prolonged activity of momentarily stimulated nerves. Pro. Nat. Acad. Sci. Washington, 20: 306-310. SHANES, A. M., AND R. F. NIGRELLI, 1941. The Chromatophores of Fundulus heteroclitus in polarized light. Zoohgica, 26: 237-245. SPAETH, R. A., 1916. Evidence proving the melanophore to be a disguised type of smooth muscle cell. Jour. E.rp. Zool., 20 : 193-215. INDEX A BELL, RICHARD G., AND WILLIAM M. PARKINS. Gelatin as a plasma substitute, with special reference to pseudo-agglutina- tion, 162. \listracts of scientific papers presented at the Marine Biological Laboratory, Summer of ' 1944, 153. Albumen, of developing avian egg, 223. . \Ilium, cytology of, 163 AMBERSON, WILLIAM R., JOVE J.JENNINGS, AND C. MARTIN RHODE. Recent experience with hemoglobin-saline solution, 161. Annual report of the Marine Biological Labora- tory, 1. Arsenite, effect on respiration in Pelomyxa, 138. Astropecten marginatus, ecological observa- . tions on, 177. gEHAVIOR and tube building habits of Polydora ligni, 164. BELDA, W. H. See D. M. PACE, 138. BERGER, C. A. Experimental studies on the cytology of Allium, 163. BKRTHOLF, L. M., AND S. O. MAST. Meta- morphosis in the larva of the Tunicate, Styela partita, 166. Biochemical factors in the maximal growth of Tetrahymena, 107. Biology of the California sea-mussel (Mytilus californianus) III. Environmental condi- tions and rate of growth, 59. Blue crab, Callinectes sapidus Rathbun, mor- phology of zoeal stages of, 145. BTCK, J. B. The click mechanism of elaterid beetles, 165. BU.I.OCK, THEODORE H. Oscillographic stud- ies on the giant nerve fiber system in Lumbricus, 159. I'.i MI-, GARDINER. See LEONARD B. CLARK, 134. Q ALIFORM A sea-mussel (Mytilus cali- fornianus) biology of, 5(>. Callinectes sapidus Rathbun, morphology of zoeal stages of, 145. Capillary bed of the central nervous system of certain invertebrates, 52. Chemical organization of the cytoplasm, 156. Chromatophores, mechanism of pigment dis- plarcuiciit in, 252. Chromosome cycle, primitive, in coccid, 167. CLARK, LEONARD B., AND GARDINER BUMP. X-rays and the reproductive cycle in ring- necked pheasants, 134. Click mechanism in elaterid beetles, The, 165. Central nervous system of certain inverte- brates, capillary bed of, 52. Coccid chromosome cycle, primitive, 167. COE, WESLEY R., AND DENIS L. Fox. Biology of the California sea-mussel (Mytilus cali- fornianus) III. Environmental conditions and rate of growth, 59. Crystallography, native protein, and diffraction patterns, 157. Current, voltage, and resistance characteristics of injured nerves, 158. Cyanide, effect on respiration in Pelomyxa, 138. Cytoplasm, chemical organization of, 156. "TRAVIS, JAMES O. Photochemical spectral analysis of neural tube formation, 73. DEWEY, VIRGINIA C. Biochemical factors in the maximal growth of Tetrahymena, 107. DEWEY, VIRGINIA C. See KIDDER, GEORGE W., 121. Diffraction patterns, native protein crystal- lography and, 157. gCHINODERMS, Puerto-Rican, ecological observations on, 177. Ecological observations on two Puerto-Rican echinoderms, Mellita lata and Astropecten marginatus, 177. Effects of peripheral factors on motor neuron differentiation in the chick embryo, 153. Effects of potassium cyanide, potassium ar- senite, and ethyl urethane on respiration in Pelomyxa carolinensis, 138. Egg, avian, developing, hydrogen ion concen- tration of albumen and yolk of, 223. Elaterid beetles, click mechanism in 165. Embryonic growth in the viviparous poeciliid Heterandria formosa, 37. Energy source of the nerve action potential, 158. Evidence of perpetual proximo-distal growth of nerve fibers, 160. Experimental studies on the cytology of Allium, 163. 