TRANSACTIONS OF THE ACADEMY OF SCIENCE OF ST. LOUIS. VOL. VIII. JANUARY 1898 to DECEMBER 1898. PUBLISHED UNDER DIRECTION OF THE COUNCIL. 8T. LOUIS : NIXON-JONES PRINTING CO. xr CONTENTS. PAGE. Table of Contents iii List of Members. Revised to Dec. 31, 1898: 1. Patrons v 2. Active Members v History of the Academy (Abstract) xiii Record. January 1 to December 31, 1898 xvii Papers Published. January 1 to December 31, 1898: 1. Francis E. Nipher. — A method of measuring the pressure at any point on a structure, due to wind blowing against that structure. — Plates I. -II. — Issued January 14, 1898 1 2. Hermann von Schrenk. — The trees of St. Louis as in- fluenced by the tornado of 1896. —Plates III. -IX.— Issued January 29, 1898 25 3. Charles Robertson. — New or little known North American bees. — Issued March 3, 1898 43 4. A. S. Hitchcock. — Ecological plant geography of Kansas. — Issued March 8, 1898 55 5. Frank Collins Baker. — Themolluscan fauna of West- ern New York. — Plate X. — Issued March 22, 1898.. 71 6. Calvin M. Woodward. — The efficiency of gearing under friction. — Issued- May 10, 1898 95 7. Carl Kinsley. — Discussion of series dynamo-electric machines. — Issued May 20, 1898 107 8. Edmund A. Engler. — The normal to the conic sec- tion. — Issued July 15, 1898 137 9. Walter C. G. Kirchner. — Contribution to the fossil flora of Florissant, Colorado. — Plates XI.-XV. — Issued December 10, 1898 161 10. Hermann von Schrenk. — On the mode of dissemina- tion of Usnea barbata. — Plate XVI. — Issued Decem- ber 17, 1898 189 iv Contents. Papers Published — Continued. page 11. L. H. Pammel. — The histology of the caryopsis and endosperm of some grasses. — Plates XVII. -XIX. — Issued December 29, 1898 199 12. Title-page, prefatory matter and index of Vol. VIII. Record from January 1 to December 31, 1898. — Issued January 30, 1899. List of Authors 221 General Index 222 Index to Genera 224 MEMBERS. 1. PATRONS. Harrison, Edwin 3747 Westminster pi. 2. ACTIVE MEMBERS. Adkins, James 3901 Park av. Alden, John T Colonial bldg. Alt, Adolph 3036 Locust st. Bailey, Charles 87 Vandeventer pi. Bain, Robert E. M Century bldg. Ball, David C 310 Merchants' Exch. Barck, C 2715 Locust st. Bartlett, George M 215 Pine st. Baumgarten, G 2643 Chestnut st. Becktold, Wm. B 212 Pine st. Bernays, A. C 3623 Laclede av. Biebinger, F. W 1421 S. 11th st. Bixby, W. K 13 Portland pi. Bliss, M. A 4929 Lotus av. Boogher, John H 4034 Delmar boul. Bouton, Charles L 2709 Park av. Branch, Henry 4314 Washington av. Bremer, Ludwig 3723 West Pine boul. Brennan, Martin S 1414 O'Fallon st. Brookings, Robert S 2329 Locust st. Brown, Daniel S 2212 DeKalb st. Bryson, John P 209 N. Garrison av. Budgett, S. P 3810 Washington av. Burg, Wm 1756 Missouri av. Burnett, E. C University Club. Burroughs, W. S 240 Dickson st. Busch, Adolphus Busch pi. Cale, George W., Jr 4403 Washington boul. Carpenter, George O Russell and Compton avs. Carter, Howard Planters' Hotel. vi Trans. Acad. Sci. of St. Louis. Chaplin, Winfield S 3636 West Pine boul. Chase, E. C 3325 Morgan st. Chauvenet, Louis 5501 Chamberlain av. Chouteau, Charles P 918 Security bldg. Clifford, Alfred 4168 West Pine boul. Close, James A 2031 Olive st. Collins, Robert E 3811 Westminster pi. Compton, P. C 4156 Washington boul. Comstock, T. G 3401 Washington av. Coulter, John M University of Chicago, Chicago, 111. Crandall, Geo. C 620 N. Garrison av. Crunden, Frederick M Public Library. Curtis, Wm. S , St. Louis Law School. Cushman, Allerton S Harvard University, Cambridge, Mass. Dameron, E. C. Clarksville, Mo. Davis, H. N 56 Vandeventer pi. Davis, John D 421 Olive st. Diehm, Ferdinand 1835 Kennett pi. Dodd, S. M 415 Locust st. Douglas, A. W .9th and Spruce sts. Drake, George S Boatmen's Bank. Duenckel, F. W 1936 Louisiana av. Durant, George F 9 Benton pi. Eggert, Henry 1001 Collinsville av., East St. Louis, 111. Eberlein, W. P 3655 Cleveland av. Eliot, E. C 5468 Maple av. Eliot, Henry W 2635 Locust st. Engler, Edmund A Washington University. Erker, A. P 617 Olive st. Espenschied, Charles 3500 Washington av. Euston, Alexander 3730 Lindell boul. Evers, Edward 1861 N. Market st. Ewing, A. E 3333 Washington av. Ferriss, Franklin 5828 Cabanne av. Fischel, W. E 2647 Washington av. Flersheim, George W 1008 Locust st. Forbes, S. A Urbana, 111. Members. vii Fordyce, John R 3634 Washington boul. Francis, D. R 4421 Maryland av. Frankenfield, H. C U. S. Weather Bureau, Washington, D. C. Frerichs, Frederick W Herf & Frerichs Chem. Co. Fruth, Otto J 3066 Hawthorne boul. Fry, Frank R 3133 Pine st. Funkhouser, R. M 3534 Olive st. Gazzam, J. B 514 Security bldg. Glasgow, Frank A 2608 Locust st. Glasgow, William C 2847 Washington av. Goodman. 0. H 3329 Washington av. Gottschalk, Fred. F 619 Pine st. Graham, B. B 3500 Morgan st. Graves, W. W 1943 N. 11th st. Gray, M. L 604 Houser bldg. Green, John 2670 Washington av. Gregory, E. H., Jr 3525 Lucas av. Grindon, Joseph 509 N. Theresa av. Grocott, Willis H 1812 Coleman st. Gurney, James Tower Grove and Magnolia avs. Haarstick, Henry C Main and Walnut sts. Hambach, G Washington University. Hammon, W. H U. S. Weather Bureau, San Francisco, Cal. Hardaway, W. A 2922 Locust st. Hartmann, R 14 S. 2d st. Herthel, Adolph 1739 Waverly pi. Herzog, William 2321 Whittemore pi. Hirschberg, F. D 3818Lindell boul. Hitchcock, A. S Manhattan, Kas. Hitchcock, Henry 709 Wainwright bldg. Hodgman, Charles 300 N. 14th st. Holman, M. L 3744 Finney av. Holmes, J. M 3810 Page av. Hough, Warwick 3877 Washington boul. Hugunin, F. U 915 Olive st. Hunicke, H. A 3532 Victor st. Hurter, Julius 2346 S. 10th st. Ives, Halsey C Museum of Fine Arts. viii Trans. Acad. Set. of St. Louis. James, John A. James 2836 Lafayette av. Jester, E. T 0/o Liggett & Myers Co. Jewett, Eliot C Monterey, Mex. Johnson, Dr. J. B 4244 Washington boul. Johnson, Professor J. B Washington University. Johnson, R. DeO Flat River, Mo. Jones, Breckinridge 4th and Pine sts. Judson, Frederick N 421 Olive st. Kennett, A. Q 2916 Lucas av. Keyes, Charles R 944 Fifth st., Des Moines, la. Kinealy, J. H... Washington University. King, Goodman 78 Vandeventer pi. Kinner, Hugo 1103 Rutger st. Kinsley, Carl Washington University. Kinsman, G. C Decatur, 111. Kline, George R 215 Pine st. Kodis, Theo 3450 Sidney st. Kolbenheyer, F 2006 Lafayette av. Krall, George W Manual Training School. Kribben, B. D 701 Bank of Commerce bldg. Kromrey, Hugo 513 Walnut st. Lackland, R. J 1623 Locust st. Langsdorf, A. S 3133 Laclede av. Lazell, E. W Spencer, Mass. Leighton, George B 803 N. Garrison av. Leighton, George E 803 N. Garrison av. Lemoine, E. S 1622 Washington av. Letterman, George W Allenton, Mo. Lichter, John J., Jr 5305 Virginia av. Loeb, H. W 3536 Olive st. Lumelius, J. Geo 1225 St. Ange av. Mack, Charles J 113 N. Broadway. Madill, George A 4140 Lindell boul. Mallinckrodt, Edward 26 Vandeventer pi. Markham, George D Colonial bldg. Matthews, Leonard 300 N. 4th st. McElwee, L. C 215 S. Jefferson av. McMillan, Wm 25 Portland pi. Meier, Frederick 2340 Whittemore pi. Meier, Theo. G 421 Olive st. Members. ix Merrell, Albert 3814 Washington boul. Miller, Charles F 1751 Missouri av. Moore, Robert 61 Vandeventer pi. Morton, I. W 9th and Spruce st. Mudd, Harvey G 2604 Locust st. Mudd, Henry H 2604 Locust st. Muegge, Aug. H 2712 Franklin av. Nagel, Charles 2044 Lafayette av. Nelson, N. 0 8th and St. Charles sts. Neville, J. E 3960 Delmar boul. Nipher, Francis E Washington University. Norton, J. B. S Mo. Botanical Garden. Olshausen, E. P 1115 Rutger st. Olshausen, George R 13 Artillerie st., Berlin, Germany. O'Reilly, A. J 33 City Hall. O'Reilly, Robert J 3411 Pine st. Overstolz, Herman 100 N. Broadway. Pammel, L. H Ames, la. Parker, Geo. W 417 Pine st. Parsons, Charles 2804 Pine st. Pegram, George H 195 Broadway, New York City. Pettus, W. H. H 2834 Chestnut st. Pickard, John ..Columbia, Mo. Pike, S. B Amer. Central bldg. Pitzman, Julius 1900 S. Compton av. Post, M. Hayward 3015 Lucas av. Potter, William B ....3312 Washington av. Preetorius, Emil c/0 Westliche Post. Prewitt, T. F 3101 Pine st. Primm, Alexander T C/Q J- Kennard & Sons. Pritchett, Henry S Supt. U. S. Coast Survey, Washington, D. C. Pulsifer, William H Newton Centre, Mass. Quaintance, A. L Lake City, Florida. Ramel, A Washington University. Ravold, Amand 2806 Morgan st. Robert, E. S Union Trust bldg. Robertson, Charles Carlinville, 111. x Trans. Acad. Sci. of St. Louis. Roever, William H 3628 St. Louis av. Rolfs, P. H Lake City, Fla. Runge, Edward C Supt. Insane Asylum. Russell, Colton 4652 Wagoner pi. Sander, Enno 129 S. 11th st. Sanger, Charles R Washington University. Schmalz, Leopold 2824 Shenandoah av. Schneck, Jacob Mt. Carmel, 111. Schwarz, Henry 1723 Chouteau av. Schweitzer, P Columbia, Mo. Scott, Henry C 64 Vandeventer pi. Seddon, James A 1515 Locust st. See, T. J. J Montgomery City, Mo. Seever, William J 1600 Locust st. Senseney, E M 2829 Washington av. Sheldon, W. L 4200 Morgan st. Shepley, John F 60 Vandeventer pi. Simmons, E. C... 9th and Spruce sts. Simmons, W. D Simmons Hardware Co. Sluder, Greenfield,.. 2647 Washington av. Smith, D. S. H 3646 Washington boul. Smith, Holmes Washington Universi ty . Smith, Irwin Z 87 Vandeventer pi. Smith, Jared G Dept. Agriculture, Washington, D. C. Spencer, H. N 2725 Washington av. Spiegelhalter, Joseph 2166 Lafayette av. Starr, John E West and Horatio sts., N. Y. City. Staudinger, B 1703 Olive st. Stedman, J. M Columbia, Mo. Stevens, Charles D 1749 S. Grand av. Stone, Chas. H Harbor Commr., City Hall. Straub,A. W 1553 S. Grand av. Strauss, Julius C 3516 Franklin av. Sutter, Otto City Hospital. Taussig, Albert E 2647 Washington av. Taussig, William 3447 Lafayette av. Teichmann, William C 1141 Market st. Terry, Robert J Mo. Medical College. Thacher, Arthur 4109 Washington boul. Thomas, John R 4th and Vine sts. Members. xi Timmerman, A. H Rolla, Mo. Tittmann, Eugene C 1811 Kennett pi. Tittmann, Harold H 3726 Washington boul. Trelease, William Mo. Botanical Garden. Turner, John W 717 N. Garrison av. Tyrrell, Warren A c/o St. L., P. & M. Ry. Co., Springfield, 111. Updegraff, Milton Columbia, Mo. Valle, Jules F 3303 Washington av. Vanlandingham, A. J 500 Chamber of Commerce. Vickroy, W. R 3029 Washington av. Von Schrader, George F Wainwright bldg. VonSchrader, Otto U 3749 Westminster pi. Von Schrenk, Hermann 1724 Washington av. Wall, L. J. W 21st and Morgan sts. Walsh, Edward, Jr Miss. Glass Co. Watts, M. F 4362 Morgan st. Wheeler, H. A 3124 Locust st. Whelpley, Milton H 2342 Albion pi. Whitaker, Edwards 300 N. 4th st. Whitten, J. C Columbia, Mo. Whittier, Charles T 2727 01ivest. Winkelmeyer, Christopher 3540 Chestnut st. Winslow, Arthur Lyceum bldg. , Kansas City, Mo. Wislizenus, Fred 1817 Longfellow av. Witt, ThomasD 6th and Olive sts. Wittenberg, Paul 2d and Pine sts. Wood, O. M 3016 Caroline st. Woodward, C. M 3013 Hawthorne boul. Yeatman, James E 1410 E. Grand av. Abstract of History. xiii THE ACADEMY OF SCIENCE OF ST. LOUIS. ORGANIZATION. The Academy of Science of St. Louis was organized on the 10th of March, 1856, in the hall of the Board of Public Schools. Dr. George Engelmann was the first president. CHARTER. On the 17th of January following, a charter incorporating the Academy was signed and approved, and this was accepted by vote of the Academy on the 9th of February, 1857. OBJECTS. The act of incorporation declares the object of the Academy to be the advancement of science and the establishment in St. Louis of a museum and library for the illustration and study of its various branches, and provides that the members shall acquire no individual property in the real estate, cabinets, library, or other of its effects, their interest being usufruc- tuary merely. The Constitution, as adopted at the organization meeting and amended at various times subsequently, provides for hold- ing meetings for the consideration and discussion of scientific subjects; taking measures to procure original papers upon such subjects; the publication of transactions ; the establish- ment and maintenance of a cabinet of objects illustrative of the several departments of science, and a library of works relating to the same ; and the establishment of relations with other scientific institutions. To encourage and promote special investigation in any branch of science, the formation of special sections under the charter is provided for. MEMBERSHIP. Members are classified as active members, corresponding members, honorary members, and patrons. Active member- xiv Trans. Acad. Sci. of St. Louis. ship is limited to persons interested in science, though they need not of necessity be engaged in scientific work, and they alone conduct the affairs of the Academy, under its Constitu- tion. Persons not living in the city or county of St. Louis, who are disposed to further the objects of the Academy by original researches, contributions of specimens, or otherwise, are eligible as corresponding members. Persons not living in the city or county of St. Louis are eligible as honorary mem- bers by virtue of their attainments in science. Any person conveying to the Academy the sum of one thousand dollars or its equivalent becomes eligible as a patron. Under the By-Laws, resident active members pay an initia- tion fee of five dollars and annual dues of six dollars. Non- resident active members pay the same initiation fee, but annual dues of three dollars only. Patrons, and honorary and corresponding members, are exempt from the payment of dues. Patrons and all active members not in arrears are entitled to one copy of each publication of the Academy issued after their election. Since the organization of the Academy, 824 persons have been elected to membership, of whom, at the present time, 251 are carried on the active list. One person, Mr. Edwin Harrison, has been elected a patron. The present list of cor- responding members includes 204 names. OFFICERS AND MANAGEMENT. The officers, who are chosen from the active members, con- sist of a President, two Vice-Presidents, Recording and Cor- responding Secretaries, Treasurer, Librarian, three Curators, and two Directors. The general business management of the Academy is vested in a Council composed of the President, the two Vice-Presidents, the Recording Secretary, the Treas- urer and the two Directors. The office of President has been filled by the following well-known citizens of St. Louis, nearly all of whom have been eminent in some line of scientific work: George Engel- mann, Benjamin F. Shumard, Adolphus VVislizenus, Hiram A. Prout, Dr. John B. Johnson, James B. Eads, William T. Abstract of History. xv Harris, Charles V. Riley, Francis E. Nipher, Henry S. Pritchett, John Green, Melviu L. Gray, and Edmund A. Engler. MEETINGS. The regular meetings of the Academy are held at its rooms, 1600 Locust Street, at 8 o'clock, on the first and third Monday evenings of each month, a recess being taken from the second June meeting to the first October meeting, inclusive. These meetings, to which interested persons are always welcome, are devoted in part to the reading of technical papers designed for publication in the Academy's Transactions, and in part to the presentation of more popular abstracts of recent investi- gation or progress. From time to time, public lectures cal- culated to interest a larger audience are provided for in some suitable hall. LIBRARY. After its organization, the Academy met in Pope's Medical College, where a creditable beginning had been made toward the formation of a museum and library, until May, 1869, when the building and museum were destroyed by fire, the library being saved. The library now contains some 12,900 books and 9,000 pamphlets, and is open during certain hours of the day for consultation by members and persons engaged in scientific work. PUBLICATIONS AND EXCHANGES. Eight octavo volumes of Transactions, averaging 665 pages, have been published since the organization of the Academy, and widely distributed. Two quarto publications have also been issued, one from the Archaeological section, being a contribution to the archaeology of Missouri, and the other a report of the observations made by the Washington University Eclipse Party of 1889. The Academy now stands in exchange relations with 527 institutions or organizations of aims similar to its own. xvi Trans. Acad. Sci. of St. Louis. MUSEUM. Since the loss of its first museum, in 1869, the Academy has lacked adequate room for the arrangement of a public museum, and, although small museum accessions have been received and cared for, its main effort of necessity has been concentrated on the holding of meetings, the formation of a library, the publication of worthy scientific matter, and the maintenance of relations with other scientific bodies, through its active membership, which includes many business and professional men who are interested in the work and objects of the Academy, although not themselves investigators. December 31, 1898. Record. xvii RECORD From January 1, 1898, to December 31, 1898. January 3, 1898. President Gray in the chair, nineteen persons present. The President presented a report on the condition of the Academy for the year 1897.* The Treasurer submitted his annual report, showing invested funds to the amount of $6,000.00, and a balance carried forward to the year 1898 of $603. 41. f The Librarian submitted his annual report. % The nominating committee reported that 95 ballots had been counted, and the following officers for 1898 were declared elected: — President Edmund A. Engler. First Vice-President Robert Moore. Second Vice-President D. S. H. Smith. Recording Secretary William Trelease. Corresponding Secretary Joseph Grindon. Treasurer Enno Sander. Librariau Gustav Hambach . Curators Gustav Hambach, Julius Hurter. Directors M. H. Post, Amand Ravold. It was announced, as a result of the letter ballot, that Arti- cle IV. of the Constitution had been amended so as to read "Three Curators " instead of " Board of Curators." President-elect Engler, being called to the chair, briefly * Transactions 7: lxxvii. t Transactions 7: lxxviii. % Transactions 7: lxxix. xviii Trans. Acad. Sci. of St. Louis. addressed the Academy, and in the course of his remarks called attention to the long and faithful service of Mr. M. L. Gray, the retiring President. On motion, a committee, composed of Professor Nipher, Mr. Trelease and Dr. Sander, was appointed to prepare a resolution expressive of the Academy's appreciation of Mr. Gray's services. Dr. Amand Ravold spoke informally of formaldehyde gas as a disinfectant, and exhibited several forms of apparatus adapted to its use. It was stated that although, in confined spaces, the gas has proved to be an effective disinfectant, which has the merit of not injuring the most delicate fabrics or polished metal surfaces, its germicide action in dwelling rooms has thus far proved less satisfactory than that of sul- phur dioxide and chlorine, so far as it has been tested by the Health Department of the City of St. Louis, so that, as yet, the Health Department has not found it possible to employ it as a substitute for the older and in some respects more objec- tionable disinfectants. Two persons were proposed for active membership. January 17, 1898. President Engler in the chair. The committee appointed at the last meeting to take suit- able action on the withdrawal of Mr. M. L. Gray from active participation in the affairs of the Academy submitted the following report : — To the President and Members of the St. Louis Academy of Science: Gentlemen : Your committee, appointed to recommend suitable action on the withdrawal of Mr. M. L. Gray from service in your council, beg leave to report as follows : — The committee does not feel that we are in any sense taking leave of our associate and friend, but we feel that his long and devoted service to the Academy has entitled him to ask our consideration in seeking relief from the drudgery and responsibilities of business and executive details. Surely it was not the desire to do work of this kind that has caused him for so many years to be so constantly in attendance upon our meetings. We know that his request to be relieved of such duties is not due to any decrease in his interest in the cause which the Academy represents. We know that his zeal in promoting the higher interests of the Academy will still continue. But it is proper and fitting that we should at this time assure Record. xix him of our grateful appreciation of his long and faithful services to the Academy. And we hope that we shall continue for many years to enjoy his genial presence among us. A paper by Mr. Charles Robertson, entitled New or little known North American Bees, was read in abstract. Prof. A. C. Bernays addressed the Academy on biological facts as evidence of man's place in nature. He illustrated certain facts from the ontogeny of man by description and blackboard sketches, and tried to explain the anatomical pecu- liarities in the structure of man and the lower animals by the biogenetic law of Haeckel. He also made some suggestions about the best method of studying and of teaching anatomy. It was claimed that in the biogenetic law of Haeckel a scien- tific background, or rather a working hypothesis, was given, by means of which the recorded facts of zoology, botany, paleontology, etc., were made understandable and really be- came useful to science. He also gave a definition and illus- tration of the meaning of the term differentiation as used in biology. Mr. George W. Parker, Dr. Otto Sutter and Mr. O. M. Wood, of St. Louis, were elected to active membership. Twenty-four persons were proposed for active member- ship. February 7, 1898. President Engler in the chair, fourteen persons present. Mr. J. B. S. Norton read a paper by Professor A. S. Hitch- cock, on the ecological plant geography of Kansas. Professor L. H. Pammel spoke informally on the anatomi- cal characters of seeds from the standpoint of systematic botany. The following persons were elected active members : James Adkins, Robert E. M. Bain, W. K. Bixby, Dr. James A. Close, Dr. Geo. C. Crandall, Wm. S. Curtis, George W. Flersheim, D. R. Francis, Dr. R. M. Funkhouser, Fred. F. Gottschalk, F. U. Hugunin, Breckinridge Jones, Frederick N. Judson, George R. Kline, John J. Lichter, Jr., Dr. H. W. Loeb, Wm. McMillan, Dr. Harvey G. Mudd, J. B. S. Norton, Warren A. Tyrrell, Dr. Jules F. Valle, A. J. Vanlandingham, xx Trans. Acad. Set. of St. Louis. Dr. Henry Milton Whelpley, of St. Louis ; Henry Eggert, of East St. Louis, Illinois. Sixteen persons were proposed for active membership. February 21, 1898. President Engler in the chair, thirteen persons present. Dr. R. J. Terry exhibited a specimen of a cervical rib from a human subject, and discussed the occurrence of structural anomalies of this character. The following persons, resident in St. Louis, were elected active members: Win. B. Becktold, Henry Branch, Wm. Burg, John D. Davis, Otto J. Fruth, J. B. Gazzam, Dr. W. W. Graves, Willis H. Grocott, A. Q. Kennett, Leonard Matthews, Theo. G. Meier, Dr. Albert Merrell, Aug. H. Muegge, Alexander T. Primm, John R. Thomas. Seven persons were proposed for active membership. March 7, 1898. President Engler in the chair, twenty-eight persons present. Professor C. M. Woodward presented a paper embodying an analytical discussion of the efficiency of gearing under friction, spur wheels with epicycloidal and involute teeth being considered. Dr. Amand Ravold demonstrated the method, recently intro- duced by Philip Hanson Hiss, of differentiating the typhoid bacillus from Bacillus coli-communis , by the use of semi-solid acidulated media, in which, at blood temperature, the round col- onies of the typhoid bacillus assume a peculiar fimbriated form of growth, because of the motility of the bacteria in the slightly yielding medium, which in most cases readily distinguishes them from the more whetstone-shaped colonies of the colon bacillus, which does not produce the peculiar fimbriation in plate cultures. In tube cultures in the same general medium, but prepared with a slighter acidity and somewhat less solid- ity, a uniform clouding of the entire tube, due to the swarm- ing of the bacteria, was shown to be characteristic of the typhoid bacillus, while the colon bacillus was definitely con- Record. xxi fined to the immediate vicinity of the thrust. The media in both cases are made up without peptone. The formulae are: — For plate cultures : For tube cultures : Agar 10 grams. Agar 5 grams. Gelatine 25 " Gelatine 80 " Beef extract 5 " Beef extract 5 " Glucose 10 " Glucose 10 " Salt 5 " Salt 5 " Normal acid 20 cc. Normal acid IS cc. The whole increased to 1000 cc. The whole increased to 1000 cc. The growth of the two species in question, on potato and in milk cultures with litmus, was also demonstrated. The following persons, resident in St. Louis, were elected active members: David C. Ball, Dr. E. C. Burnett, Ferd- inand Diehm, Franklin Ferriss, Dr. L. C. McElwee, Chas. H. Stone, VV. S. Vickroy, L. J. W. Wall. Four persons were proposed for active membership. March 21, 1898. President Engler in the chair, fifteen persons present. Professor J. B. Johnson spoke informally on some aspects of the manufacture and testing of Portland cement, reviewing the historical development of the cement industry from the time of Smeaton, and showing that the first true Portland cement had been made and patented in Yorkshire, England, in 1825, though it was not until about 1850 that it was known in London ; and that its manufacture in Germany began at Stettin, in 1853, while it was not until 1875 that it was made in the United States. The statement was made that the increase in the use of this material now outruns the rapidly increasing production. The various methods of manufacture were outlined, and the theory was explained on which the various processes are based. The accepted theory of the hardening ingredients was given, and some of the standard methods of testing were described. Two improvements that the speaker hoped to see introduced in the standard methods of testing were considered in detail, — one, a modification of xxii Trans. Acad. Sci. of St. Louis. the form of the tension briquette, in which the stress should be more uniformly distributed ; and the other, a mixing of two fairly uniform sizes of sand, one of which should just about fill the interstices of the other, so as to develop the benefits of fine grinding of the cement, which, it was ex- plained, were lost when large gaps remained between the particles of sand, to be filled by the cement. April 4, 1898. No quorum present. April 18, 1898. President Engler in the chair, eighteen persons present. Mr. Carl Kinsley read a paper on Series Dynamo Electric Machines. Professor J. H. Kinealy made some informal remarks on the ventilation of schools, and by means of stereopticon views showed the different methods adopted for supplying the air re- quired for ventilation to the different rooms of school houses. The following persons, resident in St. Louis, were elected active members : W. P. Eberlein, Alexander Euston, Harold H. Tittman, George F. von Schrader. One person was proposed for active membership. May 2, 1898. President Engler in the chair, sixteen persons present. Mr. M. L. Holman, Water Commissioner of St. Louis, made an informal address on Present Methods in Water Filtration. He described the municipal