262 INDEX 263 External morphology of the third and fourth zoeal stages of the blue crab, Callinectes sapidus Rathbun, 145. pEATHER pigmentation, melanophore con- trol of, 153. Ferritin and iron metabolism, 155. Fever, chemical basis of, 163. Food vacuole in the Peritricha, with special reference to the hydrogen ion concentra- tion of its content and of the cytoplasm, 188. Fox, DENIS L. See WESLEY R. COE, 59. pALTSOFF, PAUL S. See EDITH MORTEN- SEN, 164. Gelatin as a plasma substitute, with special reference to psuedo-agglutination, 162. Giant nerve fiber system in Lumbricus, oscillo- graphic studies on, 159. Growth, embryonic, in the viviparous poeciliid, Heterandria formosa, 37, Growth, maximal, of Tetrahymena, biochemical factors in, 107. pJAMBURGER, VIKTOR. The effects of peripheral factors on motor neuron dif- ferentiation in the chick embryo, 153. HARRIS, DANIEL L. Phosphoprotein, 164. Hemoglobin-saline solutions, recent experience with, 161. Hemorrhage, peripheral circulatory changes during shock produced by, 161. Heterandria, formosa, viviparous poeciliid, em- bryonic growth in, 37. HOPKINS, SEWELL H. The external morphol- ogy of the third and fourth zoeal stages of the blue crab, Callinectes sapidus Rathbun, 145. HUGHES-SCHRADER, SALLY. A primitive Coc- cid chromosome cycle in Puto sp., 167. Hydrogen ion concentration of albumen and yolk of developing avian egg, 223. Hypothalamo-hypophyseal system of verte- brates, a comparison with the intercerebra- lis-cardiacum-allatum system of insects, 242. JNVERTEBRATES, capillary bed of central nervous system of certain, 52. Invertebrates, serological relationships between Mollusca and other, 96. Iron metabolism, ferritin and, 155. | ENNINGS, JOYE J. See AMBERSON, W. R., J 161. If ENK, ROMAN. Ecological observations on two Puerto-Rican echinoderms, Mellita lata and Astropecten marginatus, 177. KIDDER, GEORGE W., AND VIRGINIA C. DEWKY. Thiamine and tetrahymena, 121. T AVIN, GEORGE I. Recent developments in ultraviolet microscopy, 160. LAZAROW, ARNOLD. The chemical organiza- tion of the cytoplasm, 156. LEWIS, WARREN H. The superficial gel layer and its role in development, 154. Lumbricus, giant nerve fiber system in, 159. \/f ARINE Biological Laboratory, annual re- port of, 1. Marine Biological Laboratory, program and abstracts of scientific papers presented at, 153. MARSLAND, DOUGLAS. Mechanism of pigment displacement in unicellular chromato- phores, 252. Mass and time, significance of, 227. MAST, S. O., AND W. J. BOWEN. The food vacuole in the Peritricha, with special reference to the hydrogen ion concentra- tion of its content and of the cytoplasm, 188. Mechanism of pigment displacement in uni- cellular chroma tophores, 252. Melanophore control of the sexual dimorphism of feather pigmentation on the Barred Plymouth Rock fowl, 153. Mellita lata, ecological observations on, 177. MENKIN, VALY. Studies on the chemical basis of fever, 163. Metachromatic staining, theory of, 155. Metamorphosis in the larva of the Tunicate, Styela partita, 166. MICHAELIS, L. Theory of metachromatic staining, 155. Mollusca and other invertebrates, serological relationships between, 96. MOOG, FLORENCE. See S. SPIEGELMAN, 227. Morphology, external, of the zoeal stages of the blue crab, 145. MORTENSEN, EDITH, AND PAUL S. GALTSOFF. Behavior and tube building habits of Poly- dora ligni, 164. Motor neuron differentiation in the chick em- bryo, effects of peripheral factors in, 153. Mytilus californianus, biology of, 59. VTACHMANSOHN, DAVID. On the energy source of the nerve action potential, 158. Native protein crystallography and diffraction patterns, 157. Nerve fibers, perpetual proximo-distal growth of, 160. 264 INDEX Neural tube formation, photochemical spectral analysis ot, 73. Neurosecretion \'l. A comparison between the intercerebralis-cardiacum-allatum system of the insects and the hypothalamo-hypo- phy>cal system of the vertebrates, 242. (V\" the energy source of the nerve action potential, 158. On the interpretation of rates of regeneration in Tubularia and the significance of inde- pendence of mass and time, 27. Oscillographic studies on the giant nerve fiber system in Lumbricus, 159. pACE, D. M., AND \\. H. BELDA. The ef- fects of potassium cyanide, potassium ar- senite, ethyl urethane on respiration in Felomyxa carolinensis, 138. PARKINS, WILLIAM M. See ABELL, RICHARD G., 162. Pelomyxa carolinensis, effects of cyanide, ar- senite, and urethane on respiration in, 138. Peripheral circulatory changes during shock produced by hemorrhage, 161. Pheasants, ring-necked, X-rays and repro- ductive cycle in, 134. Phosphatase, phosphoprotein, 164. Phosphoprotein phosphatase, a new enzyme from the frog egg, 164. Photochemical spectral analysis of neural tube formation, 73. Pigment displacement in unicellular chromato- phores, mechanism of, 252. Plasma substitute, gelatin as a, 162. I'oeciliid, viviparous, Heterandria formosa, em- bryonic growth in, 37. Polydora ligni, behavior of, 164. Potential, nerve action, on the energy source of, 158. Primitive coccid chromosome cycle in Puto sp., 167. Program and abstracts of scientific papers pre- sented at the Marine Biological Labora- tory, 153. PuertO-Rican echinoderms, ecological observa- tions on, 1 77. Puto sp., primitive chromosome cycle in, 167. l> A'l'KS of regeneration in Tubularia, the in- terpretation of, and the significance of the independence of mass and time, 227. Recent experience with hemoglobin-saline solu- tions, 161. Regeneration, rate-, of, in Tubularia, 227. Relationships, serological, between the Mol- lusca and other invertebrates, 96. Reproductive cycle in ring-necked pheasant-, X-rays and, 134. Resistance characteristics of injured nerves, current, voltage, and, 158. Respiration in Pelomyxa, effects of cyanide, arsenite, and urethane on, 138. RHODE, C. MARTIN. See AMBERSON, \V. R., 161. ROMANOFF, ALEXIS L. Hydrogen ion concen- tration of albumen and yolk in the de- veloping avian egg, 223. GCHARRER, BERTA, AND ERNST SCHARRER. Neurosecretion VI. A comparison be- tween the intercerebralis-cardiacum-alla- tum system of the insects and the hypo- thalamo-hypophyseal system of the verte- brates, 242. SCHARRER, ERNST. The capillary bed of the central nervous system of certain inverte- brates, 52. SCHRADER, SALLY HUGHES-. A primitive coc- cid chromosome cycle in Puto sp., 167. SCHRIMSHAW, NEVIN S. Embryonic growth in the viviparous poecillid, Heterandria for- mosa, 37. Serological relationships between the Mollusca and other invertebrates, 96. SHANES, ABRAHAM M. Current, voltage, and resistance characteristics of injured nerves, 158. Shock produced by hemorrhage, peripheral cir- culatory changes during, 161. Spectral analysis, photochemical, of neural tube formation, 73. SriEGELMAN, S., AND FLORENCE MOOG. On the interpretation of rates of regeneration in Tubularia, and the signifidance of the inde- pendence of mass and time, 227. Studies on the chemical basis of fever, 163. Styela partita, metamorphosis in, 166. Superficial gel layer and its role in development, 154. "PETRAHYMENA, biochemical factors in maximal growth of, 107. Tetrahyrnena, Thiamine and, 121. Theory of metachromatic staining, 155. Thiamine and Tetrahymena, 121. Tubularia, rates of regeneration in, 227. T TLTRAVIOLET microscopy, recent develop- ments in, 160. Urethane, effect on respiration in Pelomyxa, 138. \7ERTEBRATES, a comparison between the intercerebralis-cardiacum-allatum system of insects and the hypothalamo-hypophy- system of, 242. INDEX 265 Voltage, and resistance characteristics of in- jured nerves, and current, 158. PAUL. Evidence of perpetual prox- imo-distal growth of nerve fibers, 160. \VILHELMI, RAYMOND W. Serological rela- tionships between the Mollusca and other invertebrates, 96. WILLIER, B. H. Melanophore control of the sexual dimorphism of feather pigmentation in the Barred Plymouth Rock fowl, 153. WRINCH, DOROTHY. Native protein crystal- lography and diffraction patterns, 157. "V-RAYS and the reproductive cycle in ring- necked pheasants, 134. VOLK, of developing avian egg, hydrogen ion concentration of, 223. ^ OEAL stages of the blue crab, external mor- phology of, 145. ZWEIFACH, BENJAMIN W. Peripheral circula- tory changes during shock produced by hemorrhage, 161. ERRATUM In the October 1944 issue in the article "The theory of metachromatic stain- ing" by L. Michaelis the formula of thionin on page 156 was incorrectly stated and should read as follows: NH- formula of thionin (univalent cation, as existing in neutral or slightly acid solution) Volume 87 Number 1 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University L. IRVING, Swarthmore College M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University A. C. REDFIELD, Harvard University F. SCHRADER, Columbia University DOUGLAS WHITAKER, Stanford University H. B. STEINBACH, Washington University Managing Editor AUGUST, 1944 Printed and Issued by LANCASTER PRESS, Inc. PRINCE &. LEMON STS. LANCASTER, PA. SERIAL LIST A SERIAL list of the holdings of The Marine Biological Labora- tory has been published as a separately bound supplement to The Biological Bulletin. This supplement lists with cross references the titles of journals in the Library; additional titles and changes are