systems of water filtration used at Hamburg, Berlin and other German cities which he had visited, and spoke of the problems to be solved in designing a filter plant for St. Louis. He spoke of the different methods of coagulating and getting rid of mud and sediment in water, and said he thought it was probable that Record. xxiii a combination coagulating and filtering system would have to be used to purify the water for St. Louis. Mr. Julius Pitzman, of St. Louis, was elected an active member. May 16, 1898. President Engler in the chair, thirteen persons present. Mr. J. A. Seddon delivered an informal address on Floods in the Mississippi System. He discussed the floods of the Missouri and upper Mississippi rivers with regard to their relations to one another and to the floods of the lower Missis- sippi. He spoke of the great damage done by the floods to plantations and towns on the banks of the lower Mississippi, and of the means to be adopted for controlling the floods and preventing, as far as possible, this damage. His remarks were illustrated by maps and diagrams showing the flooded regions and the various quantities of water discharged by the rivers at different stages. One person was proposed for active membership. June 6, 1898. President Engler in the chair, nineteen persons present. Dr. Amand Ravold read a paper on Consumption from an Economic and Humanitarian Standpoint. He quoted statistics showing the yearly loss of life by consumption in St. Louis, and stated that on account of disabilities and deaths by con- sumption the loss to the city was not less than one and' one- half million dollars yearly. He said that he did not believe in the isolation of consumptives, as the communication of the disease to well persons could be prevented by the exercise of simple sanitary precautions. He condemned the habit of spit- ting as unnecessary, dirty, and detrimental to the health of the community. Dr. Howard Carter made a few remarks in regard to the occurrence of tuberculi in milk, and the danger of using milk from unhealthy cows. Mr. Carl Kinsley showed the different methods of deter- mining the frequency of an alternating current, and dem- xxiv Trans. Acad. Set. of St. Louis. onstrated that the frequency could be determined by the vibration of an air column, in a brass or glass tube, set in motion by the current. Mr. W. C. G. Kirchner presented a paper entitled Con- tributions to the Fossil Flora of Florissant, Colorado. Mr. M. L. Holman, of St. Louis, was elected an active member. One person was proposed for active membership. October 17, 1898. President Engler in the chair, ten persons persent. Mr. C. H. Thompson spoke of some interesting stylar movements of certain Marantaceae, connected with their pol- lination, his observations referring to Maranta, Galathea and Thalia . Seven persons were proposed for active membership. November 7, 1898. President Engler in the chair, fourteen persons present. Mr. James A. Seddon read a paper on Resistance to Flow in Hydraulics, in which the point was made that relatively a small part of this resistance, so far as open streams were concerned, was directly attributable to friction against the bottom and limiting banks, but that the resistance was found acting between accelerations and impacts, and showed in forced distortions of the free surface, from which forms the energy passed into internal motion. The following persons were elected active members: William Herzog, Dr. Theodore Kodis, A. S. Langsdorf, J. George Lumelius, Herman Overstolz, of St. Louis; A. L. Quaintance, of Lake City, Florida; Dr. T. J. J. See, of Montgomery City, Missouri. Two persons were proposed for active membership. Record. xxv November 21, 1898. President Engler in the chair, eleven persons present. Dr. G. Hambach presented an oral abstract of a paper en- titled A Revision of the Morphology of the Blastoideae, illustrating his remarks with numerous specimens and photo- graphs. The following persons, resident in St. Louis, were elected active members : Dr. Howard Carter, William H. Roever. December 5, 1898. President Engler in the chair, twenty-six persons present. Mr. H. von Schrenk presented by title a paper On the Mode of Dissemination of Usnea barbata. Professor L,. H. Pammel presented by title a paper on the Histology of the Caryopsis and Endosperm of some Grasses. Dr. Theodore Kodis stated the results of some experiments on overcooling animal and vegetable tissues, in which it was shown that, as water may, under favorable conditions, be cooled to some distance below zero, centigrade, without freezing, — the temperature immediately rising to the freez- ing point the moment that freezing begins, and remaining there until the water is entirely solidified, then beginning once more to drop, — so, when animal and vegetable tissues are experimented on, they may be cooled to a temperature de- cidedly lower than the freezing point, under favorable condi- tions, before freezing begins, but that, when it begins, the temperature at once rises to the freezing point (which is always somewhat lower than that of pure water), remaining there until the process of freezing is complete, when it once more begins to fall. The speaker gave a short account of the current theories as to the mechanical constitution of proto- plasm, and discussed the bearing on them of the phenomena when the solidification of overcooled tissues began. Professors Nipher, Budgett and Green were elected a com- mittee for the nomination of officers for the year 1899. xxvi Trans. Acad. Sci. of St. Louis. December 19, 1898. President Engler in the chair, sixteen persons present. The nominating committee reported the following nomina- tions for officers for the year 1899 : — President Edmund A. Engler. First Vice-President Robert Moore. Second Vice-President D. S. H. Smith. Recording Secretary William Trelease. Corresponding Secretary Joseph Grindon. Treasurer Enno Sander. Librarian G. Hambach. Curators G. Hambach, Julius Hurter, Hermann von Schrenk. Directors M. H. Post, Amand Ravold. Mr. O. M. Wood addressed the Academy on the Sociology of the Negro, emphasizing the negro's faithfulness, cheerful- ness, and love of music ; and in speaking of the prospects of the negro race for the future he laid stress on the fact that the negro is ever grateful for assistance from the white man, and fully realizes that his advancement is impossible without such assistance. The speaker's remarks were illustrated by lantern slides showing typical negroes in the savage state. Reports of Officers for the Year 1898. Submitted January 9, 1899. The President addressed the Academy as follows: — Gentlemen: It devolves upon me, according to the time-honored custom of Presi- dents of this Academy, to make some remarks upon the present status o our society and the prospects for the future. I am glad to be able to con- gratulate the Academy upon its present condition. This Academy has never been without a nucleus of earnest workers in science, who form the life of the society, whose devotion to science enables us, in our modest way, to carry on our meetings in a truly scientific spirit, to devote our thoughts and attention, for at least two nights every inontb, to subjects which are aside Record. xxvii entirely from the thoughts which actuate the greater part of our community. We are not led astray, and never have been led astray, by the spirit of com- mercialism which rules not only in our city but throughout our country; and I can safely say that we have kept the flame of pure science burning, though we have not been able to spread its light as broadly as we should like. Yet it is a satisfaction to know that we do not lack for the devotion which has always characterized the members of this society, and in which the hope of the future lies. I am glad to be able to report, as you can see in detail from the reports of the other officers, a very satisfactory progress in a number of different directions. In the membership I can report that during the past year we have added a net increase of 49, making at present a total active membership of 251. While the larger part of this membership consists of persons who do not attend our meetings, yet we are assured of their sympathy and encourage- ment, if by no other thing, by the mere fact that they are willing to con- tribute their annual dues for the sake of enabling the few who do the work to carry on the society and to print its publications. But in retaining and striving to increase this class of members, the Academy assumes a recip- rocal obligation, and it is no doubt desirable that at least some of the meetings of the Academy should be devoted to the consideration of the growth of scientific knowledge, not necessarily such growth as has been brought about by the members of our own Academy, but the world's progress in science. To do this implies that we should have, at least occasionally, a statement of results in scientific work, in a form not too technical, in order that we may appeal to the larger part of the membership that we now have, retain their interest, and also increase the membership. The Council of the Academy has had such general lectures upon scientific progress in mind, and has made some effort to provide such popular even- ings. It is a matter, however, of no little difficulty to find the man and the subject which will appeal to this larger class of members. Still, it is a thing which should be kept in mind, and I have no doubt that, as occasion offers, such discourses will be provided. With reference to the meetings of the Academy, I may again congratulate the society. We have held fifteen regular meetings during the year, and the average attendance has been sixteen. While this is a comparatively small number, I think it will compare favorably with that of similar meet- ings of very much larger bodies in other cities, so that we have nothing to be ashamed of. At the same time, I think it is desirable, and I think it entirely feasible, by the means that I have indicated and by others, to increase the attendance. We have made progress, also, with reference to the library. During the year we have increased the number of exchanges which are now receiving our Transactions, and from which we receive periodical publications, quite considerably. The total number of exchanges, as I am informed by the librarian, is now 527, scattered through various parts of the world; and from these societies we have received, during the year, very nearly a thou- sand numbers, — about 390, I think it is, of volumes, and 551 pamphlets, which, as you see, make a very considerable addition to our already exten- sive library. xxviii Trans. Acad. Sci. of St. Louis. Besides the actual increase in the contents of the library, arrangements have been made by the Council to increase the usefulness of tbe library, and I speak of this with considerable satisfaction, because I feel sure that the library is now more accessible and can be more easily used by scientific workers than ever heretofore. I refer particularly to the fact that the library room is now suitably heated, lighted, and furnished with tables on which any student may place the books taken from the shelves and use them. The library is always open, not only to every member of the Acad- emy, but to any person, not a member of the Academy, who satisfies the officers that he is really a student in science and wishes to use and not abuse the books. We have also made some progress with reference to our museum. As you are aware, one of the purposes of the Academy has been, from the very start, and still is, to provide a suitable museum. Progress in that direc- tion has been very slow. The reason is very evident, namely, lack of money; and I doubt very much whether we shall ever be able to make much progress in that line until funds have been secured to enable us to build a permanent museum building, in which may be displayed the collections which we have and which we may acquire. But it would not only be futile, but I doubt whether it would be wise, to attempt to accumulate large collections, or collections of any considerable value, before we have a safe place in which to keep them. This of itself is a sufficient reason for effort in the direction of securing a permanent home for the Academy. This is one of the things that we have always had before us, and it is one which perhaps requires as much effort as anything which the Academy has before it. But notwith- standing the difficulties, we have begun the exhibition of at least one col- lection : I refer to the collection of meteorites, in which there are, I believe, twenty-five specimens, the value of which, as estimated by our librarian, is somewhere between $750 and $ 1,000. This value, of course, is somewhat fictitious, because we do not expect to offer them for sale. These speci- mens are now displayed in a suitable case, and the case stands on the second floor of this building. Space for exhibiting it has been kindly given us by the Historical Society. "We have made progress, also, with reference to our own publications. I think I am safe in saying that the publications of this Academy stand, from a scientific point of view, upon the level of those of any scientific society in the world, not only as to the quality of the material which we publish, — and I congratulate the Academy on the fact that great care has always been exercised in publication, — but also as to the amount in which the Transactions have been issued. To be definite in the statement, I will announce that we are just completing Volume VIII. of our Transactions, which consists of eleven numbers, each a scientific paper, and a twelfth number, which gives the record of the society and various information con- cerning the Academy. This volume will contain, when completed, some- where about 250 pages. This has been done during the year which is just closed, and I am glad to be able to announce that the Council has decided,— with some trepidation, it is true, but still with the purpose of carrying out the plan if it is possible to do so, — to publish an annual volume, a thing which has never been done before by this Academy. The practice has been to publish each paper as ordered by the Council, and, when enough pases Record. xxix had been printed, to close the volume. We have now decided to made a slight change in the manner of publication and to close each volume at the end of the year. This will make the volumes not exactly of the same size, but approximately so. I regard this as a very decided step in advance; but of course it remains for the members of the Academy and others interested to assist in the work, to enable us to do this. It is, however, another source of congratulation to know that never in the history of this Academy, from the eartiest time until now, has a suitable paper been offered, and accepted by the Council, which means have not been at hand or presently found to publish; and I am glad to say that we are still in that pleasant situation, ready to publish any good material that is offered. I therefore think that, looking at the situation of the Academy as it stands at present, we have every reason for congratulation. Effort must be made, not only to retain the present membership but to increase it, to interest scientific workers in the advantages of the Academy, to let them know that there is a means for publishing their work; and such results cannot be achieved without effort, particularly in a busy commercial town like ours. But I feel confident that the future will take care of itself if the earnest workers that we have in our midst will continue in the work simply of adding to scientific knowledge and of spreading that knowledge whenever they have the opportunity. I need not say that I fully appreciate the compliment and the responsi- bility of election for another year. The Treasurer reported as follows : — RECEIPTS. Balance from 1897 $ 603 41 Interest on invested money 379 16 Membership dues 1,515 00 Invested capital returned 1,000 00 $3,497 57 EXPENDITURES. Rent $ 450 00 Current expenses 254 52 Publication of Transactions 677 67 Capital invested [including accrued interest] 1,410 50 2,792 69 Balance to 1899 $704 88 INVESTED FUND. Investment on security $6,400 00 After the Treasurer's report had been read, the President called attention to the mention of an increase in the invested fund from $6,000 to $6,400, and remarked that this was a larger sum than had ever before been invested, adding : — " I wish also to call attention to this point, which I forgot to mention when I was on my feet, namely, that small donations for the permanent xxx Trans. Acad. Sci. of St. Louis. fund will probably be as useful, and can be obtained with perhaps as little effort, as any contributions that could be received for the Academy. We cannot expect that the invested funds will grow very rapidly, but, if we suc- ceed in increasing them even four or five hundred dollars per year, we shall soon have a very respectable income without relying wholly or even in part upon the dues of members." The Librarian reported that during 1898 exchanges had been received from 276 societies, of which five had this year been added to the exchange list. In all 941 numbers were reported as having been added to the library, an increase of 104 as compared with the preceding year. It was reported that during the year the Transactions of the Academy had been distributed to 527 societies or institutions, chiefly by exchange or donation. 7 * Li {ft Transactions of The Academy of Science of St. Louis. VOIi. VIII. No. 1. A METHOD OF MEASURING THE PRESSURE AT ANY POINT ON A STRUCTURE, DUE TO WIND BLOWING AGAINST THAT STRUCTURE. FRANCIS E. NIPHER. Issued January 14, 1898. A METHOD OF MEASURING THE PRESSURE AT ANY POINT ON A STRUCTURE, DUE TO WIND BLOW- ING AGAINST THAT STRUCTURE.* Francis E. Nipher. Various investigations on wind pressures have been made from time to time, by methods which differ essentially from each other. In one form of pressure recorder, an open tube has its mouth directed to the wind by a vane upon which the tube is mounted. The leeward end of the tube connects with some device for measuring the pressure of the air within the tube. This method has been used for a century. Newton's idea of this device is very well known, and wind gauges of this kind have been much used. The air stream may be supposed to flow from a reservoir of air of the same density, and having a height h. If the density be d the pressure in dynes per square cm. which would cause the sup- posed efflux at velocity v is P = gh d. The velocity of efflux is v = V2gh. Eliminating h » P=-v2. (1) 2 v J If v is given in centimeters per second, and o be given in grammes per cubic centimeter, then P will be given in dynes per square centimeter. At ordinary temperatures and pres- sures, d may be taken as 0.0012. If v be in miles per hour and P be in lbs. per square foot, P = 0.0025v2. (2) * Presented to The Academy of Science of St. Louis, December 20, 1897. 2 Trans. Acad. Sci. of St. Louis. An obstacle which wholly checks the wind, would develop this pressure on its windward surface.! This is the pressure collected by an open tube directed by a wind-vane. I have verified this formula by experiments on railroad trains and find it very exact. In pressure-board work, the experimenter determines by means of calibrated springs, the total force required to hold the board against the wind, and divides this force by the area of the board. This gives an average pressure. And here the result is due to a summation of compression on the windward side, and rarefaction on the leeward side. In the tube col- lector, the effect of rarefaction is wholly eliminated. It is not uncommon to hear statements that the want of agreement of different observers in their determination of the constant in eq. (2) is very unsatisfactory. Part of this dis- agreement is due to the fact, that the phenomena were essentially different. Neither the tube collector nor the pressure-board yields any information of value concerning the distribution of wind- pressures over buildings, or other structures. It is this information which the writer has sought to obtain. The pressure of the quiescent air in a building, against the inner surface of its walls, is given by a barometer. In a wind storm, the air masses surging against the building, and sweep- ing around it, compress the air on the windward side, and produce a rarefaction on the leeward side. These differences in pressure are continually changing, as the wiud velocities change, and they are continually dissipating by a flow of air into and out of the building. This flow takes place, not only through windows, doors and crevices, but apparently through heavy brick walls and plastered interiors. I have found by methods that will appear later, that in a tightly closed brick house, the effect of wind is to increase the pressure within the building. This was the case when windows and doors were provided with weather strips, and two stoves were burning on the first floor, and an open grate fire was burning on the second floor where the internal pressure was taken. The air was rushing upward in the chimney leading from the open grate with such velocity as to yield the fluttering sound which t Newton's Theorem. Nipher — Method of Measuring Pressure on a Structure. 3 was almost on the point of breaking into a deep musical note. And still the pressure within the building was increased by every increase in wind velocity. Within the building the barometer measures the air pressure, modified by these wind effects. If however we take the barometer outside of the building, where it is exposed to the wind, it fails to measure the true pressure of the air. The barometer is then an obstacle in the wind. Some of the small openings leading into its air-spaces may collect pressure from compressed air on the windward side, and some from the rarefied air on the leeward side. In addition the air sweeping across the mouths of such openings produces rarefactions within, by an action which is utilized in the atomizer. The barometer will then record the result- ing air-pressure within its air spaces, just as inside the build- ing it measures the air-pressure in the building. But this gives no indication of what the pressure is outside of the barometer. If we shield the barometer from the wind by inclosing it in a box or casing, then we have simply placed it in another house. The air-spaces of the box are then in the uncertain condition which existed before in the barometer. The device which is to be used for measuring the pressure at any point on a building, must consist of a collector, to be placed at the point, and an indicator or gauge, to be placed at any point where convenience demands, and a tube for trans- mitting the pressure from collector to gauge. The collector must be small and portable. It must be un- affected by the wind, but must collect and transmit changes in pressure due to wind. Compressions and rarefactions due to the collector itself, must not be allowed to affect the gauge. The suction or atomizer action of wind blowing across the end of a tube must also be prevented. After many failures the collector shown in Fig. 1 was devised, and seems entirely satisfactory. The tube leading from the pressure gauge, and shown in longitudinal section, pierces a thin circular disk, having a diameter of *2\ inches. The end of the tube is flush with the face of the disk. The disk is ground to a thin knife edge as shown. A second disk of the same size is screwed down upon the former. 4 Trans. Acad. Sci. of St. Louis. The screw head should be soldered or brazed to the second disk, in an air-tight joint. The screw threads are held in a perforated nut brazed into the end of the tube.* Between the disks is clamped a circular, porous layer, which extends half an inch or more outside of the metal disks. The material best suited to use in this porous layer seems to be wire screen. I have tried the finest brass wire cloth, having 130 wires to the lineal inch, and ordinary iron window screen, having 12 wires to the inch. The number of thicknesses of screen has been varied from two to twelve. The results were not appreciably affected by such changes. For use in rain storms the coarse iron wire is to be preferred. Three sheets of wire screen, making angles of 30° with adjoining sheets, give very — *n <£ Fig. 1. satisfactory results. The wire may be dipped into a thin oil, in order to prevent water from clogging the layer. When the fine wire is used, the central part of the layer around the screw should be punched out to form an air chamber. According to the plan which dictated the design of the col- lector, it was to be unaffected by wind, when placed edge- wise in the air current. The air currents between the two disks are so checked that no atomizer action is possible. The compression on the windward edge of the porous layer, is dissipated by lateral outflow before it penetrates between the metal disks. The rarefaction on the leeward edge is similarly prevented from affecting the air between the disks. This is accomplished by a lateral inflow cf air. The arrows in Fig. 1, roughly indicate the action described. This collector was placed under a bell-jar resting on a glass plate. The tube led * Soldered joints will serve unless the collector is to be used in the hot air of a chimney. Nipher — Method of Measuring Pressure on a Structure. 5 out through a stopper, and was taken to the air space in a cistern half full of water. The capacity of the cistern was about 100 cubic centimeters. From the bottom of the cistern, a glass tube one meter long, and having an inclination of 5 in 100, served as a gauge for indicating air pressures in the cistern. A second tube entered the stopper of the bell-jar. The effect of breathing gently into the mouth of this tube, and producing a slight compression of the air in the bell-jar, is instantly seen in a rise in the water column of the gauge. In order to test the collector in a high wind the apparatus was taken upon a passenger car. The water gauge was clamped to a window sill. The doors were opened at each end of the car. The ventilators and windows were also opened, so that there was a free circulation of air through the car. There was no appreciable atomizer action on the gauge. The collector was thrust out of the window, and held edgewise in the wind at a distance of thirty inches from the side of the car. The connection between collector and gauge was then made and broken. No effect was produced on the gauge. This experiment has been repeated at various times, and in one case when the train traveled between two stations fifteen miles apart in fifteen minutes by the watch. When the hand is placed tangent to the windward or leeward edge of the collector, the gauge at once shows a marked decrease or increase in pressure. The wind in free air does not affect the gauge, but any change in pressure is at once in- dicated. It was also found that the collector might be held in any other position in the wind, with no appreciable change in the result.