published annually. A few extra copies of the original list are still available. Orders may be directed to The Marine Biological Laboratory. MICROFILM SERVICE 1 HE Library of The Marine Biological Laboratory is now pre- pared to supply microfilms of material from periodicals included in its extensive list. Through the generosity of Dr. Athertone Seidell, the essential equipment has been set up and put into operation. The Staff of The Marine Biological Laboratory Library is anxious to extend the Microfilm Service, particularly at this time when dis- tance makes the Library somewhat inaccessible to many who nor- mally use it. Investigators who wish films should send to the Li- brarian the name of the author of the paper, its title, and the name of the periodical in which it is printed, together with the volume and year of publication. The rates are as follows: $.30 for papers up to 25 pages, and $.10 for each additional 10 pages or fraction thereof. It is hoped that many investigators will avail themselves of this service. Your Biological News You would not go to the library to read the daily newspaper — probably you have it delivered at your home to be read at your leisure. Why, then, depend upon your library for your biological news ? Biological Abstracts is news nowadays. Abridgments of all the im- portant biological literature are published promptly — in many cases before the original articles are available in this country. Only by having your own copy of Biological Abstracts to read regularly can you be sure that you are missing none of the literature of particular interest to you. An abstract of one article alone, which otherwise you would not have seen, might far more than compensate you for the subscription price. Biological Abstracts is now published in six low priced sections, as well as the complete edition, so that the biological literature may be avail- able to all individual biologists. Write for full information and ask for a copy of the section covering your field. BIOLOGICAL ABSTRACTS University of Pennsylvania Philadelphia, Pa. LANCASTER PRESS, Inc. LANCASTER, PA. THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best. We shall be happy to have workers at the MARINE BIOLOGICAL LABORATORY write for estimates on journals or monographs. Our prices are moderate. INSTRUCTIONS TO AUTHORS The Biological Bulletin accepts papers on a variety of subjects of biologi- cal interest. In general, a paper will appear within three months of the date of its acceptance. The Editorial Board requests that manuscripts conform to the requirements set below. Manuscripts. Manuscripts should be typed in double or triple spacing on one side of paper, S1/^ by 11 inches. Tables should be typewritten on separate sheets and placed in correct sequence in the text. Explanations of figures should be typed on a separate sheet and placed at the end of the text. Footnotes, numbered consecutively, may be placed on a separate sheet at the end of the paper. A condensed title or running page head of not more than thirty-five letters should be included. Manuscripts must be returned to the Editor with the galley proof. Page proofs will be sent only on request. Figures. The dimensions of the printed page, 5 by 7% inches, should be kept in mind in preparing figures for publication. Illustrations should be large enough so that all details will be clear after appropriate reduction. Explana- tory matter should be included in legends as far as possible, not lettered on the illustrations. Figures should be prepared for reproduction as line cuts or half- tones; other methods will be used only at the author's expense. Figures to be reproduced as line cuts should be drawn in black ink on white paper or blue- lined co-ordinate paper; those to be reproduced as halftones should be mounted on Bristol board and any designating letters or numbers should be made di- rectly on the figures. The author's name should appear on the reverse side of all figures. Literature cited. The list of literature cited should conform to the style set in this issue of The Biological Bulletin. Papers referred to in the manuscript should be listed on separate pages headed "Literature Cited." Where there are several papers cited, by the same author, the author's name should be repeated in each