* The wire layer was cut off even with the edge of the metal disks. When placed edgewise in the wind, a marked increase in pressure was then shown. When turned 7-^° around a diameter at right angles to the lines of flow, this increase in pressure disappeared. For greater angles the gauge showed an exhaust. The disk collector thus appeared to satisfy all of the re- quirements, and justified more serious attempts to test its limitations. * This has only been tried with the fine brass wire, but with from two to twelve thicknesses. 6 Trans. Acad. Sci. of St. Louis. From what has been said it is apparent that the open leg of the water gauge should be connected with some standard pressure, and not allowed to communicate with the air in the room. The method suggested by Abbe was employed.* Two thin circular steel plates eighteen inches in diameter were ground to a sharp beveled edge. They were clamped together and held rigidly one-eighth of an inch apart, by four screws near the edge having accurately turned bushings be- tween the plates. A small screw was also used to clamp the plates near the center, in order to prevent vibrations of the plates. A face plate was screwed to the center of one plate into which an iron pipe could enter, and which served as a vertical support for the plates. Inside of the pipe a rubber tube made connection with a small brass tube which pierced the large plate flush with its surface. This device is placed above the building in the undisturbed stream of air, with the plane of the disks horizontal. The tube terminating between the two disks was connected with the inclined leg of the pres- sure gauge. The cistern of the gauge was connected with the collector shown in Fig. 1. This collector was mounted above the building near the other device. When the large disks were as near as one-eighth of an inch it was found that the two collectors balanced against each other per- fectly, in the strongest winds of last winter. The conditions of steady flow of a fluid between two large plates have been discussed by Sir William Thomson. f Abbe states the result as follows: "Steady motion becomes easy and turbulent motion becomes difficult when the distance between the plates is equal to or less than the diameter of the plates, multiplied by the ratio — coefficient of viscosity of the air, divided by coefficient of skin friction of the air on the plate. " The pressure of the air within the tube is then the same as that between the plates, and in the free air around them." When the cistern of the gauge was allowed to open into the air of the room, the variation of pressure within the building, due to gusts of wind, was plainly shown. * Report of the Chief Signal Officer, 1887. 2 : 144. t Phil. Mag. Sept. 1887, and verbally at the Manchester meeting of the British Association. See Report of the Chief Signal Officer before referred to. Nipher — Method of Measuring Pressure on a Structure. 7 With a view of making a further test of the disk collector of Fig. 1, it was decided to make a study of the distribution of wind pressure over a large pressure board, carried above the roof of a railway car. The President and the General Superintendent of the Illinois Central Railroad readily con- sented to co-operate with us, and furnished a car which was changed to adapt it to the work. The shops of the road at East St. Louis and Centralia, Illinois, were practically placed at our disposal, and the apparatus, which had been con- structed in the shop of the Physical Laboratory of Washing- ton University, was sent to the railroad shops, and mounted upon the car. The pressure board was three feet wide in vertical dimen- sion, and four feet long. It was made of dry pine seven- eighths inch thick, the wood having been boiled in oil, in which it was allowed to cool. In this way it was hoped to prevent warping due to sun and rain. The boards were grooved together, and were bound at the top and bottom by cross pieces of wood in the plane of the board. The board was provided with two strap hinges of iron, one-fourth of an inch thick, four inches wide, and crossing the entire board. They covered the horizontal strips marked b and h in Figs. 3 and 4. The hinges were each bolted by twelve f- inch bolts at the center of the four inch squares shown in the figures.* The hinges extended beyond the board and locked to a vertical iron pipe one and five-eighths inch external diameter which passed down through the roof of the car to the floor. The board was mounted upon the pipe somewhat as a flag is mounted upon its staff. The distance from the edge of the board to the center of the pipe was six inches, the clearance between board and pipe being intended to isolate the board from the pipe. The hinges were only an inch wide in vertical dimension where they crossed this interval, but were made broader laterally. The lower edge of the pressure board was one foot above the center of the roof. The pipe turned on ball bearings at the bottom, capable of taking up any thrust in a vertical or a horizontal direction. * The pressure board is not shown to scale in these figures, since the small areas are not squares. 8 Trans. Acad. Sci. of St. Louis. A similar ball bearing surrounded the pipe at the roof. The pipe was clamped in collars near the floor and roof, and ten- sion rods spread out from collar to collar, passing over com- pression members, which braced against the pipe in order to stiffen it in the plane of the pressure board. The vertical pipe was provided with a T into which a hori- zontal pipe was screwed to serve as a lever. A spring-balance was attached to this lever on an arm of 36 inches. The spring balance rested on a horizontal shelf, forming part of a mov- able windlass around which a small rope attached to the ring of the spring balance was wound. From the hook of the balance, a wire was attached to the lever arm. A gradu- ated arc attached to the lever under this wire showed the angle between the normal to the lever, and the direction of the pull indicated by the spring balance. This angle was kept small by moving the windlass, but the angle was read whenever the balance reading was taken. The board turned so easily that its friction was wholly inappreciable. The pres- sure board could be set at any angle with the wind, and the moment of the force required to hold it there was determined. The setting of the board was accomplished by means of a wind vane near the front of the car mounted in ball bearings exactly like those of the pressure board. Pointers a foot in length clamped to the pipes of pressure board and wind vane moved over graduated scales. They were set to 180° on both scales, when vane and board were set in a head wind coincid- ing with the axis of the car. A second movable pointer on the pressure board circle could be set to any desired angle with the former. Bringing this pointer to the scale division indicated by the wind vane pointer, would set the pressure board to the desired angle with the wind. In that part of the work of which account is here given, this angle was 90°. The pressure board was then at right angles to the wind. In a head wind, the vane was very steady, but when the wind blew strongly across the car, there was always more or less of fluctuation, and the pressure board was held in a mean position, and favorable conditions were seized upon as they presented themselves. Nipher — Method of Measuring Pressure on a Structure. 9 The work of setting the pressure board, reading the posi- tion of the wind vane, the index of the spring balance and the angle of pull, was done by Professor A. H. Timmerman, formerly ray assistant, and now Professor of Physics in the Rolla School of Mines. It was necessary to give constant attention to all of the indications to be read, in order that values practically simultaneous could be obtained when an adjustment was secured. The pressure board was divided into 108 four-inch squares. The vertical rows of squares were numbered from 1 to 12, the former being nearest to the axis of rotation. The horizontal rows of squares were lettered from a to i, the former being at the top. At the center of each square a hole f of an inch in diameter was bored through, and two disk collectors could be me- chanically coupled to each other through any one of them so as to close it in an air-tight joint. All other holes were closed by closely fitting corks, flush with the front side of the board. One collector was fur- nished with a screw, which was seated in a screw hole in a hub attached to the other collector, and fitting into the hole in the pressure board. The manner of mounting these collectors is shown in Fig. 2, where the fragment of wood between represents the pressure board. Rubber tubes passed down through the iron pipe into the car and connected the disks on the front and rear of the pressure board with separate pressure gauges. In this way the pressure could be determined at 109 different points on the board, the pressure in front being measured separately from that on the back of the board. The pressure gauges were four in number. The cisterns were Fig. 2. 10 Trans. Acad. Sci. of St. Louis. 2% inches in diameter and 4 inches high. The glass tubes lead- ing out from the bottoms of the cisterns were all mounted in parallel grooves in a board graduated to centimeters, from 0 to 100. The cisterns had a vertical screw adjustment, so that the water levels could all be set to 40 centimeters on the scale. The tubes were all inclined 5 in 100. The support for the tubes was set in a longitudinal position in the center of the car and was provided with a pivotal adjustment iu leveling on varying grades. A spirit level was used in adjusting the level when the car was at rest, but during motion of the train, one of the three gauges was used as a level, its scale reading being maintained at 40. The tube and cistern of this gauge were closed on each other by a rubber tube. The gauges were mounted on a pendulum weighing about 150 pounds, suspended on knife edges from the roof of the car, and swinging transversely. A paddle at the bottom dipped into a trough containing fifty pounds of crude glycerine. This paddle was adjustable as a valve, so that the time of swaying of the pendulum could be adjusted to the rocking of the car. It was, however, necessary to constantly steady and check violent swinging with the hand. By this means, however, a very satisfactory leveling was maintained. See Plate 1. The cistern of the third gauge was connected with an open cup collector attached to the wind vane two and a half feet above the center of the car. The vane consisted of light divergr- ing wings of wood, three feet in vertical dimension and two and a half feet long. The vane was 16^ feet from the pipe supporting the pressure board and the pipes supporting vane and pressure board were five feet two inches from the side of the car. When free in the wind, board and vane always set parallel to each other, and in a calm, they set parallel to the axis of the car when the train was in motion. When the pressure board was set at right angles to the car, its center was in the middle vertical plane of the car. The car was reversed at the end of each run, so that the vane was always in front. The external mounting is shown in Plate 2. The three gauge tubes were separately connected by rubber tubing with an iron tank fifteen inches iu diameter and twenty- Nipher — Method of Measuring Pressure on a Structure. 1 1 eight inches high, which was connected by a large rubber hose with the Abbe collector before described. The latter was exposed above the roof of the car. It was placed symmetri- cally opposite the wind vane, and 38 inches above the roof. At all points where rubber tubing might become sharply bent, a connection of brass pipe was inserted. This pipe was bent into a smooth curve by first filling it with lead. The tubes were all adjusted so that gauge columns would respond with equal promptness, when they were affected by a common pres- sure change. The iron tank containing the common standard pressure against which all pressures were measured, was made large enough, so that a rise to the gauge limit affecting only one gauge would not appreciably affect the others. The resistance of the fine brass wire cloth in the collectors, was found to be equal to that of 20 to 25 feet of the rubber tubing used in connecting. This was determined by a method equivalent to a Wheatstone bridge arrangement. A divided circuit consisting of two equal tubes led from a common source at an air blast to the extremities of a long horizontal glass tube, which drooped in the middle and contained a short column of water as an index. The other two arms of the bridge consisted of the disk collector, and a tube whose length was so chosen that the pressures on the water index were balanced when an air blast was applied. These circuits were "grounded" in the air. The final adjustment on the car was, however, made by trial, by blowing gently for a moment into the tank. This adjustment was effected by means of screw clamps, which pinched the rubber tubing connecting the cisterns and the gauge tubes. See Plate 1. The gauge readings were all made by myself. The instru- ment was kept constantly leveled and steadied, and on a call of " read " from Mr. Timmerman, the three gauges were simultaneously determined. The columns vibrated constantly through two or three centimeters, and it was necessary to be always ready to take an average reading of them all. Two wooden pins were set on the divisions to be read on two of the scales, and the third was read first. The other two were then read later. All of the work was recorded by Mr. Louis Schlossstein, of St. Louis, who also made a simultaneous 12 Tra?is. Acad. Sci. of St. Louis. leading of train speed, as shown on the dial of a Boyer speed recorder. The work was all done on the Illinois Central Railroad between Champaign and Centralia. The car was the second car from the locomotive in the fast freight trains running between Chicago and New Orleans. The start from Cham- paign was made at 5:35 a. m. and Centralia would be reached at 11:50 a. m. Sometimes we made the return trip in the afternoon, starting from Centralia at 2:45 p. m. and reaching Champaign again at 9 :00 p. m. The work dur- ing the latter part of this run was done by the light of caboose lamps. The double run proved too much for our endurance, and we were generally satisfied with the single run. The speeds varied from 20 to 50 miles per hour, as shown by the recorder. The indications of the recorder were checked by watch observations on the mile posts. We found a small index error which could easily have been corrected by an adjustment of the driving pulley of the instrument, but we preferred to make no change. The experiments were begun on June 26, 1897, and were continued every day until July 20, although part of this time was devoted to work not within the scope of this paper. About 1,500 independent measurements were made upon the pressure board. It was decided to make a very thorough determination of pressures along the middle lines of the board. Such observations were made along the horizontal line of squares from 1 to 12, e, Figs. 3 and 4, and along the two vertical rows a to i, 6 and 7. In addition the half of the board furthest from the axis was well explored, and observa- tions were made at a few symmetrically located points in the other half of the board in order to detect any substantial dif- ference which might exist. It was to have been expected that slight flexures of the board might result in some differences in the lateral halves, although no appreciable difference was found. It was however found that the dragging of air along with the train caused the pressures on the front side to be greater at the top than at the bottom of the board. This effect was least when strong winds blew across the trains. It was also found that the rarefaction was greater near the Nipher — Method of Measuring Pressure on a Structure. 13 bottom than near the top on the back side of the board. This was doubtless due to the obstructing effect of the car roof. Fifteen observations of pressure in any square were usually taken at one time and the collectors were then removed to another. In the tables which follow the means of these observations are given. Where several determinations are given for the same square, they were usually made on differ- ent days. Some discrepancies appear which seem too large, but all observations have been included. The tremendous shocks which our improvised laboratory sometimes received made it necessary to exercise constant vigilance in detecting loose adjustments, and some sources of error have doubtless escaped us. It was found that increase of pressure on the front side of the board, and decrease of pressure on the back side were linear functions of the total force required to hold the board to the wind, as measured by the spring balance. If .F repre- sent the pull of the spring balance on an arm of 3 feet, h1 the increase in the scale reading of gauge No. 3 above the datum reading, 40.0 and h2 the decrease of gauge reading of No. 4, below 40.0, then for the front and rear pressures we have respectively, ft, = AYF h2 = A2F where A1 and A2 are constants A2 being essentially negative. They denote the increase or decrease of scale reading per pound of pull on the spring balance. Ax and A2 are reduced to vertical water column by multiplying by 0.05. This may be taken as the pressure in grammes per square centimeter. The correction for density of water is about half of one per cent., and is slightly overcompensated by the change of level in the cistern. The further factor 2.048 reduces the pres- sures to lbs. per square foot. The factor for reducing Ax and A2 to lbs. per square foot is therefore 0.1024. The values of Av and A2 have been entered in the proper squares on diagrams of the front and the back of the pressure board. Such diagrams are shown in Figs. 3 and 4. The board was divided into strips where the conditions were evi- 14 Trans. Acad. Sci. of St. Louis. © OS os os os OS OS OS © o OS OS OS OS OS OS OS co • ■**! CO CO CO CO CO CO CO o O ■* •»*< -* -* -# o co eo OS I— . »— < »— 1 1— « »— 1 o (M •«*< 1Q CO CO CD CO CO CO o o o »a «5 »o o 1" CO o CM >a lO U3 CO o 00 • • ■ • • ■ • CN ■* CO CO co co co co co CO o ■* co co CO CO »^ o oo CO OS OS OS OS OS co o CM ■<* ia CO co CO CO t~ CC CO o -# CO CO l~ 00 b- o 00' CO OS o -r 1-1 -# co <— CM «♦ lO co c~ l^ t~ t- co CO o «# 00 t^ CO t^ o 00 eo OS OS T 1-1 ■* CO o CM «* lO CO t~ t~ t- t^ CO co o «# cc b- 00 t~ o CO CO -- OS ■**■ 1-1 "■* CO o CM "* us CO t>. t~ t~ l~ cc CO O «# CO b- 00 t— o oo CO os OS •*< i— i -* CO © CM • • • • • • • • >* >c CO t- t^ t^. t> CO O -* O CO o f~ o 00 eo OS OS OS OS c CO © . t~ co CO o o o cc CO 00 o o 00 CO OS »o "* ■>*< ■* U5 © CM • ■ • • • • • • • «* 1Q CO t-» t- t-- co IQ co o o ■** «* •>* •* -*< © 00 co o OS OS OS OS OS CM IM "* >o »o id ia 1Q «5 -* co o o o o O o O © © o CO CO CO CO CO CO CO © "* -*> ■* ■* «*■ ■* -* -* co ft K •< O « D co cc K - Pi H X H b O H © '- © — P=4 ■ts Nipher — Method of Measuring Pressure on a Structure. 15 o OS OS OS OS Ol 01 Ol © o CI OS OS OS 01 Ol Ol ta CO CO CO CO eo eo CO eo CO o o OS OS f— i f—t © © © t- o OS 01 co CO co HI o co ■* CO eo ■* •* ■* •>*< -* 00 o © ^ © © © o >o o 1— 1 CO co 00 00 t- CO CO •& -"*! ■* ■* -* ■■* ■«*< >* CO © CD © .-I 1— 1 CN © © >o o CN CO 00 CO © t- eo CO •* ■>* «* «* •* IQ ■"* -* CO o © © N © © 10 o CM eo CO CO © l>- eo • • • • • • • ■ • CO ■<*< ■* -* >* ■<»< ta ■* •* 00 o © © .— < * •«* •>* "* to ■* •«* 00 «£> © © »— 1 r-t eo © © ta © N eo co CO © fc» eo CO <* -*• •* «* ■* 1C «* -* CO o © © 1— I CM © O «5 o IN eo CO co © t-- eo CO <# «* ■* -* •* »o ->*< «* CO © © © »— 1 t— 1 N © © IQ o CM •>*< co CO © t~ eo CO ■>* ■«*! ■* -*1 ■"*" lO ->♦< ->*< 00 o © *~t t— 1 09 © © © © U5 o t- 01 t~ © b- eo • • • • • • • • a eo ■* -* ■"*< -* ■* HI ■* -* 00 © »— 1 f— i l-> © © © © IQ ■* co to co co CO »o CO • • • • • • • • i eo ■«*< "* "* •* -*i ■mM ■*! -* © © © © © © © © © o eo eo CO eo eo eo CO IN • eo ■* -* ■* >* 1 ■* •* -*l e* 0 < O fa « P x CD W K Pm H w H fa O « V t» 16 Trans. Acad. Sci. of St. Louis. dently similar, and the average of observed values made in this area was determined. Thus, in the row of squares adjacent to the edge, the top line of squares constitutes such an area. These squares are marked 1 to 12, a. The line of squares 1 to 12, i, is another such area. The averages in the vertical rows a to i, 7, and a to i, 12 were also separately found. The average value for such areas was then entered in all the squares of the areas, excepting that the values in the cor- ner squares were smoothed a little, in order to make adjoin- ing side rows unite with each other. Over these corner squares the pressures diminish in two directions toward the nearest edges of the board. In such cases the integrated pressures must therefore be less than in squares removed from the corners. In like manner the averages in the second row of squares from the edges were determined. In this row the squares 2 to 11, b; 2 to 11, h; b to h, 2 ; and b to h, 11 were sepa- rately treated. The data contained in the table may be used by anyone to repeat these computations. The results for each square on the board are entered in Figs. 3 and 4, where the values have all been multiplied by 100. The values in Fig. 4 are of course negative in sign. These values therefore correspond to a spring balance reading of 100 lbs., when the arm is three feet. The moments of the forces applied to these four-inch squares with respect to the axis of rotation have been summed. It was assumed as sufficiently exact, that the center of pressure for each square of -^ square foot, was at its center. The arm for the vertical row a to i, 1, is f ft., the arm for vertical row a to i, 2, is one foot, etc. The sum of the moments for the front and back sides of the board was found to be, — Front 1 fi9 Back 128 Total 297 By the spring balance 300 Difference 3 A difference of one per cent. Nipher — Method of Measuring Pressure on a Structure. 17 Square. Ai A2 Square. A, A2 1 a 0-36 —0.41 7(f 0.85 —0.37 1 e 0.33 —0.46 7 i 0.23 —0.48 II 0.45 —0.36 ti 0.36 —0.40 (1 0.50 —0.38 8 e 0.73 —0.36 1 i 0.44 —0.48 it 0.67 —0.44 2 e 0.56 —0.47 9 a 0.43 —0.44 < Hutton u Thibault <( Dubuat ( Moria ) \ Piobert & \ { Didiou J << Mariotte Dubuat Thibault Langley Air (i c« (« « (i it Water Air Water cC II Air ii 0.13 0.25 0.63 0.13 0.25 0.25 1.11 1.0 0.3 to 2.7 Rotation. ii ii ii ii ii ii f Still Water rect. v = 3 to \ 6i ft. per sec. 2.18 1.25 / Vertical motion shallow \ tank . 1-856 ( Plate stationary, stream \ of water. Wind Power. /Rotation u = 4 to 11 me- \ ters per sec. 1.834 1.31 1.17 to 2.5 1 The pressures in Fig. 3 correspond to a balance pull of 100 lbs. on an arm of 3 ft. This is equivalent to a force of (3 -4- 2.5) 100 = 120 lbs. at the center of pressure. The pull per square foot is therefore 10 lbs. This gives a value for E of 10 -f- 7.17 = 1.39. This value is only in part de- termined from data independent of that involved in the other determination. If instead of taking the pressure at the middle of the pres- sure board we take the average pressure over the front face, we have by the spring balance determination of pull, K = 10 -4- 5.58 = 1.79, which agrees very well with some of the higher values of K determined by others. It is however evi- dent that this value of K is theoretically a wholly different quantity from that which we seek to determine. It seems entirely possible that the pressure at the center of the board is less than that which corresponds to the average (velocity)2 with which the wind sweeps past the board, as Nipher — Method of Measuring Pressure on a Structure. 21 computed by the Newton theorem. The determination of this wind velocity is a problem of much greater difficulty than at first sight may appear. If we assume 7.48, the highest pressure on the board to correspond to this velocity, the value of K becomes 10 -f- 7.48 = 1.34. The spring balance and pressure collector deal with the pressures which really affect the pressure board, regardless of the complexity of air currents. In this respect they differ essentially from the tube collector attached to the wind vane. In illustration of this point I may relate one incident of our work which happened after our pressure board work was finished. It was observed on starting on one of our trips, that the gauge attached to the cup collector on the wind vane showed little if any increase in pressure. As the speed in- creased, it began to show an exhaust, which seemed to increase with the speed. Every part of the pressure circuit was examined and no cause could be found for the result. Finally it was observed that unnoticed by us, a refrigerator car had been placed in front of us, and it was thought possible that the open trap-door in the roof of this car might be the cause of the trouble. The door was 3,3X35 inches, and was opened to an angle of 33°. Its distance from the wind vane was eleven feet. The door was shut down, and the gauge read- ing increased 20 centimeters in a couple of seconds. This door had deflected the air stream upwards, and it had de- scended in a cascade upon our car. The lines of flow must have made a somewhat greater angle than 60° with the horizontal axis of the tube collector. It is because of such experience that no use has been made of a great mass of data involving velocities. This material will be used in a future paper, deal- ing with the effects of the train upon the air around it. The results of this investigation seem to fully justify the use of the disk collector in a study of the distribution of pressure on large buildings. An arrangement which I have found very satisfactory may be briefly described. A window is raised a couple of inches, and a wooden bar capable of longitudinal extension is placed in the opening thus formed. A f- inch brass tube connecting with the pressure gauge passes out through a hole in the bar. Outside, this tube enters a 22 Trans. Acad. Sci. of St. Louis. similar tube at right angles to the former, and forming the arms of a T. From the ends of these arms short tubes return to the wall of the building, and serve as feet for hold- ing the conducting tube in position, when it is pulled into position from within. The ends of the feet should be plugged, and one of them should be adjustable, to provide for inequal- ities in the wall surface. The end of one arm of the T should also be closed. From the other, a tube extends six or eight inches, and an elbow and tube return towards the building and carry the pressure collector. The plane of the collector is parallel to the surface of the wall, and should be one or two inches from it. Collectors should be placed over the various walls of the building and upon its roof. The tubes should be adjusted to the same resistance, and as nearly as possible, to the same capacity, so that all will respond with equal quickness. The gauge tubes should be parallel and side by side, so that they may be photographed at any instant, when a heavy gust of wind is blowing. The gauge tubes should of course lead into a tank, connecting by a large tube with a large Abbe collector above the roof. This collector should be high enough above the roof to escape the effect of the building. The stream lines should be horizontal where this collector is placed. Near the summit of inclined roofs, ou the windward side, and over the entire leeward side, the pressure is reduced by a wind. On a, flat roof the pressure is always reduced. The pressure is less than that of the atmosphere, and this reduction of pressure increases with an increase in velocity. The standard pressure should be col- lected from a point above these disturbances. When the wind blows at right angles to one side of a build- ing, the pressure on all other sides is diminished. If a win- dow yields on the windward side the pressure within the house will be increased. The tendency is to lift off the roof, and throw three walls outwards. The windward wall is often braced by partitions so that it is less likely to fall in than the others are to fall out. I believe that it is possible to greatly increase the resisting power of ordinary farm build- ings against wind without materially increasing their cost. Nipher — Method of Measuring Pressure on a Structure. 