case. Mailing. Manuscripts should be packed flat, not folded or rolled. Large charts and graphs may be rolled in a mailing tube. Reprints. Authors will be furnished, free of charge, one hundred reprints without covers. Additional copies may be obtained at cost; approximate figures will be furnished upon request. THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Pennsylvania. Subscriptions and similar matter should be addressed to The Biologi- cal Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain: Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, $1.75. Subscription per volume (three issues), $4.50. Communications relative to manuscripts should be sent to the Manag- ing Editor, Marine Biological Laboratory, Woods Hole, Massachusetts, between July 1 and October 1, and to the Department of Zoology, Wash- ington University, St. Louis, Missouri, during the remainder of the year. Entered as second-class matter May 17, 1930, at the post office at Lancaster. Pa., under the Act of August 24, 1912. BIOLOGY MATERIALS The Supply Department of the Marine Biological Labora- tory has a complete stock of excellent plain preserved and injected materials, and would be pleased to quote prices on school needs. PRESERVED SPECIMENS for Zoology, Botany, Embryology, and Comparative Anatomy LIVING SPECIMENS for Zoology and Botany including Protozoan and Drosophila Cultures, and Animals for Experimental and Laboratory Use. MICROSCOPE SLIDES for Zoology, Botany, Embryology, Histology, Bacteriology, and Parasitology. CATALOGUES SENT ON REQUEST Supply Department MARINE BIOLOGICAL LABORATORY Woods Hole, Massachusetts CONTENTS Page ANNUAL REPORT OF THE MARINE BIOLOGICAL LABORATORY . . || i SCRIMSHAW, NEVIN S. Embryonic Growth in the Viviparous Poeciliid, Heterandria Formosa • • 37 SCHARRER, ERNST The Capillary Bed of the Central Nervous System of Certain Invertebrates . . 52 COE, WESLEY R., AND DENIS L. Fox Biology of the California Sea-Mussel (Mytelus Calif ornianus). III. Environmental Conditions and Rate of Growth 59 DAVIS, JAMES O. Photochemical Spectral Analysis of Neural Tube Formation . 73 WILHELMI, RAYMOND W. Serological Relationships between the Mollusca and Other Invertebrates . 96 Volume 87 Number 2 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board , LI ft V I 3 • E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University L. IRVING, Swarthmore College M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University A. C. REDFIELD, Harvard University F. SCHRADER, Columbia University DOUGLAS WHITAKER, Stanford University H. B. STEINBACH, Washington University Managing Editor OCTOBER, 1944 Printed and Issued by LANCASTER PRESS, Inc. PRINCE & LEMON STS. LANCASTER, PA. SERIAL LIST A SERIAL list of the holdings of The Marine Biological Labora- tory has been published as a separately bound supplement to The Biological Bulletin. This supplement lists with cross references the titles of journals in the Library; additional titles and changes are published annually. A few extra copies of the original list are still available. Orders may be directed to The Marine Biological Laboratory. MICROFILM SERVICE 1 HE Library of The Marine Biological Laboratory is now pre- pared to supply microfilms of material from periodicals included in its extensive list. Through the generosity of Dr. Athertone Seidell, the essential equipment has been set up and put into operation. The Staff of The Marine Biological Laboratory Library is anxious to extend the Microfilm Service, particularly at this time when dis- tance makes the Library somewhat inaccessible to many who nor- mally use it. Investigators who wish films should send to the Li- brarian the name of the author of the paper, its title, and the name of the periodical in which it is printed, together with the volume and year of publication. The rates are as follows: $.30 for papers up to 25 pages, and $.10 for each additional 10 pages or fraction thereof. It is hoped that many investigators will avail themselves of this service. Your Biological News You would not go to the library to read the daily newspaper — probably you have it delivered at your home to be read at your leisure. Why, then, depend upon your library for your biological news ? Biological Abstracts is news nowadays. Abridgments of all the im- portant