23 But it is very desirable to determine what these pressures really are. The observations on pressures should be supplemented with simultaneous observations on wind directions. A vane with a cup collector connected with a gauge below will determine pressure, due to the free wind, from which velocity may be computed. Two metal brushes 180° apart are attached to two insulated rings surrounding the vane-tube. The poles of a battery connect with these rings by sliding contact. A flat commu- tator with 72 segments surrounds the tube and the two brushes slide upon it, touching opposite segments. The wires lead in a cable to the closed copper windings on an iron ring at the pressure gauges. The wires of the cable connect to these windings at equidistant points, those coming from opposite commutator bars leading to opposite points in the winding, and adjacent wires at the commutator being adjacent at the ring. The arrangement is exactly that of the Gramme armature. The polarity of the ring will follow the shifting of the vane. A magnetic needle whose pivot is at the center of the ring will follow the vane. This needle may be photo- graphed with the scales. With comparatively small expense it will be easy to obtain information of great value, even with ordinary winds. At some of the mountain observatories, destructive winds are not uncommon, and it is desirable that such work should be undertaken there. In conclusion I wish to express my grateful thanks to Professor Timmerman and Mr. Schlossstein for their enthusiastic aid under circumstances which were very far from inviting to one seeking rest and quiet. My thanks are also due to Mr. Stuyvesant Fish and Mr. A. W. Sullivan, President and General Superintendent of the Illinois Central Railroad, through whose co-operation we were enabled to do the work, and to Mr. Joseph Boyer for the loan of his Speed Recorder. There are dozens of others who have assisted us greatly in various ways, whom I cannot name. The names of some of them I do not even know. But it was a constant pleasure to receive their aid, and I feel that much of the success we may have attained was due to such friendly assistance. 24 Trans. Acad. Sci. of St. Louis. On p. 13 reference is made to an uncorrected error in dis- placement of the zero point. This is a differential error, due to differences in the gauge tubes. The correction is practi- cally all involved in the factor 0.05. Issued January 14, 1898. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate I. ^^^fy^/fnfif m ■ PENDULUM AND PRESSURE GAUGES. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate II. c H C K "O ?5 P3 CO ts O Transactions of The Academy of Science of St. Louis. VOL.. VIII. No. 2. THE TREES OF ST. LOUIS AS INFLUENCED BY THE TORNADO OF 1896. HERMANN von SCHRENK. Issued January 29, 1898. THE TREES OF ST. LOUIS AS INFLUENCED BY THE TORNADO OF 1896.* Hermann von Schrenk. The tornado, which swept over the southern part of St. Louis on May 27th, 1896, was one particularly destructive to the trees and vegetation of that section, and in the following an attempt has been made to enumerate some of the phe- nomena which followed the period of destruction. On the morning after the storm many of the immediate injuries to trees were noted. Most of the observations were made in the storm center, which included Lafayette Park and the adjoin- ing streets, and observations which follow are to be taken aa applying to this region. Nortonf has described many of the injuries which the trees suffered, and to these I would add several, necessary to an understanding of the after-effects. There was hardly a tree which escaped injury of one kind or another, with the possible exception of several cypresses, Taxodium distichum, which, with their conical forms, yielded to the force of the wind. The maples and elms, as stated by Norton, were in many cases uprooted, and when this took place in the first few moments of the period of greatest wind velocity, the trees were simply turned over, and when straight- ened up some days after, resumed their former growth, and to-day these trees are almost the only normal trees in the Lafayette Park district. By far the larger number lost all their principal branches ; many were reduced to the trunk with perhaps two forks. The ragged ends of such trunks were sawed off, painted and covered with cloth, and then they looked much like very heavily pollarded trees. The time at which the injury was done, fell in the period of the tree's greatest activity. The new leaves had not been * Presented to The Academy of Science of St. Louis, in preliminary form, December 6, 1897. t Norton, J. B. S. Garden & Forest 10: 292. 1897. (25) 26 Trans. Acad. Sci. of St. Louis. unfolded for many weeks and many were still unfolding. In accounts of the injuries sustained by trees in former storms, little is said of leaves and leaf injuries. The manner in which leaves are attached to stems enables them to present the smallest surface to the wind, and thus to escape injury. Peculiarities in the structure of leaves have been adduced as another protecting factor.* When moist, leaves are far more liable to injury than when dry; several leaves cling together, and the force of the wind pounds and rubs them against twigs and branches, tearing and lacerating them.f In the present case the destruction of foliage was very marked, particularly in such trees as the soft maples and sycamores ; several trees of the honey locust retained most of their leaves. For a large area about the trees, some minutes after the storm, the ground was covered with torn and shredded leaves of all kinds. The leaves were wet, for it had been raining for sev- eral minutes before the windstorm, and the injuries evidently were largely due to rubbing against branches. Flying missiles •of various kinds aided in the destruction. Grains of sand and small bits of wood and stone, flying through the air at velocities ranging from 50 to 80 miles per hour, were well able to shred the tender leaves. As a result of these combined causes many maples and sycamores were left with hardly a sound leaf on the remaining branches. Besides the injuries to the branches and leaves there were others not so evident at first. Numerous trees had trunks of sufficient elasticity to bend before the force of the wind, with- out breaking. In swaying to and fro, the bark on one side was considerably stretched, and compressed on the opposite side, and in the next instant the conditions were reversed. Where this took place repeatedly the bark was torn horizon- tally for several feet, sometimes on but one side, more often on both. The violent wrenching of a tree with a large top * Kuy, L. Uber die Anpassung der Laubblatter an die mechanischen Wirkungen des Regens und Hagels. Berichte d. deutsch. Bot. Gesellschaft 3:207. 1885. t Magnus, P. Verhandl. d. Botauischen Ver. d. Prov. Brandenburg 18: viii. 1876. — Caspary, R. Beschadigung der Rosskastanieublatter durch Reibung mittelst Wind. Botanische Zeitung 27: 201. 1809. von SchrenJc — Trees of St. Louis as Influenced by Tornado. 27 like the maple, produced considerable strain upon the bark, especially when the force applied was a twisting one. Where the strain was too great, the bark came off in sheets, or split longitudinally. This was particularly noticeable in narrow streets like Nicholson Place. In many trees there was no outward sign of this injury for several months; not until the loosened bark died, did any shrinkage take place, but then it split and curled up. This could be seen on many maples in September, 1896. Flying pieces of slate and stone removed large pieces of bark, and often became imbedded in the wood. Wounds of this nature healed rapidly, and many pieces of slate are to-day surrounded by new wood. Immediately after the storm attempts were made to save as many trees as possible. The fallen ones were righted, the torn stumps were cut off and covered, and those trees which had suffered material injuries to their bark, were covered with cloth to keep out insects and fungi as far as possible. The weeks succeeding the storm were warm, and an abundant amount of rain fell. For some weeks in June the trees ap- peared dormant, then gradually, on such twigs of that year as were left, the axillary buds for the year 1897 began to unfold, and produced new leaves, a process much akin to that known as " Johannisgrowth." By September, a growth of six inches or more had been made from those buds. But by far the majority of the trees had no such buds to fall back upon, as all small branches had been torn away. In these trees, numerous adventitious buds broke out from all parts of the trunk and remaining larger brauches. In the maples these grew less vigorously than in the elms or sycamores, but in all a round mass of foliage grew during the summer. As the injuries to many of the trees had been very great, this growth was very small and such trees failed to revive this spring (1897). This was true of all the lindens (Tilia Americana} . On the morning following the storm some notes and draw ino-s were made of various twigs, which were marked, with a view to adding something to the discussion of the question whether a tree can form more than one ring of wood in a year under unusual circumstances. The trees noted had lost 28 Trans. Acad. Set. of St. Louis. all their leaves, and in every one new leaves were formed from the buds on the twigs of that year. In the spring of 1897, all the trees which had survived the winter, continued to grow normally from the buds formed on the adventitious shoots. Numerous sections of branches were made early in May to determine the character of the wood formed by the trees during the previous summer. Such sections were made again in the fall (1897) of wood taken from trees which bad died during the summer, as well as from living trees. The contention that more than one ring of wood may be formed in one year is one of long stauding, and much has been written both for and against it. Cotta* found that where trees had been injured by insects or frost, such trees made little or no growth subsequent to the injury, and where they did, the new wood did not differ from that previously formed. On the other hand, a very dry period following a spring very favorable to rapid growth, caused a stoppage in the activity of the cambium, which then formed fall wood. A good rain started new activity, spring wood resulted, and thus two rings were visible. Sometimes the two rings could be seen on one side only. Unger,f in a discussion as to whether the development of several generations of buds in any way influences the growth of annual rings, concludes that there is no causal relation, both being- due to changing: con- ditions of temperature and moisture. He found but one or two species where there were any signs of double rings. A full discussion of the earlier views is presented by Kny.f He had occasion to study this problem near Berlin on some trees defoliated by caterpillars. In several trees of Tilia parvi flora the leaves were destined early in the year, and in July the buds formed for the following year developed. Kny was able to determine positively the presence of two distinct rings of wood as a result of premature development of these * Cotta, H. Naturbeobachtungen uber die Bewegung u. Funktion des Saftes etc. 75. Weimar 1806. t Unger, F. Uber den Grund der Bildung der Jahreslagen dicotyler Holzpflanzen. Bot. Ztg. 5: 2C5. 1847. X Kny, L. Uber die Verdoppelung des Jahresringes. Verhandl. d. Bot. Ver. d. Prov. Brandenburg 21 : 185. pi. 1. 1879. von Schrenk — Trees of St. Louis as Influenced by Tornado. 29 buds. He determined furthermore that the presence of two rings was evident for several years back, and gradually became obliterated, varying in different branches. The phenomenon was a variable one in other respects, for the two rings did not always occur even in branches from which all the leaves had been removed. He proved the presence of two rings in Sor- bus aucuparia and Quercus pedunculata . Fagus silvatica var. pendula, which had lost all its leaves, and where a second growth took place, did not form a second ring. Kny suggests at the end of his account, that to prove the relationship be- tween defoliation and wood formation a number of trees ought to be entirely defoliated, and the resulting changes studied. Acting upon this suggestion, Wilhelm* removed the leaves from some trees of Quercus sessilifiora on June 7, and from several others on July 9. Incisions were made in the trunks to study the wood formation. The same was done for a series of control trees of similar age and thickness. In the fall the trees were cut down and examined. A double ring was found, but only on the side of the tree on which the incisions had been made, both in the defoliated and control trees. The youngest ring was characterized by lesser width, and a sudden decrease in the size of the vessels in the center of the ring. The differences were not great enough to warrant the conclusion that a distinct double ring had been formed. He concludes by saying: " The appearance is accordingly a result of injury to the woody cylinder, and is much furthered bv defoliation." He found the abnormal growth of wood in the trunk, but no sign of it in the branches, an observation directly opposed to that of Kny. Those trees defoliated in July showed no interruption in the woody ring, although all had produced a second growth of leaves. The formation of double rings of wood in Pinus sp., owing to late spring frosts, has been shown by Hartigt and others. Hartig found branches which had been frozen repeatedly for * Wilhelm, K. Die Verdoppelung des Jahresringes. Ber. d. d. bot. Gesell. 1: 216. 1883. t Hartig, R. Doppelringein Folge van Spiitfrost. Forstlich naturwiss. 2;eitschrift4: 1. 1895. 30 Trans. Acad. Sci. of St. Louis. a number of years; in one instance a branch fifteen years old had no less than ten double rings. An icemantle is formed between the bark and new wood, crushing the latter. When this thaws it remains in its crushed condition and a break is thus formed which is partially rilled by the expanding cells of the medullary rays, and by parenchymatous cells, developed from the cambium. Ultimately normal tracheids are formed. In the course of time the masses of wood are separated by a fine brown line. It will be noted that this manner of separating wood rings has nothing in common with that seen in normal wood, where there is a distinct formation of different kinds of cells in the fall and the spring. On examining the trees in the storm district on May 28, very few trees were found which had twigs of that year, for in the majority of trees, these twigs had been torn away. Thus it came that the observations had to be re- stricted to the silver maples and sycamores. Fig. 1 is from a sketch of a sycamore twig made after the storm, to which has been added a sketch of the same twig in the spring of 1897. At o is the growth of 1895. It had two lateral buds which developed in April, 1896, into two branches. The storm tore these away leaving but the lowest part of each, with one axillary bud. In June these buds grew out into short branches (6) which bore normal buds in the fall. At c and c are seen the broken ends of the branches. In April, 1897, the buds of the previous fall grew normally, and in May a short branch was at each axil. A cessation of growth had then existed for about a month from Ma}' to June, 1896, and it re- mained to be seen whether this had caused any marked Fig. i. von Schrenk — Trees of St. Loxiis as Influenced by Tornado. 31 changes in the wood. It is very unreliable, as Kny has shown, to judge of the age of a twig by its external markings, i. e., leaf and bud scale scars, branching, etc., and in deter- mining the age only a personal knowledge of the twig is to be considered. The confusion prevailing on the day after the storm prevented the drawing of many twigs, hence the re- sults are not as full as might be desired. Examinations of numerous twigs this spring (1897) showed appearances similar to those in the twig figured and may be considered as con- firmatory in many respects. Sections were made of the twig described, and these are shown diagrammatically in fig. 1, the dotted lines indicating the point at which the section was made. It was very evi- dent that two rings had been formed during the year 1896. The diagrams of the sections need no explanation. A closer examination of the wood of that year was made. The spring wood of the sycamore consists of very large angular vessels with few wood elements. The latter increase in number toward the fall wood, and at the end of the season's growth are much flattened radially (PI. V. fig. 1). The wood formed be- fore the storm was normal spring wood. This rather abruptly passed over into wood (PI. V. fig. 2, b) looking much like normal fall wood. When the twig began its second growth in June a second spring wood appeared (PI. V. fig. 2, a) which was very distinctly separable from the wood found a month before, thus presenting in all respects the appearance ordinarily seen in normal wood. In fact, the two figures (PI. V.) are almost identical. On PI. III. the upper figure represents the cross-section of a sycamore twig taken from am uninjured tree. The inner ring is that formed in 1895, the next one in 1896, the lower figure is from a section of the twig already described. The inner ring is that for 1895 ; the narrow one following was formed before the storm in 1896; the wood beyond, after the storm. PI. IV. shows similar conditions existing in a twig of a soft maple. In the latter the double ring was in all respects similar to that found in the sycamore. The various twigs were studied with a view to determine how far back the abnormal ring could be traced. In the 32 Trans. Acad. Sci. of St. Louis. sycamore noted, undoubted traces were found going back for several years, but there was no evidence of the same in the larger branches. The wood taken from the trunks of some thirty maples and sycamores, examined this fall, failed to show any sign of a double ring. These observations agree closely with those made by Kny, who found the double rings in the branches but not in the trunk. If leaf activity and cambial activity stand in any causal relation, it would appear that this is more marked in the younger internodes, where a sudden stoppage in the supply of assimilatory products would be more directly felt than in the trunk. When the leaves stop their activity suddenly, growth in the younger internodes gradually comes to a standstill, to be resumed with renewed leaf activity. It has often been noted, that in those cases, where axillary buds for some reason develop into branches in the same year in which the branch itself develops upon which they are situated, the wood of the parent branch shows no differentiation, probably because the growth was is no way interrupted. From the observations made on the trees of the storm district, it would appear, that interrupted leaf activity on a branch is felt in the younger internodes, causing the formation of an apparent fall wood, which is followed by spring wood when new leaves are formed on that branch, thus giving rise to a double ring. The distance back where this cessation may make itself felt is very variable, depending much upon the position of a branch upon a tree and upon the num- ber of leaves borne by such a branch. The study of the wood of the trunk had to be necessarily rather conjectural, for there were absolutely no means of determining positively to what year a certain ring belonged. Under conditions of unfavorable nutrition or stunting, trees such as pines, spruces, etc., form no ring whatever.* With the trees of the storm district the conditions surely were unfavorable enough to warrant the assumption of an entire cessation of growth in the months succeeding the storm. With this fact in mind, it is evident, that the last ring formed * Hartig, R. Zur Lehre vom Dickenwachsthum der Waklbaume. Botan- ischeZtg. 28: 508. 1870. von Schrenk — Trees of St. Louis as Influenced by Tornado. 33 might be either that belonging to 1896 or 1897, and where the outer ring was of peculiar structure, following a normal ring, it was, as a rule, considered as being that formed after the storm in 1896. The wood of the trunk was affected by the loss of leaves and branches in a manner which was to have been expected. A very much decreased rate of growth was very evident in most cases, irrespective of the species. Some wood had been formed before the storm and in many trees it seemed as if this had been the last, for the outermost ring was very narrow and followed a ring of the usual average width. In other trees two narrow rings were found, outside a wider one, each, perhaps, one-tenth to one-twentieth the width of the fore- going. An excellent opportunity wa3 given for an examina- tion of the wood of the trunk, as most of the injured trees had died during the past summer and were removed this fall. In eertain trees, such as the elms, where the rings were not so markedty narrower, the cells of the fall wood were almost parenchymatous. With the loss of the leaves the entire physiological activity of the tree was deranged. A powerful effort was made by the tree immediately after the storm to produce new leaves, to enable it to carry on the necessary activities. With the loss of the leaves the transpiration current gradually lessened and finally stopped ; the flow of assimilatory compounds ceased and to all intents the plant stopped growing. The root system was practically uninjured, and, in the warm weather of the weeks succeeding the storm, in a ground abun- dantly supplied with water, the activity of the roots must have been a very great one. The rainfall for these months was as follows: May, 9.12 in., June, 4.57 in., July, 4.67 in., August, 2.12 in., — as against a normal rainfall of: May, 4.58 in., June, 5.08 in., July, 3.76 in., August, 3.50 in. The average mean temperature was, June, 74° F., July, 78.9° F., August 79.4° F., with an average maximum temperature of: June, 82.6° F., July, 87.6° F., August, 88.7° F. These data show condi- tions very favorable to great root activity. Great volumes of water must have been driven into the trunk, more particu- larly into the new wood. This stimulated the growth of ad- 34 Trans. Acad. Sci. of St. Louis. ventitious buds, which appeared at all points on the trunk and branches. But, with an entire absence of transpiration (or almost so, for the evaporation from the branches and trunk must have been very slight in an atmosphere with high humidity) — this high root activity must soon have stop- ped, owing to the pressure of the water already absorbed. When the new buds developed into leaves, the number of such leaves formed was entirely incomparable to the number which had been on the tree before, and consequently the surface from which evaporation might have taken place was corre- spondingly small. This would set up, at best, but a very feeble current, which could in no way balance the volume of water absorbed by an uninjured root system. Under such conditions, a high turgor might be expected to have existed in the new wood and cambium cells in the trunk and branches. After the storm the available supply of assimilatory products could not have been very great, and with little material and a high internal tension, it would seem that cells of unusual form, if not unusual size also, might have been formed by the cam- bium. In other words the conditions seemed favorable for the development of an oedematous wood. That such did actually happen was plainly evident in a number of trees. These were such as were deprived of most of their branches, and on which but few buds developed. A large maple tree in the eastern part of Lafayette Park may serve as an example. This tree died during the summer of 1896 presumably, for it failed to grow any in 1897 and was cut down in June, 1897. The normal wood of the maple consists of wood fibers of small lumen, arranged in exact radial rows, with an occasional large vessel. (PI. VI. fig. 1.) The fall wood is marked by several rows of cells, flattened radially. In this maple the outer ring to the unaided eye looked watery and very pulpy. A portion of the bark had been removed from the eastern side by flying mis- siles during the storm. In June a small crown of leaves was formed but it was evident in September that there was little hope of saving the tree. Examined microscopically this pulpy outer ring appeared like the fig. 2, PI. VI., which shows the inner part of this ring together with the fall wood of the pre- ceding year. The appearances might be explained as follows. von Schrenk — Trees of St. Louis as Influenced by Tornado. 35 In April, 1896, the normal radial rows of cells were formed until May 27th. They were thin walled and full of sap. After the storm the cambium continued to form cells, but now under the conditions above described. The new cells, highly turgescent, pressed on the cells already formed, and flattened them in many cases (fig. 2, b), into disc-shaped cells. As the cambium cells continued in their growth, with abundant water passing into the new wood, the resulting cells were very much enlarged and much distorted, bearing in no way any resemblance to normal wood. Numerous large spaces or rifts were formed, probably owing to unequal pressure (PI. VI. fig. 2, r). This oedematous wood was formed all around the trunk, and was alike at all points, with a possible variation in thickness on different sides. Among the many maples examined there were but few which showed this condition as markedly as this one. This was a disappointment, as it was hoped that the death of so many trees might in some way be connected with this oedematous wood. Its absence in many trees may be ascribed to their greater vigor, and lesser interruption in the transpiration cur- rent. The only other trees with marked oedematous wood were several willows, and in every one of these the malady was very marked. These willows grew near a pond and were close to each other. Two of them had but a trunk and two forks, the others had more branches. It was a surprise to find these willows succumbing to this trouble, and in so marked a manner. The wood of the willow is soft and is composed of very large vessels found in all parts of the wood ring. On PI. VII. fig. 1, such wood is shown, that is to say, a small piece of fall wood followed by spring wood. Fig. 2 is a portion of the wood from the outermost ring of one of these willows. The cells are thin-walled, of very irregular shape, and much compressed in places. The large vessels are in all cases divided into several parenchymatous cells, which in noway resemble the normal vessels (fig. 2,1). A comparison of figs. 1 and 2 will make this clear. Unfor- tunately these oedematous trees all died during the year, so no opportunity was given to see what would have been the character of the wood, after more normal conditions had set 36 Trans. Acad. Sci. of St. Louis. in. It is very probable that some trees still standing are thus affected, and in the years to come, when any "cyclone" trees are felled, one ought to look for traces of the oedematous wood. Many of the injured trees struggled along during the summer of 189(5, produced leaves in the spring of 1897, and there seemed to be some chance that they might continue to live. As the summer came on, however, another factor entered into their environment which proved fatal to most of them. The trees had a mass of foliage developed in the spring which looked healthy and normal in every respect. After several months of hot and dry weather the bark began to break off in large sheets both from the trunks and branches, particularly from the soft maples. The accompanying plate, (VIII.) reproduced from photographs, shows several maples, an elm and a honey locust from which the bark is breaking away. It will be noted that these trees were not then dead, as most of them had green leaves still on them (figs. 1-4). I attribute this breaking away of the bark to powerful insola- tion, causing what is known as bark scorching. Hartig* defines the phenomenon of bark scorching (Rindenbrand) as the drying and subsequent dying of the bark of smooth- stemmed trees, which have been suddenly exposed to the sun's rays. He says : " It has been determined that death is a result of intense heating brought about by direct insolation in midsummer." In another place | he discusses the conditions existing in some spruce woods, in which the trees had been de- foliated by the nun moth. " On Aug. 19th, the warmest day of 1892, when the thermometer registered at 96.8° F. in the shade and 104.9° F. on a felled area that was not exposed to the wind, it was found that on the southwest side of eighty-year- old spruces, fully exposed to the sun, the temperature was 131° F. between the wood and the bark." The subject of the temperature inside a tree, as compared with that of the air, has been frequently considered. Some observers, such as * Hartig, R. Uber Sonnenbrand oder die Sonnemisse der Waldbiiume. Uuter3 a. d. forstbot. Inst. z. Muuchen. 1880. f Hartig, R. Diseases of Trees (translated by Somerville & Ward) 294-95. 1894. von Schrenk — Trees of St. Louis as Influenced by Tornado. 37 Bohm and Breitenlohner,* Becquerel,t Krutzsch,! find the internal temperature of a healthy tree normally lower than that of the air, but later researches have shown that, as a rule, the temperature of the layers under the bark is higher by several degrees than the surrounding air. Rameaux § finds that in a poplar branch, 4 cm. in diameter, the internal tem- perature was 4°, 8°, and 13° C. warmer than the outside air. Ihne || finds the maximum air temperature at about 3 P. M. to be always below that of the tree, often many degrees. Vonhausen ** measured the temperature of a large number of beeches under varying conditions and computes the maximum temperature on the W. S. W. side between the bark and wood 120° F. with an air temperature of 91°. He emphasizes the fact that tender cells, like those of the cambium, cannot with- stand so high a temperature for along time. Russell ft says: " The temperature of the tree as a general rule ranged higher than the outside." When the leaves were removed from a tree of Abies balsamea the temperature between the wood and bark was 4-5° higher than in an uninjured tree. Ram- eaux records a similar increase in temperature in trees whose branches had been removed, amounting to 8-10° C. Hartig explains this increase by " the fact that the trees had small crowns, and that little water found its way into the younger wood rings." tt The trees in which for some reason the transpiration current had been stopped or much reduced * Bohm u. Breitenlohner, Die Baumtemperatur in ihrer Abh'angigkeit von ausseren Einfliissen. Wiener Akad. Sitzber. d. math. nat. Classe 75:615. I. Abth. t Becquerel, Compt. Rendus etc. 47: 717. 1858; 48:764. 1859. X Krutzsch, H. Untersuchungen iiber die Temperatur der Baume etc. Ko'q. Sach. Akad. f. Forst u. Landwirte zu Tharand 10. 1854. § Rameaux, M. Des temperatures vege"tales. Ann. d. Sci. nat., Bot, 19:1. 1843. || Ihne, Egan. Uber Baumtemperatur unter der Einfluss der Insola- tion. Allg. Forst u. Jagdzeit. (Supplement) 12. Heft 4. ** Vonhausen, W. Untersuchung iiber den Rindenbrand der Baume. Allgem. Forst u Jagdzeit. 49 : 8. 1873. ff Russell, H. L. Observations on the temperature of trees. Bot. Gazette 14:216. 1889.— See also Hess, iiber den Rindenbrand. Forstschutz 541- 49. 1878. XX Hartig, R. Diseases of Trees (tr. by Somerville & Ward) 295. 1894. 38 Trans. Acad. Sci. of St. Louis. always show a marked rise in temperature above that of the atmosphere. Thus the highest is met with in dead trees. In live trees, unless there is a flow of water in the new wood, there is a tendency to equalize the differences in temperature caused by excessive insolation. In the trees of the storm district the stoppage of this flow of water was felt very soon. During the summer of 1896 the roots continued to send water into the new wood, and the new leaves kept up a feeble current. In the spring of 1897, however, this current was probably very small, for by this time the roots no longer absorbed very much water. The months of June, July, August and September had some un- usually hot weather. In June, from the 11th to the 19th, the temperature rose daily above 91° F. (to 96° F. on the 17th); again, from June 29th to July 10th, the same was true. In August, for four days it was 98° F, and from August 26 (101° F) to September 16th, almost three weeks, the average maximum temperature was 95° F. These are temperatures as recorded at the St. Louis Weather Bureau, and in unshaded places, such as Lafayette Park, they were probably some degrees higher. Judging from previous observations, the temperature between wood and bark in the trees of Lafayette Park must have risen to a height sufficient to destroy the delicate cambium cells. With such a small water supply, rapid drying took place, and the bark, breaking away from the woody cylinder, finally peeled off in large pieces. This, as has been said, was true particularly of the maples, which had a smooth bark. Those trees whose trunks had been wrapped with cloth, suffered as much as the unwrapped ones, in many cases more so. The bark peeling took place mainly on the south and southeast sides, though cases of scorching on western exposures were met with. The bark broke away from the branches as well as the trunk, and by the end of September, there was many a tree left standing with no bark whatever on the southwestern side. Under the powerful insolation, rapid drying of the wood occurred, causing longitudinal split- ting, which bore much resemblance to frost cracks. PL VIII., fig. 3, shows a tree where the wood on the southern side has split in this manner, while the northern section was von Schrenk — Trees of St. Louis as Influenced by Tornado. 39 full of sap. The tree bore a crown of green leaves until the end of September. When this tree was cut down in November, it was found that the whole trunk had become almost dry, except a segment one inch in width on the north side. As the trees were being cut down in November numerous sections were taken from all parts of trunks and branches to determine the nature and extent of the drying in both wood and bark. The amount of such drying which was found was very large. Of some fifty or more maples the large number when seen in section showed but one-quarter to one-half of the woody cylinder in the normal condition. The drying usually took place in the form of a wedge, which was narrow at first, and gradually increased in width. (See PI. IX.). Analyses were made to determine the exact amount of water lost. These were made for the soft maple, box elder and linden. The summed-up results are as follows: — Number of wood rin^s. Per cent, of water. Acer dasycarpum. outer 4 outer 5 next 4 25.3 Wood from south or scorched side of trunk 12.7 15.5 Negundo aceroides. outer 11 next 8 outer 11 next 8 outer 20 outer 20 46.1 28.6 17.7 22.1 59.0 Wood from scorched side of branch 17.1 Tilia Americana. Wood from north side of trunk outer 12 outer 12 56.6 28.2 40 Trans. Acad. Sci. of St. Louis. These analyses were all made in November after the leaves had fallen from the trees. It is doubtful whether any of the trees, from which the wood was taken, would have lived another year. In comparing the results it will be seen that a marked difference existed in all cases between the scorched and healthy sides of the trees. The difference was more marked in the branches than in the trunk, which may be ascribed to their smaller diameter and consequent rapid drying. In those trees where the bark had sustained injuries during the storm, various fungi had found a foothold and their myce- lium flourished in the wood of that part, drawing the water to- the infected places. This is illustrated by fig. 1, PI. IX.,. where the small dark triangle was full of hyphae of a Basid- iomycete, and contained very much more water than the adjoining wood. On the opposite side the wood was dry and brittle, falling apart readily. The saw in cutting the limb broke the wood fibers, which accounts for the rough ap- pearance of that half in the photograph. The presence of fungus mycelium was by no means universal. Numerous maples dried very materially without losing their bark, and no fungus could enter on that account. In one tree almost 80 per cent, of the wood was dried and yet the tree showed no sign of this on the outside of the trunk. The leaves on that side were dried, and the whole tree looked sickly. No myce- lium could be detected in the wood. Where the fungus threads were present, the limits of the areas in which they were found were usually indicated by dark lines. (See PI. IX. fig. 2.) The bark scorching was very universal. Hardly a tree standing in the open was exempt, and, as a result, it may be said that wherever the branches were removed to any extent, especially in the maples, the trees lost their bark this summer. No buds were formed on such trees this fall, and the larger number had to be removed, so that, to-day, of the many beau- tiful trees, which adorned this section of the city, but a small per cent, still stand, and it seems likely that many of these will not be able to withstand the strain of another summer. von Schrenk — Trees of St. Louis as Influenced by Tornado. 41 EXPLANATION OF ILLUSTRATIONS. Plates III.-IX. Plate III. — Platanus occidentalis. Above, transection of twig two years old, X 65. Below, transection of twig injured by the storm, X 75. Both from photomicrographs, a, Wood-ring formed in 1895. b, Wood-ring formed during April and May, 1896. c, Wood-ring formed after May, 1896. Plate IV. — Acer dasycarpum. Above, transection of twig two years old. Below, transection of twig injured by the storm. Both from photomicro- graphs, X 75. a, Wood formed during April and May, 1896. 6, Wood formed after May, 1896. Plate V. — Platanus occidentalis. Transection of wood, a, Wood formed after May, 1896. b, Wood formed during April and May, 1896. /, Fall wood, 1894. s, Spring wood, 1895. Plate VI. — Acer dasycarpum. Transections of wood. 1, Fall wood fol- lowed by spring wood. 2, 6, Wood cells formed in April, 1896, and com- pressed by the wood outside of it. /, Wood formed in the fall of 1895. o, Large-lumened cells formed after May 27, 1896. r, Break in the tissue. PI. VII. — Salix nigra. Transections of wood. 1, Healthy wood. /, Fall wood, s, Spring wood. 2, Oedematous wood. I, Division of large vessel. PI. VIII. — Trees showing effects of bark scorching. 1, Bobinia Pseud- acacia. 2, 1, 6, Acer dasycarpum. 5, Ulmus Americana. PI. IX. — 1, Transection of linden trunk (Tilia Americana) showing drying effects of bark scorching. 2, Transection of maple trunk (Acer dasycar- pum), showing effects of bark scorching. In the wood between the dark lines numerous fungus hyphae occur. Issued January 29, 189S. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate III. &'' *'\ V-'"' *&v" ' - PLATANUS OCCIDENTALIS. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate IV. a b ACER DASYCARPUM. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate V. CD 5^f^% PLATANUS OCCIDENTALS. Trans. Acad. sci. of St. Louis, Vol. VIII. Plate VI. 0 0 ^0 0- u^n uOo^W^ £> — - ooO/lU'u n c=> O O = <=> q CD C3 o :0 D"0f Br ACER DASYCARPUM. Tkans. Acad. sci. of St. Louis, vol. VIII. Plate VII. SALIX NIGRA. Trans. Acad. Sci. of St. Loots, Vol. VIII. Plate VIII. EFFECTS OF BARK SCORCHING Trans. Acad. sci. of St. Loois, Vol. VIII. Plate IX. Fig. 1. tWVv « J ¥ Fig. 2. EFFECTS OF DRYING ON WOOD. Transactions of The Academy of Science of St. Louis. VOL. VIII. No. 3. NEW OR LITTLE KNOWN NORTH AMERICAN BEES. CHARLES ROBERTSON. Issued March 3, 1898. NEW OR LITTLE KNOWN NORTH AMERICAN BEES. Charles Robertson.* Prosopis thaspii, n. sp. 5. — Agrees in size with the female of P. affinis; head in front appearing more circular than usual, the eyes and vertex more convex; eyes short and broad, meeting bases of mandibles ; cheeks short and with the vertex rather strongly produced behind the eyes; head shining and rather sparsely punctured, except in front of ocelli ; an impressed line extending forward from anterior ocellus ; mesonotum and scutellum finely and rather sparsely punctured ; metathorax rather strongly sloping, hardly truncate, inclosure rugose on basal middle; abdomen shining, impunctate; insect black, flagellum beneath, mandibles at tips and tarsi dull ferru- ginous ; a short triangular spot on each side of face, two spots on collar, sometimes wanting, and basal third of hind tibiae, yellow ; the other tibiae also show a faint indication of yellow at extreme base; wings subhyaline, uervures, stigma and tegulae dark. Length 6 mm. Carlinville, Illinois ; 2 5 specimens, taken June 9th and 14th on flowers of TJiaspium aureum var. trifolialum. COLLETES NUDUS, U. Sp. 5. — Black, shining, pubescence thin, pale, except on ver- tex and thorax above, where it is blackish ; face punctured densely below, more sparsely and coarsely above, sides, of vertex shining, somewhat doubly punctured; clypeus coarsely striate punctate, apex emarginate; labruin with five more or less evident pits; mandibles, except base, rufo-piceus, space between base and the eye narrow ; antennae black or nearly so; prothorax with evident sharp spines; mesonotum punc- tured closely in front, more coarsely and sparsely on disc, like the scutellum; post-scutellum with a transverse series of sub- * Presented in abstract to The Academy of Science of St. Louis, January 17, 1898. (43) 44 Trans. Acad. Set. of St. Louis. quadate shining pits, posteriorly densely punctured ; meta- thorax with usual transverse series of pits, truncation with lateral faces bordered by salient rim ; anterior coxae with short, stout spines; wings subhyaline, second submarginal cell longer than third, tegulae and nervures dark : abdomen somewhat shining, sparsely and rather feebly punctured, ex- cept towards apex; segments somewhat depressed at base and apex, base of second and apical margins of segments 1-5 with thin fasciae of whitish pubescence. Length 10-12 mm. g. — Antennae long, fourth joint as long as second and third together; pubescence on face, cheeks and front femora rather long and dense; space between eye and base of man. dible hardly wider than in female; legs slender, abdomen a little more shining. Length 9-10 mm. Carlinville, Illinois; 3 5, 2^ specimens. Of the species known to me this is the most bare. Of those with black hairs on mesonotum the male has the most evident prothoracic spines and the shortest space between eye and mandible, the female has more evident coxae and prothoracic spines than any except O. armatus Pttn. Halictds pectoralis Sm. Halictus pectoralis Smith, Brit. Mus. Cat. Hym. 1: 68. $. 1853. $. — Closely resembles the female; clypeus without pale markings; the joints of the antennae, except the scape, of about equal length, making the antennae quite short. Length 5-6 mm. Carlinville, Illinois ; Inverness and Orlando, Florida ; 50 5, 12 $ specimens. In the female the second and third segments of abdomen have a patch of whitish pubescence on each side of base. This species varies with the punctures quite fine and sparse aud the inclosure of metathorax poorly defined. Halictus lerouxii Lep. Halictus Lerouxii Lepeletier, Hist. Ins. Hym. 2: 272. $. 1841. Halictus parallelus Smith, Brit. Mus. Cat. Hym. 1 : 72. 1853. Halictus parallelus Packard, Am. Nat. 1 : 602. 1868. $. — Resembles the female; black, clypeus anteriorly, labrum, middle of mandibles sometimes, knees, tibiae except Robertson — New or Little Known North American Bees. 45 spot in front and behind, and tarsi yellow; ventral segments 4 and 5 short, euiarginate. Length 9-12 mm. Carlinville, Illinois ; 34 §, 38 $ specimens. The male is larger than that of H. ligatus Say and smaller than that of H. parallelus Say (= H. occidentalis Cr.) and may be distinguished from both by its darker antennae and mandibles. Halictus lustrans Ckll. Panurgus lustrans Cockerell, Trans. Am. Ent. Soc. 24: 147. . June, 1897. Halictus lustrans Robertson, Ent. News 8: 172. Sept. 1897. Hemihalictus lustrans Cockerell, Can. Ent. 29: 288. Dec. 1897. I have a specimen from Cockerell. This and H. anomalus have only two submarginal cells. They are each more closely related to certain groups of Halictus than those groups are to each other. Sphecodes minor, n. sp. $. — Black, the abdomen red ; head closely punctured, more sparsely on clypeus and on each side of ocelli ; mandibles bidentate, rufo-piceus at apex ; antennae dull testaceous at apex beneath, joints of flagellum about as long as wide; mesonotum smooth and shining, sparsely and rather coarsely punctured, a fine median line slightly impressed, or even raised anteriorly; scutellum sparsely punctured; metathorax truncate, coarsely reticulated ; wings somewhat clouded be- yond middle, nervures, stigma and tegulae dark, second sub- marginal cell narrowing above ; abdomen shining, almost impunctate, entirely red. Length 8-10 mm. Carlinville, Illinois ; 10 5 specimens. This species runs smaller than 8. dichrous Sm. (=8. arvensis Pttn.), though some examples of the latter are no larger, and a little larger than 8. confertus Say (= 8. falcifer Pttn.). It differs from 8. dichrous in having the punctuation of the head and thorax finer, though of the same style, the metathorax more rugose, the abdomen brighter red, without black apex and without the strong punctuation. 46 Trans. Acad. Set. of St. Louis. Andrena illinoensis Rob. form bicolor, n. f. Two female specimens agree with the normal form in all respects except that the abdomen is reddish, except two black spots on second segment indicating the position of the usual foveae. A\DRENA MARIAE Rob. form CONCOLOR, n. f. This species was based on examples with the abdomen red. I separate as a form the specimens which are entirely black. The abdomen is usually all red or entirely black, only rarely black with a trace of red. All of the males I have seen have the abdomen black, or sometimes with more or less red. Andrena forbesii Rob. Andrena forbesii Robertson, Trans. Am. Ent. Soc. 18: 59. 9- 1891. . — Resembles A. forbesii; the head, except the elypeus, more coarsely and less closely punctured, the pubescence thinner, shorter and paler ; thorax above more coarsely and sparsely punctured, pubescence densely plumous, but short and thin, except about shoulders ; metathorax coarsely reticulated, the inclosure less distinct ; abdomen more strongly and much more densely punctured, except the base of first segment, which is more shining and less closely punctured ; fasciae more dense and whiter, interrupted on second segment. Length 10-11 mm. Carlinville, Illinois, 4 o_ specimens, taken July 19th, 26th and 29th, on flowers of Pycnanthemum linifolium Ph. Of forty-two species occurring in my neighborhood, this is the only one flying at that time. Of the species of Andrena known to me this has the abdomen most densely punctured. The scutellum and post- scutellum resemble those of A. nuda; the pubescence also like that of A. nuda, but more abundant on mesonotum. Robertson — New or Little Known North American Bees. 47 Parandrena andrenoides (Cr.) Rob. form bicolor, n. f. All males I have seen are black. I assume the black female to be normal and propose the form name for the female with red abdomen. Intermediate forms are rather rare. Of the four species, Andrena erythrogasira Ashm., A. illi- noensis Rob., A. mariae Rob. and Parandrena andrenoides, the first is the only one in which the abdomen of the female is not commonly, or usually, entirely black. A. erythrogastra and A. mariae are the only ones in which the males show even a trace of red. A specimen of A. sphecodina received from Prof. Cockerell is the only male I have seen with the abdomen quite red. This is probably a synonym of A. mariae. By itself the red color is of no value except as indicating a mere variant form. In Trans. Am. Ent. Soc. 24: 150, Prof. Cockerell says : ** Robertson has suggested that P. andrenoides may be a Scrapter, but Dalla Torre places Scrapter as a synonym of Macropis, and certainly andrenoides is not a Macropis." Nevertheless, P. andrenoides may be a Scrapter and also a Macropis in the sense of Dalla Torre. It only follows that his Macropis must be defined to include Scrapter, just as his Podalirius must be defined to include Enteclinia and his Eucera to include Emphor. In Bull. Soc. Ent. France, 1897, 63, Vachal discusses the characters of Scrapter and shows the composite character of Macropis D. T. It appears that Scrapter differs from An- drena only, or mainly, in having but two submarginal cells. He says : " De tout ceci il resulte que le chapitre Macropis de M. Dalla Torre doit etre remanie et les especes qu'il contient etre reparties an moins en trois genres, et merae en quatre, s'il resulte de l'examen des types que'les especes australiennes de Smith ue peuvent entrer dans aucun desautres." Whether Parandrena is a section of Andrena or the same as Scrapter remains to be seen. P. andrenoides and ivelles- leyana are quite distinct. The males agree in having the angles of the sixth ventral segment reflexed. Andrena crataegi, which is not related, is the only Andrena in which I have found a similar structure. 48 Trans. Acad. Sci. of St. Louis. Stelis lateralis Cr. Stelis lateralis Cresson, Proc. Eat. Soc. Phil. 2: 410. ? (?). 1864. £. — Typically the abdomen has ten spots which are elongate and indented posteriorly, the first three segments usually showing the rounded lateral portions of these spots. It may become 12-16 spotted by the marks on segments 3-5 breaking into spots. . — The abdomen is 8-14 spotted, in various combinations, the spots on 5th segment wanting or nearly so, and those on 4th small. 8-spotted examples may have four spots on each extreme side, or three on each side and two sub-discal on the 4th segment. Carlinville, Illinois: 9 £, 8 $ specimens. I think the type specimen is a male. Megachile exilis Cr. Megachile exilis Cresson, Trans. Am. Ent. Soc. 7 : 224. 1879. This was determined for me by Mr. Cresson. Thirty local males differ from the type by having the anterior tarsi black iustead of pale ferruginous. Calliopsis andreniformis Sm. Calliopsis andreniformis Smith, Brit. Mus. Cat. Hym. 1: 128. 9. 1853. In the Synopsis, 295, Cresson indicates that C. flavipes Sm. is the male. To be sure, Dalla Torre retains the latter name, but the American part of his catalogue is not connected with the literature. Several times I have taken the sexes in copula. Panurginus labrosus Rob. Calliopsis labrosus Robertson, Trans. Am. Eat. Soc. 22: 123, in part. J*?. 1895. §. — Black, shining, thinly pubescent ; face with shallow punctures of medium size, rather sparse, except above anten- nae ; labrum with median space striate, strongly narrowed to a truncate or rounded apex ; third joint of antennae about as long as two following together, middle joints of flagellum testaceous beneath; mandibles rufopiceus ; anterior and middle knees and tubercles yellow ; prothorax and about tubercles with short ochraceous tomentum ; mesonotum shin- Robertson — New or Little Known North American Bees. 49 ing, sparsely punctured except anteriorly; wings hyaline, or nearly so; second submarginal cell shorter than first, nar- rowed about one-third towards marginal ; nervures and stig- ma fuscous, tegulae testaceous; legs blackish, scopa white; disc of metathorax short, irregularly striate, posterior face shining and impunctate; abdomen shining, especially the first segment, which is almost impunctate, following segments finely punctured, anal fimbria pale ochraceous. Length 6-7 mm. $. — Mesonotum more closely punctured; clypeus, a spot above, sides of face nearly as high as antennae, labrum, man- dibles except tips, tubercles, tibiae at base and apex, anterior pair, and sometimes middle and hind pair, in front, and tarsi, yellow ; middle joints of antennae pale testaceous beneath ; median face of labrum narrowed toward apex and slightly emarginate. Length 6-7 mm. Carlinville, Illinois ; 8 §, 5 ^ specimens. The sexes were taken in copula. Panurginus labrosiformis , n. sp. Calliopsis labrosus Robertson, Trans. Am. Ent. Soc. 22: 123, in part. c?$. 1895. . — A little smaller than P. labrosus, more closely punc- tured; mesonotum more coarsely punctured and presenting three impressed lines in front ; antennae shorter, third joint not equaling next two together, flagellum more testaceous beneath towards tip ; anterior margin of clypeus, labrum, and mandibles often of same color; tubercles dark; nervures paler; posterior face of metathorax shining and impunctate above, evenly punctured below. Length, 6 mm. <£. — Resembles the female; clypeus, a small spot above, sometimes wanting, narrow spots on each side of face; labrum, mandibles except tips, knees, anterior tibiae in front, tarsi, except apical joints of hind pair, yellow ; compared with male of P. labrosus smaller, less yellow, antennae shorter, darker, mesonotum more coarsely punctured and with three impressed lines in front, median face of labrum more strongly narrowed anteriorly. Length, 5 mm. Carlinville, Illinois ; 15 5, 7 $ specimens. 50 Trans. Acad. Sci. of St. Louis. Panurginus solidaginis Rob. Calliopsis solidaginis Robertson, Trans. Am. Ent. Soc. 20: 274, £. 1893. ^. — Closely resembles the male of the preceding, but the median face of labrum is not so strongly narrowed as in that species or P. labrosus. Length, 5-6 mm. Carlinville, Illinois; 11 $, 7 $ specimens. Panurginus rugosus Rob. Calliopsis rugosus Robertson, Trans. Am. Ent. Soc. 22: 121. 9cT" 1895. Calliopsis fraterculus Cockerell, Can. Ent. 28: 159. tfQ. 1896. Pseudopanurgus fraterculus Cockerell, Can. Ent. 29: 290. 