biological literature are published promptly — in many cases before the original articles are available in this country. Only by having your own copy of Biological Abstracts to read regularly can you be sure that you are missing none of the literature of particular interest to you. An abstract of one article alone, which otherwise you would not have seen, might far more than compensate you for the subscription price. Biological Abstracts is now published in six low priced sections, as well as the complete edition, so that the biological literature may be avail- able to all individual biologists. Write for full information and ask for a copy of the section covering your field. BIOLOGICAL ABSTRACTS University of Pennsylvania Philadelphia, Pa. LANCASTER PRESS, Inc. LANCASTER, PA. THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best. We shall be happy to have workers at the MARINE BIOLOGICAL LABORATORY write for estimates on journals or monographs. Our prices are moderate. INSTRUCTIONS TO AUTHORS The Biological Bulletin accepts papers on a variety of subjects of biologi- cal interest. In general, a paper will appear within three months of the date of its acceptance. The Editorial Board requests that manuscripts conform to the requirements set below. Manuscripts. Manuscripts should be typed in double or triple spacing on one side of paper, 8^/2 by 11 inches. Tables should be typewritten on separate sheets and placed in correct sequence in the text. Explanations of figures should be typed on a separate sheet and placed at the end of the text. Footnotes, numbered consecutively, may be placed on a separate sheet at the end of the paper. A condensed title or running page head of not more than thirty-five letters should be included. Manuscripts must be returned to the Editor with the galley proof. Page proofs will be sent only on request. Figures. The dimensions of the printed page, 5 by 7% inches, should be kept in mind in preparing figures for publication. Illustrations should be large enough so that all details will be clear after appropriate reduction. Explana- tory matter should be included in legends as far as possible, not lettered on the illustrations. Figures should be prepared for reproduction as line cuts or half- tones; other methods will be used only at the author's expense. Figures to be reproduced as line cuts should be drawn in black ink on white paper or blue- lined co-ordinate paper; those to be reproduced as halftones should be mounted on Bristol board and any designating letters or numbers should be made di- rectly on the figures. The author's name should appear on the reverse side of all figures. Literature cited. The list of literature cited should conform to the style set in this issue of The Biological Bulletin. Papers referred to in the manuscript should be listed on separate pages headed "Literature Cited." Where there are several papers cited, by the same author, the author's name should be repeated in each case. Mailing. Manuscripts should be packed flat, not folded or rolled. Large charts and graphs may be rolled in a mailing tube. Reprints. Authors will be furnished, free of charge, one hundred reprints without covers. Additional copies may be obtained at cost; approximate figures will be furnished upon request. THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Pennsylvania. Subscriptions and similar matter should be addressed to The Biologi- cal Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain: Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, $1.75. Subscription per volume (three issues), $4.50. Communications relative to manuscripts should be sent to the Manag- ing Editor, Marine Biological Laboratory, Woods Hole, Massachusetts, between July 1 and October 1 , and to the Department of Zoology, Wash- ington University, St. Louis, Missouri, during the remainder of the year. Entered as second-class matter May 17, 1930, at the post office at Lancaster. Pa., under the Act of August 24, 1912. BIOLOGY MATERIALS The Supply Department of the Marine Biological Labora- tory has a complete stock of excellent plain preserved and injected materials, and would be pleased to quote prices on school needs. PRESERVED SPECIMENS for Zoology, Botany, Embryology, and Comparative Anatomy LIVING SPECIMENS 0 for Zoology and Botany including Protozoan and Drosophila Cultures, and Animals for Experimental and Laboratory Use. MICROSCOPE SLIDES for Zoology, Botany, Embryology, Histology, Bacteriology, and Parasitology. CATALOGUES SENT ON REQUEST Supply Department MARINE BIOLOGICAL LABORATORY Woods Hole, Massachusetts CONTENTS Page DEWEY, VIRGINIA C. Biochemical Factors in the Maximal Growth of Tetrahymena 107 KIDDER, GEORGE W.