1897. From comparisons with P. aethiops Cr. it is evident that Prof. Cockerell was not aware that that species already had three " j "rater 'cidi" in P. mexicanus Cr., P. scaberYox and the above. The form of the tubercle at vertex varies in both sexes. 1 have 14 $, 6 $ specimens. Cockerell on Panurgus and Calliopsis, Can. Ent. XXIX., 287-90. — From comparisons with European material received from Friese of Innsbruck, the author refers Calliopsis com- positarum and rudbechiae to Panurginus. I accept this and refer C . asteris and parvus to the same genus, as also the other species indicated above. The reference of Panurgus marginatum to Halicloides is probably correct and P. novae- angliae goes with it. Prof. CockerelPs treatment of these genera is likely to create an erroneous impression. In Trans. Am. Ent. Soc. 24: 150, he says: " It is perfectly evident that the so-called species of Panurgus of North America are not all of the same genus," and, in place cited above, *« The result is extremely interesting, and seems to show that we have for many years been placing bees in genera to which they by no means belong." To those acquainted with the literature and the bees in question this has been clearly understood, at least since it was distinctly stated by Cresson ten years ago. In the Synopsis, 134, Cres- son says: "The genera Panurgus, Calliopsis and Per dita, have been made the receptacle for a number of species which do not properly belong to either of those genera, and have been placed there provisionally until more abundant material can be obtained, when a more careful study may be made of their characters." Robertson — Neio or Little Known North American Bees. 51 NOMADA RUBICUNDA OHv. Nomada rubicunda Olivier, Enc. M£th. Ins. 8: 365. 1811. Nomada rubicunda Cresson, Proc. Ent. Soc. Phil. 2: 299. 1863. Nomada torrida Smith, Brit. Mus. Cat. Hym. 2: 250. £. 1854. I have a female specimen taken at Orlando, Florida. Nomada obliterata Cr. Nomada obliterata Cresson. Proc. Ent. Soc. Phil. 2: 301. £. 1863. Nomada bisignata var. obliterata Cresson, Trans. Am. Ent. Soc. 7: 1879. Nomada bisignata var. obliterata Cresson, Synopsis 297. 1887. Nomada viburni Robertson, Trans. Acad. Sci. St. Louis 7: 341. ^ . 1897. This is a good species. I did not capture the female until after I had described the male. I have 2 § and 6 $ specimens all agreeing with the type specimen in having the first two submarginal cells united, except one which has two submar- ginal cells in one wing and three in the other. One female agrees with the description of the type. The other has more yellow on face and metathorax, a line on each side of mesono- tum over the tegulae, two spots on pleura and line on post- scutellum yellow. Length $t 7-9 mm. ; 0., 8-10 mm. Nomada articulata Sm. Nomada articulata Smith, Brit. Mus. Cat. Hym. 2: 248. . — Does not require description because of close resem- blance to male. Carlinville, Illinois; 14 $, 37 $ specimens. There is a large tubercle on each side of front. Epeolus pusillus Cr. Epeolus pusillus Cresson, Proc. Ent. Soc. Phil. 2: 398. $ . 1864. Epeolus pusillus Fox, Ent. News 3: 29. ,j\ 1892. ? Epeolus mercatus Fabricius, Syst. Piez. 389. 1804. This seems to be the best guess for E. mercatus. Cresson credits E. pusillus to Mass. and N. H., and Fox captured it in New Jersey. Cresson's argument does not hold against E. pusillus, for he had evidently forgotten it when he wrote : " This \_E. cressonii Rob.] is probably the true mercatus, although the very short description given by Fabricius will apply quite as well to several other species, not found, how- ever, east of the Mississippi River." f Fox captured E. com- pactus in New Jersey, but this is a little larger, has the legs darker and is probably less abundant. Melissodes comptoides, n. sp. 5. — Black; head clothed with pale pubescence, except on the vertex where it is long, fulvo-ochraceous and usually mixed with black in front; flagellum testaceous beneath; mandibles rufous in middle, at apex with an orange streak, often reduced or wanting ; pubescence of thorax above thick and fulvo-ochraceous, on the sides pale, beneath fuscous or black ; legs with pubescence largely black or fuscous, mixed with oehraceous on femora and tibiae above, anterior tarsi fulvous beneath ; scopa fulvo-ochraceous, hind metatarsus with black pubescence beneath; base of abdomen with pubes- t Trans. Am. Ent. Soc. 7: 88. 1878. Robertson — Neiv or Little Known North American Bees. 53 cence like that of thorax above, second segment with narrow basal and median fasciae of pale appressed pubescence, third with a broad basal fascia, fourth with a broad band with a subtriangular notch posteriorly on the disc, pubescence be- neath black ; wings subhyaline, apical margins clouded, nervures dull ferruginous. Length 12-13 mm. $. — Head, thorax and legs without black or fuscous hairs ; thorax above, and usually the vertex, fulvo-ochraceous, below paler, but the legs sometimes with pubescence fulvo-ochra- ceous; clypeus, labrum and base of mandibles yellow; anten- nae beneath fulvous, third joint hardly longer than second, fourth joint nearly as long as next two together; abdomen with fasciae, resembling those of female, the last three seg- ments with black pubescence, beneath the segments have black or fuscous pubescence, pale laterally ; tarsi ferruginous Length 11-13 mm. Caiiinville, Illinois: 11 5; 44 g specimens. The male suggests M. compta on account of the black tip of abdomen. MELISSODES AMERICANA Lep. Macrocera americana Lepeletier, Hist. Ins. Hym. 2: 92. . — The interrupted pubescent fasciae on third and fourth segments are either white or black, so that the abdomen has two or four white spots or none. g. — More like the female than any other male I have seen ; base of clypeus labrum and base of mandibles pale yellow ; third joint of antennae about half as long as fourth, fiagellum ful- vous beneath; third segment of abdomen with an oblique white patch on each side. Length 14-16 mm. Carlinville, Illinois; 5 £, 2 ^ specimens. Issued March 3, 1898. Transactions of The Academy of Science of St. Louis. VOL. VIII. No. 4. ECOLOGICAL PLANT GEOGEAPHY OF KANSAS. A. S. HITCHCOCK. Issued March 8, 1898. ECOLOGICAL PLANT GEOGRAPHY OF KANSAS.* A. S. Hitchcock. The present paper is but a preliminary study of our plant geography. An attempt is made to outline the classes of plant communities found within our borders, but a presenta- tion of the adaptation peculiar to the several classes has been reserved for a future paper. The nomenclature is that of Gray's Manual, sixth edition. In giving representative lists the rare and local species have been omitted. Only vascular plants are considered. The eastern fourth of the State is carboniferous, mainly Upper Coal Measures extending west to Cowley, Chase and Riley Counties. West of this is a rather narrow Permian belt, fifty miles in width, more or less. West of this and along the southern border only are the Red-beds, extending from Sedgwick to Clark County. Two narrow areas, first Dakota and then Benton, occupy the north central portion of the State from Republic, to Rice and Hodgeman Counties. The western portion of the State is Tertiary, occupying something less than half the area of the State. There are smaller areas of some other formations, such as Niobrara and Comanche. The State is 400 miles in length east and west, by 200 miles in width, north and south. The altitude varies from about 750 feet at the east (in the valleys) to nearly 4,000 feet on the high plains at the west. The western portion of the State is occupied by the Great Plains. This region extends eastward approximately as far as the western limit of the Permian. The Great Plains is a tree- less grassy area, nearly level except as broken by the deep valleys. These valleys cut through the underlying strata, giv- * Presented to The Academy of Science of St. Louis, February 7, 1898. (55) 56 Trans. Acad. Sci. of St. Louis. ing rocky bluffs, such as the Chalk Bluffs of the Smoky Hill river. In some places the erosion gives rise to canons and buttes. Some of the rivers have a scant supply of timber, which increases as one goes east, but often the valleys are treeless as the Cimarron. The region covered by the Red- beds is characteristically sandy, as is much of the region to the north as far as the Arkansas river. In the eastern part of the State, the timber along the water- courses increases, and in many places extends over the uplands. The prairie regions are restricted to smaller areas. Limestone hills are a prominent feature of the landscape. In 4Ql£ti \\s ISLi tOOOft. S000JU HOOft. MAP OF KANSAS. (Reduced, from 8th Rept. Kansas Bd. Agriculture.) the southern part are found the Flint hills extending from Chase to Chautauqua County. Salt plains are found in Stafford and Reno Counties. One of the most important factors in determining the dis- tribution of vegetation is the rain-fall. The two eastern tiers of counties have above 41 inches. The next tier or so 30 to 40 inches. Three-fourths of the State has usually less than 30 inches, the amount decreasing westward. There is usually not much snow during the winter, and when it comes it usually remains upon the ground only a short time. The temperature varies between — 25° F. and 110° F. Hitchcock — Ecological Plant Geography of Kansas. 57 or even more. The minimum temperature is probably 20° higher in southern Kansas than in northern. In discussing the ecological plant geography, I will follow the classification of Warming, dividing the various communi- ties into four divisions: Hydrophytes, Xerophytes, Halo- phytes and Mesophytes. Hydrophytes. Plants which love the water. They include water plants and swamp plants. There are three groups : A. Plants which are not attached to the substratum and may be sub- mersed or floating. B. Plants attached to the soil but the vegetative portion not extending much above the surface of the water. They include submersed plants and also forms with floating leaves. C. Swamp plants in which the vegeta- tive portion extends into the air. The hydrophytic vegetation is distributed throughout the State. There is some along every water-course, but the amount found here is comparatively scant. In eastern Kan- sas bayous and lakes left in old river channels have usually an abundant supply of hydrophytes. Here and there are found isolated swamps in depressions that are not directly connected with a water-course. Where springs break forth there is usually a swampy area. In western Kansas there are in many places perennial slow- flowing streams of clear water which support a rich hydro- phytic vegetation. In some localities, as in Meade and Kiowa counties, large springs break forth into a small river at once. Around these springs one finds besides a rich hydrophyte flora, many species of eastern Kansas. One other habitat for hydrophytes is found in the buffalo wallows. These are depressions found on the high plains, which vary in diameter from a few feet to many rods. CLASS I. FLOATING PLANTS. These are found only where there is open water. They float on the surface or at a short distance below. Of flowering plants there are but a few species. Utricularia vulgaris, Ceratophyllum demersum, Spirodela polyrrhiza, 58 Trans. Acad. Sci. of St. Louis. Lemna perpusilla, and var. trinervis, Wolffia Columbiana. Jussiaea repens is also often found floating free. CLASS II. WATER PLANTS ROOTED IN THE SOIL. This class is represented here and there over the entire State, in ponds, springs, the central part of marshes where there is visible water, and (especially westward) in slow-flow- ing streams. Some are submersed, when the leaves are finely divided, others have floating leaves, but the vegetative portion does not rise above the surface of the water, except as the water may evaporate and leave the plant in the mud. The flowers are pushed above the surface but the fruit is usually submersed. Ranunculus divaricalus, R. multifidus, Kelumbo lutea. The leafy blades of the last are usually raised out of the water, yet the plant should undoubtedly be classed here. Nympliaea odorata, JSTuphar advena (leaf blades frequently emersed), Nasturtium officinale, Myriophyllum spicatum, M. scabratum, Callitriche heterophylla, Jussiaea ?*epens, Ilerpestis rotundi- folia (also a swamp plant), Elodea Canadensis, Heteranthera limosa, H. reniformis, H. graminea (these three grow also in the mud), Polamogelon (several species), Ruppia maritimay Zannichellia palustris. CLASS III. SWAMP PLANTS. The plants are rooted in the soil, which is saturated with water and may be submersed, but the vegetative organs extend above the water when it is present as a sheet. This class is usually found around the borders of the preceding class and also along rivers and ponds in which the preceding class may be absent. The species are too numerous to give a detailed list. But we may distinguish a few communities which often have many species in common but which possess a distinctive physiognomy. Reed-Sioamps. These occur in swampy depressions or along water-courses. The plants are mostly with upright leaves and are chiefly monocotyledons. During the rainy season (spring and early summer) the swamp may show much open water between the plants, but later the water may Hitchcock — Ecological Plant Geography of Kansas. 59 entirely dry up leaving the plants to winter in the hard ground. The following would be characteristic: Typha latifolia, Sparganium eurycarpum, Alisma Plantago, Sagit- taria (several species), Eleocharis (several species), Scirpus lacustris and other species, Carex (several species, as O. fili- formis latifolia, C. hystricina). Bogs. These are found in the neighborhood of springs and hence are supplied with water the year around. Many of the plants of the preceding sub-class, are found here, but particularly the following: Cardamine rhomboidea, C. parvi- Jiora, LytJirum alatum, Ludwigia alternifolia, Epilobium lineare, E. coloralum, Berula angustifolia, Eupatorium per- foliatum, Mimidus Jamesii, Gerardia tenuifolia, Lycopus rubellus, L. lucidus Americanus, L. sinuatus, Juncus (several species), Cyperus (several species, as (J . esculentus) , Phrag- mites communis (local). Buffalo Wallows. Depressions in the soil found scattered over the plains in western Kansas. During the wet season they hold water but are dry in the latter part of the year. The plants are mostly annuals. Ammannia Wrightii, Oenothera canes- cens, O. triloba parvijiora, Aclinella odorata, Lippia cunei- folia, Eriochloa polystachya, Panicum colonum, Eragrostis pilosa . Borders of ponds and marshes (more rarely borders of streams). This formation might more properly come among the mesophytes but it will be included here for convenience. It occurs between the swamp formation and the true meso- phyte flora. Lippia lanceolata, Leersia oryzoides, L. Vir- ginica. /Sandy or muddy borders of streams. This formation is found on sand-bars or mud-banks in the latter part of sum- mer when the water has subsided. The plants are mostly small annuals. Conobea mullifida, Ilysanthes riparia, Cy- perus aristatuSf C. diandrus, Eragrostis replans. CLASS IV. SHRUB-SWAMPS. These are found only through the eastern part of the State. There is much open water, and herbaceous vegetation is more or less suppressed by the shrubby growth. In between the 60 Trans Acad. JSci. of St. Louis. patches of shrubs will usually occur reed swamp vegetation. The most common shrub is Gephalanthus occidentalis . In southeastern Kansas, the tall herb Hibiscus mililaris often occurs in the community. Salix longifolia may occur in swamps, but forms characteristic communities along river banks, especially portions liable to inundation. Xerophytes. As one would suppose from its geographical position, the flora of Kansas is largely xerophytic. The group may be conveniently divided into three classes : A. Rock vegetation. B. Sand vegetation. C. Prairie. CLASS V. ROCK VEGETATION. The exposed rock is chiefly limestone. There is some hostyles pauciforus, Cassia chamae- crista, Liatris squamosa, Heterotheca Lamarckii, Chrysopsis villosa, Aplopappus divaricafus, Gnaphalium polycephalum, Helianthus petiolaris, Artemisia caudata, TAthospermum angust folium, Ipomoea leptophylla, Peutstemon grandiflora, Monarda punctata, Amaranthus Torreyi, Froelichia Flori- dana, F. gracilis, Cycloloma platyphyllum, Euphorbia hexagona, Cyperus Schiveinitzii, C. filicuhnis, Paspalum setaceum, Cenchrus tribuloides, Sporobolus asper, 8. cryptan- drus, Ammophila longifolia, Triodia cuprea, Fragrostis tenuis, F. pectinacea, F. capillaris. Hitchcock — Ecological Plant Geography of Kansas. 63 In the western part of the State some of these disappear and the following become common : (Jleomella angustifolia , Talinum calycinum, Indigo/era leptosepala (S. W. Kansas), Gaura villosa, Mentzelia mulliflora, Erigeron divergens, Aster tanacetifolius, Hymenopappus jlavescens, Polypteris Hooheriana, Gaillardia pulchella, G. arista ia, Artemisia fili- folius, Gilia aggregata, Heliotropium convolvulaceum, Kry- nitzkia Jamesii, Asclepias arenaria, Physalis lanceolata pu- mila, P. hederaefolia, Oxybaphus sp. (allied to O. albidus), Abronia fragrans, (Jorispermum hyssopi folium, Eriogonum annuum, Euphorbia Geyeri, E. petaloidea, Slillingia syl- valica, Commelina angustifolia, Andropogon Hallii, Stipa, comata, Redfieldia flexuosa, Gymnopogon racemosa (S. Kansas), Eragrostis oxylepis. A large number of the species are not found, within our limits, in any other than sandy soil. Several plants that are characteristic of the plains in west Kansas extend eastward in the sand-hills, such as Ipomoea leptophylla. Some plants of the surrounding prairie may extend more or less into the sand-hills, such as Ambrosia psilostachy a. The cactuses are rarely found in the sand-hills. A large proportion of the species are annual. The peren- nial plants frequently have extensively creeping root-stocks, such as the binding grasses mentioned. Others form deep slender roots which appear to reach down to permanent moisture, as Physalis and Asclepias. Artemisia forms a thick crown and helps materially in retarding the shifting sand, as do some of the bunch grasses. Ipomoea leptophylla and Cycloloma are tumble weeds. In the eastern sand-hills a shrub, Prunus angustifolia, fre- quently forms quite extensive thickets which tend to bind the sand. Amorpha canescens is also frequent. This small shrub is also common on the prairie. CLASS VII. PRAIRIE. Prairie is more extensive than all the other formations together. The western half of the State is in the region known as the Great Plains, which extends west to the moun- tains, south into Texas and far northward. In this region the 64 Trans. Acad. Sci. of St. Louis. land is nearly level, broken here and there by valleys cf streams. One can often travel many miles on the uplands with no perceptible undulation to break the monotony of the level plain extending to the horizon on all sides. There are no trees, no shrubs (though many of the plants have a lig- nescent base), no tall herbs. The distribution of Malvastrum coccineum shows fairly well the extent of this region. Our collections show the eastern limit of this species to be the following counties: Kepublic, Cloud, Ottawa, Ellsworth, Kice, Stafford, Pratt, Barber. East of this region there is much prairie on the uplands, but it decreases as one goes eastward, so that in the eastern counties the amount is com- paratively small as to total and limited in extent as to individual areas. The following plants are characteristic of the western plains: Erysimum, asperum, Poly gala alba, Malvastrum coc- cineum, Linum rigidum, Sophora tomentosa, Psoralea tenui- flora, Cereus viridiflorus (S. W.Kansas), Opuntia Rafinesquii> O. Missouriensis , 0. fragilis, Gutierrezia Euthamiae, Aplo- pappus spinulosus, Evax prolifera, Engelmannia pinnalijida, Thelesperma gracile, Artemisia Wrightii; Senecio Douglasii, Cnicus ochrocentrus, Asclepias Jamesii, Krynitzkia crassise- pala, Ipomoea leptophylla, Solarium trifiorum, (J hamaesaracha sordida, Verbena bipinnatifida, Cladothrix lanuginosa, (Jhe- nopodium olidum, C. Fremonti incanum, Allium Nuttallii, Aristida purpurea, Munroa squarrosa, Elymus Sitanion. The following also extend further east: Kulinia eupatorioides , Liatris punctata, Solidago Missouriensis, Ambrosia psilosta- chya, Lepacliys columnaris, E chinospermum liedowskii occi- dentale, Evolvulus argenteus, Solanum rostratum, Oxybaphus angustifolius, Andropogon furcatus, Andropogon scoparius, Chrysopogon nutans (these three grasses are predominant in the eastern prairies), Schedonnardus Texanus, Bouteloua oli- gostachya, B. racemosa, Buchloe dactyloides, Koeleria cristata, Eatonia obiusata. In the so-called " second bottom " of the valleys are found: Cleome integrifolia, Astragalus mollissimus, A. gracilis, Cucurbita foetidissima, Sporobolus airoides.- Hitchcock — Ecological Plant Qeography of Kansas. 65 The physiognomy of the western plains is peculiar. The predominant plant is buffalo grass (Buchloe daclyloides) with often considerable grama grass (Bouteloua oligostachya). This forms a close mat two or three inches high of a grayish- green color. The other plants are scattered here and there in this sod. The Opuntias are conspicuous though they creep along edgewise, and rise but little above the grass. The most conspicuous plants are Asclepias Jamesii and Cnicus ochrocentrus, which are one to two feet high. Certain of the plants in the list are found chiefly in new ground, such as fire-guards and prairie-dog towns. These are: E 'chinospermum , Krynitzkia, Solatium triflorum, S.ros- tratum, Chamaesaracha, Verbena, Cladotlirix, Chenopodium Fremonti incanum, Schedonnardus, Munroa. I have spoken previously of the xerophytic character of the plants of the plains (Contr. Nat. Herb. 3: 538). Though the species vary somewhat as one goes west, the physiognomy remains the same and the community cannot be broken up readily into subformations, as is the case with most of the other classes. In eastern Kansas the species in addition to those mentioned above are: Anemone decapetala, Delphinium azureum, Cal- Jirrhoe triangulata, C. alcaeoides, Linum sulcatum, Baptisia australis, B. leucophaea, Psoralea argophylla, P . floribunda, P. esadenta, Petalostemon violaceus, P. candidus, Astragalus caryocarpus, A. Plaltensis , Desmodium Illinoense, Lespedeza capitata, Schranhia uncinata, Liatris pycnostachya, L. scari- osi, Grindelia squarrosa, Solidago rigida, Silpliium lacini- atum (compass plant), Helianthus 7'igidus, Coreopsis palmata, Cnicus undulatus, Asclepiodora viridis, Lithospermum an- gustifolium, Salvia azurea grandifiora, Panicum scoparium, P. depauperatum, Stipa spartea. Through eastern Kansas occur small areas the vegetation of which probably should be classed with the xerophytes. They are usually referred to as " sterile places." They are open barren spots owing their dryness to the character of the soil. The following plants are frequently found in such localities: Draba Caroliniana, Silene antirrhina, Sagina decumbens, Hypericum Drummondii, Slylosanthes elatior, 66 Trans. Acad. Sci. of St. Louis. Cuphea viscosissima, Diodia teres, Androsace occidentalis, Iledeoma hispida, Plantago Patagonica gnaphalioides. The nearest representation which we have of the xerophy- tic forest flora is the oak scrub, found on dry hills through eastern and southern Kansas. The trees are mostly oaks, Quercus nigra, Q. tinctoria, Q. Mulilenbergii and Q. prinoides. Halophttes. This sub-division is rather sparsely represented in Kansas, the plants being confined to the salt marshes. CLASS VIII. SALT MARSH VEGETATION. These salt marshes are found in Stafford and Reno Coun- ties, and less frequently in some other counties. During the dry season the water disappears and there is a salt plain. In Stafford County the Big Salt Marsh covers several square miles, and during the dry season a large part of it is entirely devoid of vegetation, a level sheet of white. There are cer- tain species found only where the soil is more or less salty, but these species are often mixed in with others in soil that is only slightly salty, not enough to be called a salt marsh or plain. These areas are commonly known as alkali spots. The list of species is comparatively short. Aster exilis, Flaveria angustifolia , Ghenopodium rubrum, Atriplex ex- pansa, A. argentea, Suaeda diffusa, Scirpus maritimus, Distichlis maritima, Agropyrum glaucum. All but the last three are annual, these three perennial by creeping root- stocks. Distichlis and Agropyrum are widely distributed. The former is quite abundant at the foot of gypsum hills where the soil is impregnated with gypsum and magnesium, but is wet during a part of the year. Mesophytes. The mesophytic flora is abundant in eastern Kansas, but decreases rapidly westward, where it is confined to the vicinity of the streams. CLASS IX. MEADOWS. Under this title is included the grass vegetation lying near the streams, but beyond the hydrophytic formation. The Hitchcock — Ecological Plant Geography of Kansas.- 67 meadows are usually subject to overflow during the wet sea- son and may be quite dry during the dry season. They are commonly known as bottom lands. Also much of what are termed sloughs come under this class, though they may have a hydrophytic formation in the center. Grasses (especially Spartina cynosuroides) and sedges predominate, but have a mixture of various perennial herbs scattered through. It is scarcely worth while to give more than a short list of the species which will serve as examples. In eastern Kansas the bottom land flora would be illustrated by Desmanlhus brachylobus, Vernonia Baldwinii, Solidago lanceolata, Aster paniculalus, Erigeron Canadensis, Iva ciliaia, Ambrosia trifida, Helianthus grosse-serratus, Poly- gonum incarnatum and other species, Carex straminea and other species, Spartina cynosuroides. CLASS X. FIELDS. This formation is found on dryer ground than the pre- ceding and in the wooded region would be replaced by forest if this were not prevented by man. In other regions the tendency is to return to prairies. In cultivated fields the plants which intrude are known as weeds. These have already been discussed in the Bulletins of the Kansas Experi- ment Station (Nos. 50, 52, 57, 66, 76). In fields formed by breaking prairies the weeds should in most cases be classed as mesophytes. CLASS XI. DECIDUOUS FORESTS. As previously stated considerable of the area in the east- ern part of the State is covered with forests. These forests are best developed in the valleys of the rivers, especially the Missouri, Kansas, Marais des Cygnes, Neosho and Verdigris. The forest extends in many places onto the uplands, but becomes then quite sparse. As one travels west the belt of timber along the streams becomes narrower and the species fewer till it entirely disappears in the western part of the State. It is noticeable that in many parts of the State where the timber may be scattering or absent along the main river, one can find a much better representation at the heads of the 68 Trans. Acad. Sci. of St. Louis. draws or canons in the edges of the valley some distance from the stream. A full discussion of the distribution of Kansas trees will be found in the eighth Biennial Report of the Kansas State Board of Agriculture.* I will give here only an outline of communities as so far observed. Copses. Thickets found along the margins of woods or in the meadows previously described. Also along the banks of water-courses. They consist of Rhamnus lanceolata, Rhus glabra, Pyrus coronaria, P. loensis, Prunus Americana, Cor- nus asperifolia, Corylus Americana. Where the soil is moist Amorpha frulicosa is common. Climbing vines are com- mon, such as Celastrus scandens, Vitis riparia, Polygonum dumeiorum scandens. In the moist thickets along streams, Apios tuberosa, Amphicarpaea monoica, Echinocystis lobata and Sicyos angulata, occur. Along fences and similar habitats one finds many of these shrubs and also Rhus Toxicodendron, Rubus occidentalis, Rubus villosus and Rosa setigera. With these are found a number of tall herbs such as Solidago Canadensis, S. serotina, JErigeron Canadensis, Lactuca Canadensis. Lowland Woods. In the bottom land along streams which are not subject to overflow except in unusually high water we find the timber reaching its greatest development. The fol- lowing are characteristic. — Trees: Asimina triloba, Tilia Americana, Acer dasycarpum, Negundo aceroides, Cercis Canadensis, Gymnocladus Canadensis, Gleditschia triacan- thos, Ulmusfulva, U. Americana, Celtis occidentalis, Morus rubra, Platanus occidentalis, Juglans nigra, Carya olivae- formis, C. sulcata, C. amara, Quercus macrocarpa, Q. palustris, Salix amygdaloides, S. nigra, Poprdus moni- lifera. Shrubs: Menispermum Canadense, Xanthoxylum Americanum, Vitis cinerea, V. cordifolia, Aesadus argula, Stajrfiylea trifolia, Ribes gracile, Sambucus Canadensis, Symphoricaiyos vulgaris (found in several communities), Smilax hispida. Herbs: only a very few common ones will * Mason, A Preliminary Report upon the Variety and Distribution of Kausas Trees, p. 259. Hitchcock — Ecological Plant Geography of Kansas. 69 be given : Ranunculus abortivus, Viola palmata cucullata, V. pubescens, Claytonia Virginica, Geum album, Cryptotaenia Canadensis, C haerophyllum procumbens, Sanicula Marilan- dica, Eupatorium purpureum, E. ageratoides, Solidago ulmi- folia, Ambrosia trifida, Rudbeckia laciniata, Actinomeris squarrosa, Bidens frondosa, Lactuca Eloridana, Campanula Americana, Ellisia Nyctelea, Echinospermum Virginicum, Scrophularia nodosa Marylandica, Polygonum Virginicum, Laportea Canadensis, Pilea pumila, Erythronium albidum, Cinna arundinacea. Upland Woods. Extending along the bluffs, hillsides or uplands. Trees: Acer saccharinum, Eraxinus Americana, Carya alba, Ostrya Virginica, Quercus stellala, Q. Muhlen- bergii, Q. coccinea tinctoria, Q. nigra. Herbs: where the soil is rich with leaf mould, such as moist hillsides near the larger streams in eastern Kansas, one finds, Isopyrum biternatum, Podophyllum peltatum, Sanguinaria Canadensis, Dicentra Cucullaria, Dentaria laciniata, Osmorrhiza brevistylis, 0. longistylis, Asarum Canadense, Polygonatum giganteum, Smilacina stellata, Uvularia grandijtora, Arisaema tri- phyllum, Uniola latifolia, Poa sylvestris. In the dryer and more open woods the following are com- mon : Agrimonia Eupatoria, Triosteum perfoliatum, Soli- dago Liindheimeriana , Aster Drummondii, A. panicidatus, Antennaria plantaginifolia , Helianthus liirsutus, Cacalia atriplicifolia, Veronica Virginica, Verbena Aubletia, Phryma leptostachya, Teucrium Canadense, Pycnanthemum linifolium, Monarda fistulosa, Chenopodium Boscianum, Muhlenbergia sylvatica . Issued March S, 1898. Transactions of The Academy of Science of St. Louis. VOL. VIII. No. 5. THE MOLLUSCAN FAUNA OF WESTERN NEW YORK. FRANK COLLINS BAKER. Issued March 22, 1898. THE MOLLUSCAN FAUNA OF WESTERN NEW YORK.* Frank Collins Baker. The mollusks enumerated in the following pages were col- lected by the writer while on his vacation from July 5th to July 29th, 1897. During that time about seventy-five species were obtained, represented by over 1,500 specimens. To make the list of more value the writer has incorporated sev- eral local lists with his own,f and the paper may be con- sidered, therefore, a pretty complete catalogue of the land and fresh water shells of the western part of the State of New York. Species not found by the writer are marked by an asterisk. Local lists are seldom accompanied by a list of the exact stations (and also a somewhat detailed description of the same) and the writer believes that the present contribution will be of some value on this account, as great care has been taken in examining and recording every locality. The western part of New York is an especially good region for animals of this class; there are a large number of streams in which the fresh water forms thrive, and moist woodlands * Presented in abstract before The Academy of Science of St. Louis, Oct. 18, 1897. t The following publications have been consulted and species therein recorded added to the present list: — Walton, John. The Mollusca of Monroe County, N. Y. Proc. Roch. Acad. Sci. 2:3-18. Banks, Nathan. The Land Mollusca of the Cayuga Valley. The Nautilus 512: 137-9. 1892. Dewey, Prof. C. List of Naiades (clams), found in Western New York, and sent to the State Collection at Albany, with some chiefly from Ohio . Ninth Annual Report of the Regents of the University of the State of New York, 1856, pp. 31-38. These are all the papers relating to the Mollusca of Western New York, which the writer has been able to find. There are a number of good lists of Mollusca from Eastern New York, besides those of the whole State,, several of them by the veteran collector, Dr. James Lewis. (71) 72 Trans. Acad. Sci. of St. Louis. are very plentiful. Limestone is abundant in various places and of course teuds to produce a snail fauna of large size. The Erie Canal is one of the best collecting localities in the State, and more species can be found in it than in any other one body of water in the region. In the spring, when it is drained, the collector may don his rubber boots and literally " wade in " after these interesting creatures. At such a time a large number of species of Unionidae may be obtained, for the canal is the metropolis of this group in the western part of the State. The writer is under obligations to the following persons for assistance of various kinds : — Messrs. Harry S. Hall and Herbert S. Harris for valuable aid in collecting; Mr. John Walton for information regarding Monroe County mollusks ; Dr. Howard N. Lyon for a collec- tion of shells from Cayuga County ; and Prof. R. Ellsworth Call, Mr. C. W. Johnson and Prof. Henry A. Pilsbry for assistance in determining various species. The list of stations at which collections were made is as follows: — Station 1. Pinnacle Hill, Southeast Rochester. An esker from 100 to 200 feet above the plain upon which Rochester is built. The high hill called the Pinnacle is very steep, having a slope of from 10 to 30 degrees. The summit is well wooded, the trees extending well down the sides in a ravine, which separates the main hill from the little spur known locally as " round the world." Dead logs and leaves are very abundant, and while the soil is rather sandy in some spots, yet where the vegetation has had a chance to decay, there has been formed a rich loam, in which the mollusks love to bury themselves. Oak and maple trees predominate. July 8. Thermometer 90° in shade, collections made in the morning. Weather fine.* Station 2. Residence of Mr. John Hall, 16 Boardman * The data for the Pinnacle and Cobb's Hills are taken from Mr. Warren Upham's paper, Eskers near Rochester, N. Y., in Proc. Roch. Acad. Sci. 2: 181-200. 1892. Baker — The Molluscan Fauna of Western New York. 73 avenue, Rochester. Yard in rear of house where various species of Limax are to be found under boards and particu- larly under the garbage barrels. Station 3. Wide Waters, Erie Canal, East Rochester. Shore lined with weeds and debris of various kinds; shells sticking to stones, sticks and water weeds. July 12. Collec- tion made in the afternoon ; temperature 65° to 70°; rainy. Station 4. Genesee River Gorge, below lower falls. Banks 225 feet high, made up of the Niagara and Clinton geological formations, overlaid by about 10 feet of the Drift. Banks heavily wooded with large trees and low bushes; soil rich, mixed with pieces of decomposing limestone and shale. River very rapid ; shore lined with rocks. The river is now being contaminated by sewage from the city. July 13. Tempera- ture 60° at 8 a. m. and 85° at noon, when the clouds of the morning disappeared. Mollusks very plentiful. Station 5. Irondequoit Bay. This body of water is an inlet from Lake Ontario, and is five miles long and from one-half to one mile in width. The rolling hills about it are from 100 to 200 feet in height and are, in most cases, heavily wooded. The bay is about 80 feet deep (maximum). The shores are lined with cat-tails ( Typha) and sweet flag (Acorus). There are numerous small streams flowing into the bay and also many small bays leading off from the large bay. Irondequoit creek flows into it at its southern end. July 14. Cloudy, with occasional rains. Temperature varying from 75° to 80°. The southern part of the bay was literally covered with pond lilies (N~ymphaea), among which many mollusks were found. The shores were everywhere covered with dead beach-shells. Station 6. Sea Breeze, on Lake Ontario, at entrance to Irondequoit Bay. Sandy beach with ridge of sea-wrack, in which dead shells were plentiful. July 16. Station 7. Sea Breeze; ravine near boat house on Ironde- quoit Bay. Heavily wooded and the ground covered with dead leaves, sticks and other debris. Ground moist. July 16. Weather fine, temperature about 80°. Station 8. Locks in Erie Canal near Monroe Avenue bridge. Specimens collected from side of locks as the water flowed in to lift the canal boat. 74 Trans Acad. Sci. of St. Louis. Section 9. Charlotte and Summerville, Lake Ontario. These two places are situated on opposite sides of the mouth of the Genesee River, which is extended far into the lake by means of two long piers or docks. The water is shallow for a considerable distance and shells are brought by thousands over this area and deposited in long ridges of sea- wrack along the beach. This was an excellent place to find a large number of species. July 21. Weather fine and warm. Station 10. South Park, Genesee River, near bridge. Water shallow, shore lined with rocks. Bottom sandy and muddy. Weather very fine; temperature 80°. Mollusks very common. July 22. Station 11. New York State Fish Hatcheries, Caledonia. Grove with dead and fallen logs and numerous stones. Mol- lusks plentiful under boards. Stream running from fish tanks in which both Limnaea and Planorbis were found. July 23. Occasional rains. Temperature about 80°. Station 12. Creek near Kelly's Grove. Clear and rapid. Stony bottom. Very few mollusks. July 23. Station 13. Cobb's Hill. A part of the esker known as the Pinnacle Hills, from 75 to 125 feet in height. Heavily wooded on north side where there are several ravines. Oak and elm trees predominate. Ground moist and covered with the usual forest debris. July 27. Very rainy. Temperature 70° with a very strong wind. Station 14. Field near canal just out of Rochester. Pupa and small Limax found under chips and sticks, and about old stumps. Rainy and rather cool. Station 15. Canal below lock no. 65. Shore lined with flags and other water plants. July 28. Rainy. Station 16. Long Point, Conesus Lake, in ravine. July 1, 1891. Collected by Miss Grace M. Hall. Station 17. Owasco Lake, Cayuga County. Collected by Dr. Howard N. Lyon. Station 18. Medina. Station 19. Canandaigua Lake. The last two stations are represented only by specimens in the collection of the Chicago Academy of Sciences. Baker — The Molluscan Fauna of Western New York. 75 In the following list the writer has deviated from the beaten path which has been trodden by conchologists in the past, but he believes that the recent conclusions of well-known students should be recognized, and adopted, wherever sufficient evi- dence is shown of their validity. Thus in the Unionidae, the writer has followed Mr. Charles T. Simpson ; in the Pulmonata Mr. Henry A. Pilsbry ; and in the Viviparidae Prof. R. Ells- worth Call. References to the publications of these gentle- men will be found elsewhere. Class PELECYPODA. Order PRIONODESMACEA. Suborder Naiadacea. Family Unionidae.* Genus Anodonta (Bruguiere em.) Lamarck. 1. Anodonta Lewisii Lea. Station 10. A single small specimen collected. *2. Anodonta salmonia Lea. Erie Canal. 3. Anodonta implicata Say. Charlotte. *4. Anodonta imbecilis Say. Irondequoit Creek. *5. Anodonta fragilis Lam. Pittsford. *<3. Anodonta fluviatilis Dillw. Allen's Creek. *7. Anodonta benedictii Lea. Erie Canal. *8. Anodonta excurvata DeKay. Pittsford. * Genus Alasmodonta Say. 1819. t 9. Alasmodonta marginata Say. Station 10. Does not seem to be common in the Genesee River. Erie Canal (Walton's list). * See Simpson, Proc. U. S. National Museum 18 : 295-343. 1895. t The present classification of the Unionidae is that of Mr. Charles T. Simpson, which will be published in the writer's report on The Mollusks of Chicago and Vicinity, now in press. 76 Trans. Acad. Sci. of St. Louis. 10. Alasmodonta deltoidea Lea. Erie Canal. 11. Alasmodonta rugosa Barnes. Station 10. Typical and of good size. 12. Alasmodonta pressa Lea. Station 10. This species is said to be common in the upper Genesee River, but the writer obtained only one specimen and that was both a deformed and pathologic individual of small size (length 49; breadth 18; height 29 mill.). The whole shell is much shorter and more stumpy than the typical form and there is a a. pressa lea. peculiar bulge at the posterior pathologic specimen. en(j. xhe teeth are very long and very thin. The accompanying cut is an outline figure. *13. Alasmodonta undulata Say. (Walton's list). Genus Strophitus Rafinesque. 1820. 14. Strophitus edentulus Say. Unio femiginea Lea, Walton's list. Station 12. One specimen found. (Long Pond and Gen- esee River, Walton's list.) Genus Unio Retzius. 1788. *15. Unio gibbosus Barnes. Erie Canal. Genus Anodontopsis Simpson. 1898. *16. Anodontopsis subcylindraceus Lea. Irondequoit Creek. *17. Anodontopsis ferussacianus Lea. Erie Canal. Genus Quadrula (Rafinesque) Agassiz. 1852. 18. Quadrula rubiginosa Lea. Station 10. Common, and showing no little variation in relative lengths and heights. Baker — The Molluscan Fauna of Western New York. 77 19. Quadrula plicata Lesueur. Erie Canal. 20. Quadrula undulata Barnes. Erie Canal. Genus Lampsilis Rafinesque. 1820. *21. Lampsilis gracilis Barnes. Erie Canal. 22. Lampsilis iris Lea. Unio novi-eboraci Lea. Station 10. Erie Canal (Walton). This species is very com- mon and of large size at this locality. The variation in the size and character of the rays is very great, some being per- fectly straight while others are made up of square or oblong dashes, making the rays broken ; the color varies from light yellow to grass green. One specimen measures 76 mill, in length. At one spot in the river these mollusks were found piled up in a great heap among the rocks, showing where a muskrat had partaken of a goodly meal. *23. Lampsilis nasutus Say. Erie Canal. 24. Lampsilis luteolus Lamarck. Stations 9, 10, 18 and 19. As is usual in this species, the specimens before the writer show a great range of variation. The individuals from station 10 are very solid, and are cov- ered with a heavy deposit which renders them difficult to clean, besides making them very unsightly. At this locality they live in water from six inches to a couple of feet in depth, buried in the soft mud, their siphons and the extreme poste- rior part of their shell being the only parts of them visible. In some spots they are so close together that their shells almost touch. The sexes are very distinct in specimens from this locality, much more so than those from the canal. The shells from stations 9, 18 and 19 are smooth, polished and delicately rayed. A single specimen from station 18 has dark green rays over a yellowish-green background. The entire collec- tion shows a surface varying from perfectly plain to very strongly rayed, and the rays vary from thread-like to very wide. *25. Lampsilis ventricosus Barnes. Erie Canal. Unio occidens Lea. Trans. Acad. Sci. of St. Louis. 26. Lampsilis radiatus Lamarck. Station 18. Genesee River (Walton). Does not seem to be common. *27. Lampsilis rectus Lamarck. Erie Canal. *28. Lampsilis rosace us DeKay. Of this species Mr. Walton says: " Unio rosaceus is quite local. I have found it only at Long Pond and the margin of Lake Ontario, between Long Pond and Charlotte. DeKay's figures are very characteristic of all I have collected." 29. Lampsilis alatus Say. Station 10. Also one small specimen from Niagara Falls (Acad. coll.). Erie Canal (Walton). The specimens from the Genesee River are very large and fine, with a deep purple interior. *30. Lampsilis cariosus Say. Erie Canal. 31. Lampsilis complanatus Solander. Station 10. The specimens from this locality are very thick and heavy. 32. Unio tappanianus Lea. Station 10. Apparently not common. Order TELEODESMACEA. Suborder Veneracea. Family Corbiculidae. Genus Sphaerium Scopoli. 33. Sphaerium simile Say. Stations 5, 9, 10 and 17. Erie Canal (Walton). Very com- mon and grows to a large size, particularly at station 10, where it may be found in great abundance along the shore in the soft mud. 34. Sphaerium stamineum Conrad. Stations 9 and 10. Very common. Baker — The Molluscan Fauna of Western New York. 79 35. Sphaerium trans versum Say. Stations 4 and 10. Very common. At station 4 it may be found very abundantly just below Brewer's dock, where the specimens lie buried by thousands in the soft mud. 36. Sphaerium fabile Prime. Stations 6 and 9. Very common, particularly in the drift along the shore of Lake Ontario. *37. Sphaerium partumeum Say. Genesee River. Numbers 34, 35 and 36 are not listed by Walton. 38. Sphaerium rhomboideum Say. Station 5. Apparently not common. Genus Pisidium Pfeiffer. *39. Pisidium variabile Prime. Brighton. 40. Pisidium abditum Haldeman. Station 6. Apparently very common, especially in the drift along shore. 41. Pisidium bakeri Pilsbry.* Station 9. Very common. Class GASTROPODA. Subclass Anisopleura. Superorder Euthyneura. Order PULMONATA. Suborder Stylommatophora. Superfamily Agnatha. Family Circinariidae. Genus Circinaria Beck, 1837. Selenites Fisher. 1878. (Non Hope. 1840.) 42. Circinaria concava Say. Stations 10 and 11. Fairly common. * This is a manuscript name, which is shortly to be published by Prof. H. A. Pilsbry. 80 Trans. Acad. Sci. of St. Louis. Superfamily Aulacopoda. Family Zonitidae. Genus Vitrea Fitzinger. Zonites and Hyalina of authors. 43. Vitrea arborea Say. Stations 1, 5, 11 and 13. Pittsford (Walton). Common everywhere and with little or no variation. On the south slope of station 13 it is exceedingly abundant, just back of the Cobb house. 44. Vitrea indentata Say. Stations 1 and 13. Apparently not common. 45. Vitrea radiatula Alder. Station 5. Not a3 common as arborea. Specimens from this locality are large and very fine. t46. Vitrea binneyana Morse.* Genus Omphalina Rafinesque. 47. Omphalina fuliginosa Griff. Stations 1 and 5. Cayuga Lake Valley (Banks). This species is found in great abundance at station 1, under dead leaves and old logs; frequently found buried in the earth. The specimens are very dark colored and highly polished. 48. Omphalina inornata Say. Station 16. Pittsford (Walton). Cayuga (Banks). Ap- parently not common. The single specimen collected is very light colored. f49. Omphalina laevigata Pfr. |50. Omphalina friabilis Binney. * Names marked with a dagger are taken from Mr. Banks' paper in The Nautilus 5: 137. This is a list of The Land Mollusca of the Cayuga Valley. Baker — The Molluscan Fauna of Western New York. 81 Genus Gastrodonta Albers.* f51. Gastrodonta ligera Say. 52. Gastrodonta intertexta Binney. Station 1. Pittsford (Walton). Cayuga (Banks). "Rather common, under fallen trees and debris of various kinds. 53. Gastrodonta suppresa Say. Station 5. Pittsford (Walton). Not common. *|54. Gastrodonta multidentata Binney. Rochester and Cayuga. *f55. Gastrodonta nitida Miiller. Stations 5 and 7. Monroe County (Walton) and Cayuga (Banks). Very common near the boat house at Station 5. *|56. Gastrodonta minuscula Binney. *57. Gastrodonta limatula Ward. f58. Gastrodonta milium Morse. Genus Conulus Fitzinger. *f59. Conulus fulvus Drap. Pittsford ( Walton) ; Cayuga (Banks). Family Limacidae. Genus Limax Linne. 60. Limax maximus Linne. 61. Limax flavus Linne. Station 2. East Rochester (Walton); Cayuga (Banks). Very common and of large size. Two specimens of maximus measured 150 and 165 mill, in length; and two of flavus measured 91 and 130 mill, in length. 62. Limax agrestis Linne. Stations 2 and 14. East Rochester (Walton). * For a consideration of these genera, see Pilsbry, Proc. Phil. Acad. Sci. 1894: 11-31. 82 Trans. Acad. Sci. of St. Louis. 63. Limax campestris Binney. Stations 1, 2, 3, 5, 11, 13 and 17. Found commonly everywhere. Familv Philomycidae. Genus Philomycus Rafinesque. *f64. Philomycus carolinensis Bosc. *65. Philomycus dorsalis Binney. Family Vitrinidae. Genus Vitrina Drap. *66. Vitrina limpida Gould. Family Endodonttdae . Genus Pyramidula Fitzinger. Patula Held. 67. Pyramidula alternata Say. Stations 1, 4, 5, 7, 10, 11, 13 and 17. Rochester (Wal- ton); Cayuga (Banks). This is the most common land snail, and shows the usual variation in height of spire, coloration and coarseness of ribbing. 68. Pyramidula perspectiva Say. Station 13. Pittsford (Walton); Cayuga (Banks). But one specimen was found, which measured 10 mill, in diameter. 69. Pyramidula striatell* Anthony. Stations 1, 7, 11 and 13. Pittsford (Walton); Cayuga (Banks). Common everywhere. Section Helicodiscus Morse. 70. Pyramidula lineata Say. Station 5. Rochester (Walton); Cayuga (Banks). Com- mon. Baker — The Molluscan Fauna of Western New York. 83 Superfamily Holopoda. Family Helicidae. Genus Vallonia Eisso. 71. Vallonia pulchella Muller. Station 4. Pittsford (Walton) ; Cayuga (Banks). Common everywhere. 72. Vallonia costata Muller. Station 21. Common and strongly ribbed. Genus Polygyra (Say). 1818 ; Pilsbry. 1889. Section Mesodon Rafinesque. 73. Polygyra albolabris Say. Stations 1, 5, 7, 13, 16 and 17. Rochester (Walton); Cayuga (Banks). Very common. A form with a parietal tooth is found at Pittsford (var. denlata of Walton's list). 74. Polygyra thyroides Say. Stations 4, 11 and 18. Rochester (Walton); Cayuga (Banks). Common everywhere. The young of this and the preceding species are very diffi- cult, if not well-nigh impossible, to separate. 75. Polygyra sayii Binney. Station 1. Pittsford (Walton); Cayuga (Banks). Ap- parently quite rare. |76. Polygyra dentiferum Binney. Section Stenotrema Rafinesque. 77. Polygyra hirsuta Say. Station 7. Rochester (Walton) ; Cayuga (Banks). Com- mon. 78. Polygyra Monodon Rack. Plate X. fig. 6. Stations 1, 5 and 13. Rochester (Walton); Cayuga ( Banks). A single specimen, from the Pinnacle, has two 84 Trans. Acad. Sci. of St. Louis. teeth on the parietal wall and six old peristomes crowded together. The parietal teeth are long, thin and curved, and the whole shell is very much distorted. *78a. POLYGYRA MONODON FRATERNA Say. *79. Polygyra leaii Ward. Section Triodopsis Rafinesque. 80. Polygyra palliata Say. Station 16. Pittsford (Walton); Cayuga (Banks). Polygyra trident at a Say. Stations 1, 4, 5, 7, 13 and 17. Rochester (Walton) ; Cayuga (Banks). This species is very common, especially at Station 1, where it is found attached to the under side of logs and sticks. It is subject to very great variation and several varieties have been made ; whether there are additional varieties or whether those named are worthy of distinction is the question that confronts the student. In the present lot there are the following forms or varieties.* The novelties herein described are apparently fully adult specimens. 81. P. tridentata Say. Typical (Plate X. fig. 1). This form is characterized by two large peristome teeth and a small, almost straight parietal tooth which is directed between the two peristome teeth. 81a. P. tridentata juxtigens Pilsbry. (Plate X. fig. 3.) Station 1. Similar to the typical form, but the distance between the teeth on the peristome is considerably less and the parietal tooth is much larger, is curved and is directed toward the upper tooth on the peristome. 81b. P. tridentata (Plate X. fig. 4.) Station 1. In this form (which is a variety of the above) the teeth on the peristome are very small and are very wide apart ; the tooth on the parietal wall is like the last variety. * See Pilsbry, Proc. Phil. Acad. Sci. 1894: 11-31. Baker — The Molluscan Fauna of Western New York. 85 81c. P. TRIDENTATA BIDENTATA Vai*. T10V. (Plate X. fig. 2.) Station 1. Characterized by a perfectly smooth parietal wall and two teeth on the peristome, similar to those of typical iridentata. A half dozen specimens of this form have been seen. 81d. P. tridentata unident ata var. nov. (Plate X. fig. 5.) Station 1. In this form, of which a number of specimens have been seen, the peristome is perfectly plain and rounded, and the parietal tooth is small and points toward the upper part of the peristome. 81e. P. TRIDENTATA EDENTILABRIS Pilsbl'V. In this form (of which no specimens were found at this locality) there are no lip teeth.* *81f. P. TRIDENTATA FRAUDULENTA Pilsbiy. This is the fallax of most authors, but, according to Mr. Pilsbry, it is not the true fallax of Say, which is the intro- f evens of Bland. It is characterized by a very heavy parietal tooth, a large bifid upper tooth on the peristome and a much smaller tooth on the base of the latter. The teeth nearly fill up the aperture. f Genus Punctum Morse. f82. Punctum pygmaeum minutissimum Lea. Family Pupidae. Genus Strobilops Pilsbry. Strobila Morse. *f83. Strobilops labyrinthica Say. * See Pilsbry, The Nautilus, 712: 140. 1894. t The Tachea hortensis Miiller, recorded by Mr. Waltou, should not be included in the list of Western New York shells, because it was a stray specimen which had gotten away fromVick's greenhouse. 86 Trans. Acad. Sci. of St. Louis. Genus Pupa Drap.* Subgenus Papilla Leach. 84. Pupa muscorum Linne. Stations 5, 10 and 17. Brighton (Walton). Specimens from station 5 are very large and light colored ; those from station 10 are small, dark chestnut-colored and are found under chips near the base of the bank opposite Brewer's landing. Apparently very common. Subgenus Leucocheila Alb. and Mart. *85. Pupa fallax Say. Subgenus Columella Mart. *f86. Pupa edentula. P. simplex Gld. Irondequoit (Walton); Cayuga (Banks). Subgenus Bifidaria Sterki. *t87. Pupa corticaria Say. Pittsford (Walton). Section Albinula Sterki. *f88. Pupa armifera Say. *t89. Pupa contra cta Say. Station 11. Common. The specimens are smaller than the typical and approach P. holzingeri Sterki. The parietal tooth is somewhat smaller and does not enter as far into the aperture as in the typical form. I have not found this variety common. Section Bifidaria s. str. *f90. Pupa rupicola Say. Section Vertigopsis Ckll. Mss. 91. Pupa pentodon Say. Station 14. Rochester (Walton); Cayuga (Banks). Common. * See Sterki, The Nautilus 61: 2-8. 