} AND VIRGINIA C. DEWEY Thiamine and Tetrahymena 121 CLARK, LEONARD B., AND GARDINER BUMP X-Rays and the Reproductive Cycle in Ring-Necked Pheasants 134 PACE, D. M., AND W. H. BELDA The Effects of Potassium Cyanide, Potassium Arsenite, and Ethyl Urethane on Respiration in Pelomyxa carolinensis .... 138 HOPKINS, SEWELL H. The External Morphology of the Third and Fourth Zoeal Stages of the Blue Crab, Callinectes sapidus Rathbun 145 PROGRAM AND ABSTRACTS OF SCIENTIFIC PAPERS PRESENTED AT THE MARINE BIOLOGICAL LABORATORY, SUMMER OF 1944. 153 Volume 87 Number 3 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University L. IRVING, Swarthmore College M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University A. C. REDFEELD, Harvard University F. SCHRADER, Columbia University DOUGLAS WHITAKER, Stanford University H. B. STEINBACH, Washington University Managing Editor DECEMBER, 1944 Printed and Issued by LANCASTER PRESS, Inc. PRINCE £ LEMON STS. LANCASTER, PA. SERIAL LIST A SERIAL list of the holdings of The Marine Biological Labora- tory has been published as a separately bound supplement to The Biological Bulletin. This supplement lists with cross references the titles of journals in the Library; additional titles and changes are published annually. A few extra copies of the original list are still available. Orders may be directed to The Marine Biological Laboratory. MICROFILM SERVICE 1 HE Library of The Marine Biological Laboratory is now pre- pared to supply microfilms of material from periodicals included in its extensive list. Through the generosity of Dr. Athertone Seidell, the essential equipment has been set up and put into operation. The Staff of The Marine Biological Laboratory Library is anxious to extend the Microfilm Service, particularly at this time when dis- tance makes the Library somewhat inaccessible to many who nor- mally use it. Investigators who wish films should send to the Li- brarian the name of the author of the paper, its title, and the name of the periodical in which it is printed, together with the volume and year of publication. The rates are as follows: $.30 for papers up to 25 pages, and $.10 for each additional 10 pages or fraction thereof. It is hoped that many investigators will avail themselves of this service. Your Biological News You would not go to the library to read the daily newspaper — probably you have it delivered at your home to be read at your leisure. Why, then, depend upon your library for your biological news ? Biological Abstracts is news nowadays. Abridgments of all the im- portant biological literature are published promptly — in many cases before the original articles are available in this country. Only by having your own copy of Biological Abstracts to read regularly can you be sure that you are missing none of the literature of particular interest to you. An abstract of one article alone, which otherwise you would not have seen, might far more than compensate you for the subscription price. Biological Abstracts is now published in seven low priced sections, as well as the complete edition, so that the biological literature may be avail- able to all individual biologists. Write for full information and ask for a copy of the section covering your field. BIOLOGICAL ABSTRACTS University of Pennsylvania Philadelphia, Pa. LANCASTER PRESS, Inc. LANCASTER, PA. THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best. We shall be happy to have workers at the MARINE BIOLOGICAL LABORATORY write for estimates on journals or monographs. Our prices are moderate. INSTRUCTIONS TO AUTHORS Tin- I'.iological bulletin accepts papers on a variety of subjects of biologi- cal interest. In general, a paper \vill appear within three months of the date of its acceptance. The Editorial Board requests that manuscripts conform to the requirements set below. Manuscripts. Manuscripts should be typed in double or triple spacing on one side of paper. S1^' by 11 inches. Tables should be typewritten on separate sheets and placed in correct sequence in the text. Explanations of figures should be typed on a separate sheet and placed at the end of the text. Footnotes, numbered consecutively, may be placed on a separate sheet at the end of the paper. A condensed title or running page head of not more than thirty-five letters should be included. Fii/urcs. The dimensions of the printed page, 5 by 7% inches, should be kept in mind in preparing figures for publication. Illustrations should be large enough so that all details will be clear after appropriate reduction. Explana- tory matter should be included in legends as far as possible, not lettered on the illustrations. Figures should be prepared for reproduction as line cuts or half- tones; other methods will be used only at the author's expense. Figures to be reproduced as line cuts should be drawn in black ink on white paper or blue- lined co-ordinate paper; those to be reproduced as halftones should be mounted on Bristol board and any designating letters or numbers should be made di- rectly on the figures. The author's name should appear on the reverse side of all figures. The desired reduction should be specified on each figure. Literature cited. The list of literature cited should conform to the style set in this issue of The Biological Bulletin. Papers referred to in the manuscript should be listed on separate pages headed "Literature Cited." Mailing. Manuscripts should be packed flat. Large illustrations may be rolled in a mailing tube, but all illustrations larger than 9 by 12 inches must be accompanied by photographic reproductions or tracings that may be folded to page size. Reprints. Authors will be furnished, free of charge, one hundred reprints without covers. Additional copies may be obtained at cost; approximate figures will be furnished upon request. THE BIOLOGICAL BULLETIN THK BIOLOGICAL P>ULLETIX is issued six times a year at the Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Pennsylvania. Subscriptions and similar matter should he addressed to The Biologi- cal Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for (in-at Britain: Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New < >xt~<>nl Street, London, W/C. 2. Single numbers, Subscription per volume (three issues), $4.50. ( ommunications relative to manuscripts should be sent to the Manag- Kditnr. Marine Biological Laboratory. Woods Hole, Massachusetts, between July 1 and ( )ctober 1, and to the Department of Zoology, Wash- ington University, St. Louis. Missouri, during the remainder of the year. Entered as second-class matter May 17, 1930, at the post office at Lancaster, Pa., under the Act of August 24, 1912. BIOLOGY MATERIALS The Supply Department of the Marine Biological Labora- tory has a complete stock of excellent plain preserved and injected materials, and would be pleased to quote prices on school needs. PRESERVED SPECIMENS for Zoology, Botany, Embryology, and Comparative Anatomy LIVING SPECIMENS for Zoology and Botany including Protozoan and Drosophila Cultures, and Animals for Experimental and Laboratory Use. MICROSCOPE SLIDES for Zoology, Botany, Embryology, Histology, Bacteriology, and Parasitology. CATALOGUES SENT ON REQUEST Supply Department MARINE BIOLOGICAL LABORATORY Woods Hole, Massachusetts CONTENTS Page HUGHES-SCHRADER, SALLY A Primitive Coccid Chromosome Cycle in Puto Sp 167 KENK, ROMAN Ecological Observations on Two Puerto-Rican Echinoderms, Mellita lata and Astropecten marginatus 177 MAST, S. O. AND W. J. BOWEN The Food Vacuole in the Peritricha, With Special Reference to the Hydrogen-Ion Concentration of its Content and of the Cytoplasm 188 ROMANOFF, ALEXIS L. Hydrogen-Ion Concentration of Albumen and Yolk of the Developing Avian Egg 223 SPIEGELMAN, S. AND FLORENCE MOOG On the Interpretation of Rates of Regeneration in Tubularia, and the Significance of the Independence of Mass and Time 227 SCHARRER, BERTA AND ERNST SCHARRER Neurosecretion VI. A Comparison Between the Intercere- bralis-Cardiacuni-Allatum System of the Insects and the Hypothalamo-Hypophyseal System of the Vertebrates 242 MARSLAND, DOUGLAS Mechanism of Pigment Displacement in Unicellular Chro- matophores 252 MBL/WHOI LIBRARY UH 17 Jb