1892. Baker — The Molluscan Fauna of Western New York. 87 Subgenus Angustula Sterki. *t92. Pupa milium Gould. Genus Vertigo Miiller. *|93. Vertigo ovata Say. f94. Vertigo gouldii Binney. f95. Vertigo bollesiana Morse. Family Stenogyridae. Genus Ferussacia Kisso. Subgenus Cionella Jeffreys. 96. Ferussacia subcylindrica Linne. Stations 4, 5 and 11. East Rochester (Walton); Cayuga (Banks). Common. Family Succineidae. Genus Succinea Drap. *f97. Succinea obliqua Say. Pittsford (Walton). *97a. Succinea obliqua totteniana Lea. 98. Succinea avara Say. Station 10. Rochester (Walton); Cayuga (Banks). Com- mon. *99. Succinea aurea Lea. 100. Succinea ovalis Gould. Stations 3, 5, 10, 11, 15 and 17. Charlotte (Walton); Cayuga (Banks). Very common and variable in color, some specimens being of a rich golden hue. The specimens col- lected vary from very narrow to rather wide. The species is found in great numbers along the shores of lakes and rivers, either among the drift or crawling up the stems of the flags which line the shore. The latter case is typical of station 5. .88 Trans. Acad. Sci. of St. Louis. Suborder Basonimatopliora. Superfamily Akteophila. Family Auriculidae. Genus Carychium Miiller. 101. Carychium exiguum Say. Station 5. Rochester (Walton) ; Cayuga (Banks). Superfamily Hygrophila. Family Limnaeidae. Subfamily Limnaeinae. Genus Limnaea Lamarck. 102. Limnaea stagnalis Linne. Stations 3 and 15. Brighton (Walton). Very abundant and of large size ; found crawling over sticks and other debris in a few inches of water at station 3. There is considerable variation in the form of the aperture, some being almost round while others are long and narrow. The spires in these specimens, however, are of about the same length. Some individuals have fine spiral lines encircling the whorls. Section Radix Mont. *103. Limnaea columella Say. Section Limnophysa Fitz. *104. Limnaea reflexa Say. 105. Limnaea palustris Miiller. Stations 3, 5, G, 9, 10, 11 and 17. Erie Canal (Walton). Common and variable. The L. elodes Say of Walton's list is a synonym. 106. Limnaea catascopium Say. Stations 3, 4, 8 and 15. Erie Canal (Walton). Very com- mon and exceedingly variable. The typical catascopium is a Baker — The Molluscan Fauna of Western New York. 89 stumpy shell with the aperture longer than the spire. An- other form occurs which has a longer spire, as long as the aperture, and runs into L. expansa Hald., which latter does not seem to me to = L. palustris Miiller, but is rather a vari- ety of catascopium. A peculiar monstrosity was found by Master Hall at station 4, in which there was a large longitu- dinal ridge behind the peristome, the latter being much expanded, flaring, and the aperture twisted out of shape. Oalascojnum is found on stones and sticks in water from a few inches to several feet in depth. 107. Limnaea desidiosa Say. Stations 3, 5, 6, 9, 11 and 15. Monroe County (Walton). Found commonly on sticks and stones, and occasionally in Spirogyra. Considerable variation is found among the speci- mens collected, some being long and narrow, while others are short and stumpy. The aperture is sometimes a little longer than the spire. *108. Limnaea caperata Say. Pittsford (Walton). *109. Limnaea pallida Adams. Erie Canal (Walton). 110. Limnaea humilis Say. Station 11. Common and typical. Subfamily Planorbinae. Genus Planorbis Guettard. Section Planorbella Hald. 111. Planorbis campanulatus Say. Stations 5, 6 and 9. Pittsford (Walton). Very common. Section Helisoma Swainson. 112. Planorbis trivolvis Say. Stations 3, 4, 5, 6, 8, 10, 15 and 17. Charlotte (Walton). Common everywhere and extremely variable. Very frequently distorted. 90 Trans. Acad. Sci. of St. Louis. 113. Planorbis bicarinatus Say. Stations 5, 6, 9 and 17. Rochester (Walton). Very common everywhere and also subject to great deformity. Subgenus Gyraulus Agassiz. 114. Planorbis deflectus Say. Stations 5, 6 and 11. Brighton (Walton). Common and typical. *115. Planorbis hirsutus Gould. Mr. Walton reports rinding this species in the Erie Canal. 116. Planorbis parvus Say. Stations 5 and 6. Charlotte (Walton). Very common everywhere. Found in 8pirogyra. Genus Segmentina Fleming. Subgenus Planorbula Hald. 117. Segmentina armigera Say. Station 5. Brighton (Walton). Subfamily Ancylinae. Genus Ancylus Geoffroy. *118. Ancylus parallelus Haldeman. Charlotte. *119. Ancylus tardus Say. Irondequoit. *120. Ancylus rivularis Say. Genesee River. Family Physidae. Genus Physa Drap. 121. Physa heterostropha Say. Stations 3, 4, 5, 6, 8, 9, 10 and 11. Common everywhere and exceedingly variable. A recent study of this genus convinces the writer that of the seventy odd species described at the present time, there are really but ten or a dozen valid names. The following Baker — The Molluscan Fauna of Western New York. 91 will all prove to be synonyms of the present species, and there are several other names which will probably have to be added to this list: gyrina Say, elliptica Lea, aurea Lea, Sayii Tappan, Troostiana Lea, fontana Hald., Hildrethiana Lea, inflata Lea, cylindrica Newcomb, anatina Lea , Forsheyi Lea, Saffordii Lea, Haunii Lea, parva Lea, Nicklinii Lea, altonensis Lea, Febigerii Lea, Warrenana Lea, Halei Lea, Whilei Lea, Grosvenorii Lea, Zato Tryon, prime ana Tryon, oleacea Tryon, coniformis Tryon, deformis Currier and albofilala Ancey. The above seems like a tremendous syn- onymy, but it is probable that most, if not all, of these names will stand as they are written. 122. Physa ancillaria Say. Station 17. Rochester (Walton). This species is closely related to heterosotropha , and it may be but a variety of that species. It is always a wider and more robust form, and is distinctly shouldered, the latter being a character lacking in the former species. Genus Aplexa Fleming. *123. Aplexa hypnorum Linne. Pittsford. Superorder Streptoneura . Order CTENOBRANCHIATA. Suborder Ortliodonta. Superfamily Taenioglossa. Family Pleuroceridae. Genus Pleurocera Rafinesque. 124. Pleurocera subulare Lea. Stations 5 and 6. Common. Genus Elimia H. and A. Adams, 1854. Goniobasis Lea, 1862. *125. Elimia virginica Say. Erie Canal. 92 Trans. Acad. Sci. of St. Louis. *125a. Elimia virginica multilineata Say. Erie Canal. 126. Elimia livescens Menke. Station 10. Erie Canal (Walton). Very common in water from six to eighteen inches in depth, on or in a soft carbonaceous mud. Specimens are always covered with a deposit of algae in this locality. The specimens collected are veiy uniform. *126a. Elimia livescens depygis Say. Irondequoit. *127. Elimia semicarinata Say. Irondequoit. Family Bissoidae. Subfamily Bythiniidae. Genus Kythinia Gray. 128. Bythinia tentaculata Linne. (Plate X, figs. 7-10.) Stations 3, 4, 6, 9 and 15. Erie Canal (Walton). Very common everywhere. A number of specimens were obtained alive, and the writer is able to give a figure of the dentition from an American specimen. The central tooth is rather broader than long, very much lobed at either end, the ends of the lobes rounded. The usual rounded projection from the center of the lower surface is present. The reflection is short and wide and 7- cuspid, the center cusp large and rounded and the three side cusps short, triangular and very sharp. The denticles on the lateral lobes are 6-7 in number and are bluntly rounded. The intermediate tooth is squarish, produced at the lower outer corner and 7-dentate, the third cusp from the left side being large and roundly pointed, and the side cusps, two on the left and four on the right, are small and acutely triangu- lar. The lateral teeth are long and narrow and rounded at their lower extremity. The reflections are short and wide, the first 12 and the second 16 cuspid. The writer was not, unfortunately, able to count the number of rows of teeth. The writer regrets that he does not have access to European publications, in which the radulaof this species is figured, for comparison. Baker — The Molluscan Fauna of Western New York. 93 Family Amnicolidae. Genus Amnicola G. & H. 129. Amnicola limosa Say. Stations 6 and 9. Genesee River (Walton). Common. *130. Amnicola porata Say. Erie Canal. *131. Amnicola orbiculata Lea. *132. Amnicola pallida Hald. *133. Amnicola granum Say. 134. Amnicola lustrica Pilsbry. Station 5. Rare. Genus Cincinnatia Pilsbry. 135. Cincinnatia obtusa Lea. Stations 6 and 9. Common. 136. Cincinnatia cincinnatiensis Anthony. Station 9. Only a single specimen found. Genus Gillia Stimpson. 137. Gillia altilis Lea. Station 9. Erie Canal (Walton). Rather common. Genus Somatogyrus Gill. 138. Somatogyrus subglobosus Say. Station 9. Common. Family Valvatidae. Genus Valvata Miiller. 139. Valvata sincera Say. Stations 5, 6 and 9. Erie Caual (Walton). Very common, and subject to considerable variation. 140. Valvata tricarinata Say. Stations 3, 5 and 9. Very common and exceedingly vari- able, being either uni-, bi- or tricarinate. 94 Trans. Acad. Sci. of St. Louis. 141. Valvata obtusa Drap. Station 9. Very common. This is the first record of the occurrence of this species in this country. It is a common European species. Subgenus Lyogyrus Gill. *142. Valvata pupoidea Gould. Mr. Walton reports having found this species near Roches- ter. Family Viviparidae. Genus Vivipara Lamarck. *143. Vivipara contectoides W. G. Binney. Mr. Walton reports finding a large number of specimens of this species in the Erie Canal near Pittsford. This is the only record of this species being found in this part of New York. Genus Campeloma Rafinesque. *144. Campeloma ponderosum Say. Erie Canal. 145, Campeloma decisum Say. Station 10. Erie Canal (Walton). Very common, buried in soft mud. *146. Campeloma geniculum Conrad. Erie Canal. 147. Campeloma rufum Haldeman. Station 10. Very common, associated with decisum. Erie Canal (WTalton). *148. Campeloma integrum DeKay. Erie Canal. *149. Campeloma obesum Lewis. Erie Canal. explanation of illustrations. Plate X. 1, Polygyra tridentata Say. Pilsbry, Proc. Phil. Acad. Sci. 189-1. pi. 1. f. 7. — 2, P. tridentata bidentata Baker. Original. — 3, P. tridentata juxtig ens Pilsbry, Proc. Phil. Acad. Sci. 1894. pl.l.f. 8.— i, P. tridentata Original.— 5, P. tridentata unidentata Baker. Original. — G, P. monodon Rack., double parietal tooth. Original.— 7-10, Bythinia tentaculata Linnd. 7, Central tooth.— 8, Intermediate tooth.— 9, First lateral.— 10, Second lateral. All original. '-.' Issued March 22, 189S. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate X. POLYGYRA AND BYTHINIA. Transactions of The Academy of Science of St. Louis. VOL. VIII. No. 6. THE EFFICIENCY OF GEAKING UNDER FRICTION. CALVIN M. WOODWARD. Issued May 10, 1898. THE EFFICIENCY OF GEARING UNDER FRICTION. Calvin M. Woodward. 1. The energy lost through the friction of gears, is often ignored, and none of the treatises on Applied Mathematics give a satisfactory treatment of the subject. Moseley's ele- gant discussion is a little involved; he deals too much with conditions " behind the line of centers;" he fails to give exact results in a finite form ; he furnishes no convenient formulae for showing; at a glance the effect of the size of the rolling circle in the case of epicycloidal teeth; and he gives no con- venient material for comparing the efficiency of epicycloidal teeth with the efficiency of involute teeth. The following discussion was in substance given to my class nearly two years ago ; I have however recently reduced my equations to such form that the efficiencies of the two kinds of teeth can readily be compared without the trouble of numerical examples. / / n, \ / Figure 1, Showing two teeth in action during the "Approach." 2. General Formulae for all Teeth. Let Fig. 1 represent the geometrical conditions in a plane perpendicular to a pair of parallel axes. C1 and C2 are the centers of two wheels, /, their pitch point, i\ and r2 their radii, and T the point of con- tact of two teeth. JTis the common normal to the teeth in (95) 96 Trans. Acad. Sci. of St. Louis. action, and 6 the angle the normal makes with the line of centers. The arrows show the directions of rotations, Cx being the driver. "While the friction is not a function of the velocity, it is convenient to employ their angular velocities ax and a2. During the" approach " to the pitch point the two teeth are sliding towards each other, the velocity of sliding being measured by the familiar expression r (a1Jr a2), r being the distance of the point of contact from the pitch point. Let- ting the co-efficient of friction be f, it is evident that the line of resultant action between the teeth makes an angle

2. The magnitude of that action we will call P : so that the drivino; moment is PL = 31, : and the moment transmitted to the follower is M2 = Pl2. From the notation shown in Fig. 1, it is easily seen that l1 = rl sin (0 — „2 I" Q 2n*e2 3 n*e6 + 12 n*e 1 + 15 n*e5 + 2+ 15k2 + 15&4 7T6 -676 + &C- 90 n^e" The second fraction treated in the same way gives ir2 k'Tr3 1 + Zk'2 tt* k'(2 + 3k'2) tt5 2n12e2 3/i^e3 + 12 nfe 4*4 15 7i15e5 + &c 7T" Adding these and withdrawing — j-2 from the brackets, we get *-£+y? n^e" A;' — & 7r 7r2 ""3" ^e + 6nx2e2 + &2 + /fc'2 tt2 2(k' — k) TT3 4 ?i ,2e2 15 ?i13e3 &c. but k' — k = 2eft k2 + k2 = 2 (I + e2)/2, hence i? \Wj w,/ 1 W 2 271/ 3nn + ^ 7T< 6n2e2 + 7T 2/2 [f 2n2e2 2/2 + *2/ 2n2 4tt3/ 15/i 3e2 &c. (12). 9. In formula (12) the values of n1 and n2 are to be found 2r as explained in § 7. The value of e = — h-!,'Hi+£) (14). 12. If as is usual we let ql = q2 — q, the total energy lost becomes t7l+D, = Ut(±.+ ±;) 2q- Alog(l + 2) - A'log(l+|,) Woodward — The Efficiency of Gearing under Friction. 103 The energy exerted by the driver during this action is 17 = ** and the ratio of energy lost to energy exerted is p - % + U> _ ( 1 , l\r, U [r^ r2 ) 2q 22-/Uog(l+!)-A'log(l + f,) (15), Let Wj and n2 represent the numbers of teeth in the driver and follower. Then, but one pair of teeth being in action at a time, irr, q = — K When the obliquity is as small as possible, max. r = q sin 6 = r, cos 6 hence tan 6 = - = -» and q 7T , sin Id — = ^2/ , *-3/2 &c ^1 2ni (?i! — ^Z) 3?ll (ni — ""/ )2 ''/toK^l + V W, + 7T/ / "l/ °l «> + T/ 7T 7T2/ 7T3/2 „ 2Wl(n1+7r/)^3n1(n1+7r/)2 w Substituting these values in (16) we get *-(-+-i£ + n* — w*p 3 (n^—Tr2/2)2 (17). 7T2/2 n* + dnfir2/* 2 (n2 — 7t2/2)3 — &c. 14. Formula (17) may be still more reduced by perform- ing the divisions indicated by the fractions in the brackets. Thus v =1 + ^y2, >x i 40 80 Fig. 1 Internal Characteristic of Shunt Dynamos with Scales Changed. 120 between the three variables — current, electromotive force, and speed — the characteristic surface of the machine and the algebraic expression the characteristic equation. * Proc Inst. Mech. Eng., p. 238. 1879. t La Lumiere Eltctrique Dec. 3, 1881. Kinsley — Discussion of Series Dynamo-Electric Machines. 109 An examination of the characteristic of a dynamo, such as is given in fig. 1,* suggests that it may be a part of an hyper- bola. (See also figs. 10 and 12.) Since the total induced electromotive-force is directly proportional to the speed when the current is constant, see figs. 9 and 11, we can unite these two conditions and obtain the characteristic surface of the series dynamo. In fact fig. 2 shows how the surface may be built up by connecting together a large number of the straight lines. Current C" 0' G Fig. 2 Series Dynamo Surface. The line DB of fig. 2 is the characteristic of the machine at the speed AD. The lines CB, C'B', etc., give the rela- tions between E. M. F. and speed when the current has the constant values AC, AC, etc. In this way the characteristic surface can be defined by the actual tests of the d}^namo or motor. * The curve shown is the same that was gotten from a Weston dynamo using the armature and fields in series. The scales bave been changed for eonvenience. 110 Trans. Acad. Sci. of St. Louis. If we can obtain an algebraic expression for this relation it will be possible to use the surface for many engineering purposes. Dr. O. Frolich * started this investigation which can be profitably extended f to include many vexing problems. The actual tests of dynamo and motor are given in the second part of this paper, but in this part a more simple condition is used to illustrate the method of examination. Symbols . e = electro-motive force impressed on the terminals. E = total electro-motive force induced in the armature of the dynamo. Ea = counter e. m. f. of the motor, i = current. n1 = rev. pr. sec. n = rev. pr. min. of the dynamo, n' = rev. pr. min. of the motor. H = magnetic density in the air gap. «j = max. value of permeability of the magnetic circuit. a — a factor depending upon the saturation. A = area of the armature coil. S = number of turns of wire on the magnates. G = a constant depending upon the design of the machine. R = total resistance of the circuit, r = resistance between the terminals. The total induced e. m. f. in the armature conductors depends directly on the speed and strength of the field through which the conductors pass — see eq. (l)p. 111. The strength of the field in the air gap has a very complex relation to the other conditions of the magnetic circuit. This * Berl. Berichte 962, 1880. Electrotechinsche Zeitschrift, ii, 134, 170, 1881, vi, 128, 1885; ix Nov. 1888. t Prof. F. E. Nipher a good many years ago noticed the fact that the em- pirical equation of Frolich was that of an hyperbolic paraboloid. For some years he taught his students that this surface and its elements could be profitably used in examining the operation of dynamo-electric machines. This surface is therefore explained in la of this paper so that the methods suggested in la maybe employed in the remainder of the paper. Kinsley — Discussion of Series Dynamo -Electric Machines. Ill can be expressed by means of the empirical equation given on page 111 eq. (2). This is justified by the experimental work given in II. It will be found to be true in all the many cases where the demagnetizing effect of the armature remains nearly constant or relatively small. This condition is found when the brushes remain stationary and the change in arma- ture current changes but slightly the resultant field. The data given first were gotten from a machine which had an exceedingly strong field. Machines with relatively weaker fields have the constants a and b (p. Ill) apply to only small parts of the characteristic, see p. 127. The dynamo and motor can be considered together since in the case of the dynamo e = E — ri and for the motor e = Ea + ri the surface in which only the total induced electromotive forces are considered being similar in all respects. Assume for the present that the following equations are true. (1; E = 4An1H (2) H = GSi ^ 1 +ffSi From (2) it is seen that an increase of the magnetizing ampere turns will decrease the permeability of the circuit. from (2) and (1) n (3) E=4A«nGSi ,"■ 60 1 + aS\ from our assumptions we can let **£§& = a constant = a also S*. < 21 0 2( )0 1£ o i 1; 10 8 J -1 / / / / ■ 0 4 0 8 0 i 12 0 / / / / / / / / / Fig. 4 Series Dynamo Cli-aracteristic. Speed Equals 1400 rev.pr. min. The following four pages will be devoted to a discussion of the characteristic equation. The results obtained can be com- pared with those gotten from any available machine. Kinsley — Discussion of Series Dynamo-Electric Machines. 115 (4) E rtni 1 + b\ Impose the condition that n = a constant = 1400 It has already been found that a = 0.007 and b = 0.004 (8) E= 9'8i 1 + 0.004 i This is the equation of an hyperbolic curve having the asymptotes and E = -r- = 2450 when i = CO o i = _ = — 250 when E = co o See fig;. 4. laracteristic at a constant speed of n = 10 equation (9) E= 7'0i 1 + 0.004 i and can be plotted without difficulty. See fig. 6 JJ'. If we impose the condition that i = a constant = 120 E(l+bi) = ani from (4) (10) E= 0.567 n See fig. 5 BC when the constants are substituted. Equations (9) and (10) are the equations of the lines of intersection of planes, parallel to the co-ordinate planes, with the characteristic surface. These planes will limit the solid and bound the surface within the limits used in this particular case. 11G Trans. Acad. Sci. of St. Louis. This is shown and the method of plotting with the use of three axes is plainly evident from fig. 5. The surface used is thus bounded by the lines AD, DB, BCand CA. Fig. 5 Series Dynamo Surface.- The Condition of Power Equal to a Constant is Shown. If we wish the relation between the variables for the condition of constant power it may be gotten as follows: — Let Ei = a constant = d — 28400 (14) Ei = 28400 from equations (4) and (14) See fior 5 OP. n d db ai2 a\ (15) 4,060,000 16240 n = _-f — . — 1 2 See fig. 5 LM Kinsley — Discussion of Series Dynamo- Electric Machines. 117 n E2 bE ad a (16) n = 0.00505 E2 + 0.572 E See fig. 5 QS. Equations (14), (15) and (16) are the equations of the pro- jections upon the co-ordiuate planes of a line on the surface. See fig. 5 OT. These curves are decidedly interesting ones, but they will not be discussed at present. Fis:. 6 Series Dynamo Surface. Resistance Equal to 7 Ohms and Resistance Equal to 7.9 Ohms. It is important to know how speed, electromotive-force, and current will vary with reference to each other if the circuit is kept constant. Then R = a constant. Let R = 7 (11) E = Ri = 7 i 118 Trans. Acad. Sci. of St. Louis. from (4) and (11) E 6Ri n = - -\ a a (12) n = 1000 + 4.0 i n = — E a a (13) n = 1000 + 0.572 E See fig. 6 JF. See fig. 6 JK. o Equations (11), (12) and (13) are the equations of the projections upon the co-ordinate planes of a line on the surface which gives the simultaneous variation of E, i and u. It is instructive to note that projections on the En plane for different resistances are parallel, while the projections on the ni plane are not parallel, but vary with R as is shown by the co-efficient of i containing R in one case aud not in the other. It will be observed that from equations (12) and (13) the constants of the equation for the surface can be easily gotten, should it be experimentally easier to vary n than R. In the laboratory it is generally easy to vary R but frequently in practice n can be varied by throttling the steam, where a change of R is impracticable. See figs. 14 and 16. By the means outlined above it is a simple matter to fore- tell the result of any change in the conditions of operation. lb. Series Dynamo and Series Motor. The foregoing discussion of the series dynamo characteris- tic surface, as has been pointed out, applies equally to the series motor, when Ea is plotted instead of the E of the dynamo. As has been shown the counter electromotive-force — Ea — of the motor is analogous to the total electromotive-force of the dynamo. Kinsley — Discussion of Series Dynamo- Electric Machines. 119 It is occasionally desirable in actual practice to operate one series motor as the only load of a series dynamo. Usually then it is essential that the motor regulate for constant speed with widely varying loads. If it was desirable to run a motor whose characteristic was as given by DU of fig. 8, — data, p. 121 the necessary con- ditions could be gotten as follows: — The motor characteristic is drawn on the Ei plane of fig. 8, without reference to the speed axis. In this case the motor constants are a'=0.0067 6=0.0066 determined when the motor speed (n') was 1000 rev. pr. min. The motor characteristic equation is therefore (17) Ea = , , -,. and since 1 + b i (18) E = Es + Ri = r^L=I^ + Ri This can be arranged (18) a -g a' 4-g v J 1 + bi~~ n 1 + 6'i^n If the desired conditions that n', n, and R be constant, for varying loads and consequently varying current, are possible with the above dynamo and motor a plot of a A = x and 1 + b\ a 1 + M y will give a straight line. Through what range of current the above conditions are 120 Trans. Acad. Sci. of St. Louis. possible can be easily seen from the plot of x and y. See fig. 7. .0065 y\j .0050 / a 1+bi .0055 ^ .0050 / / ■^ I /o .0045 a' 1+b'i .0036 .0040 .0045 .0050 .0055 0060 E = E, = Fig. 7 Design of Series Power Circuit for Constant Speed of Macbines. Computed Data from Surface Constants. Total induced e. m. f. of dynamo at 1250 rev. pr. min, Total induced e. m. f. of dynamo at 1400 rev. pr. min, Total induced e. m. f. of motor at 1000 rev. pr. min. A = a .0070 1 + b\" l+.004i B = a' .0067 l+b"\ l+.0066i % = percentage variation of motive speed when dynamo sliced is 1250 r. p. m. and R = 2.25 Kinsley — Discussion of Series Dynamo-Electric Machines. 121 E Ea E, 20 181.5 118.0 167.0 40 338.0 212.0 302.0 60 474.0 288.0 423.0 80 594.0 351.0 530.0 100 7C0.0 404.0 625.0 120 795.0 447.0 710.0 Efl + Ri A B % 163.0 .00648 .00590 +2.5 302.0 .00605 .00530 0.0 423.0 .00565 .00480 0.0 531.0 .00530 .00439 —0.2 629.0 .00500 .00404 —0.5 717.0 .00473 .00372 —1.0 a a The values of . , » and , . , ,. given on page 121 are plotted. See fig. 7. As is seen a straight line averages the points very fairly from i = 110 to i = 30. Fig. 3 Series Dynamo Surface and Motor Characteristic. -Design of Power Circuit. From the line we find that (20) n n = .80 (21) R n = 0.00179 122 Trans. Acad. Sci. of St. Louis. If it is desired that (n') the speed of the motor be 1000 rev. pr. min. then 1000 10_n n = — 5— = 1250 and R=2.25 from eqs. (20) and (21). Another value of n' would require a different constant dynamo speed and a different line resistance. If on the other hand n were fixed then n' and R could be rotten from the above equations. This condition can be very nicely illustrated, or a graphical solution can be gotten by the use of the characteristic surfaces. From fig. 8. it is seen that the dynamo surface is ADBC. The motor characteristic at a constant speed of 1000 rev. pr. min. is the line DU. Project the motor characteristic upon the dynamo surface and it gives us the line VW, upon the surface which projected upon the ni plane gives the line VX. This line is not neces- sarily straight. See fig. 19. The line VX shows that the condition of constant speed of the motor would require varying speed of generator. If we add to the motor characteristic the e. m. f. = Ri, we would have as the result the line DZ which shows what the output of the dynamo must be to give the constant motor speed of 1000 r. p.m. when R = 2.25. DZ projected upon the surface gives us the conditions of operation of the dynamo in the Tine Y"Y. This line projected upon the ni plane gives the line Y"Y' which we see is a condition of a constant speed of 1250 rev. pr. min. of the dynamo for all currents. The line Y" is thus the characteristic of the dynamo at a speed of 1250 r. p. m. In the solution of the equations (20) and (21) if R had come out negative — see p. 136 — it would be impossible to use those two machines to regulate for constant speed with varying loads. From our knowledge of the necessary resistance of the circuit for best regulation, it will be possible to most success- fully design the circuit. Kinsley — Discussion of iSeries Dynamo- Electric Machines. 123 II a. Series Dynamo and Motor Tests. To what extent the foregoing discussion applies to the actual machine can be gotten from the following tests. Even if the equations did not apply the dynamo and motor surfaces could be established experimentally without any regard to their equations. 60 50 40 1=9.7 30 e 20 10 0 10 2C )0 4 6 rev. 8 pr. m: n. 10 12 00 20 • 30 40 Fig. 9 Series Dynamo . Test. Effect of Varying the Speed when the Resistance is Constant. The relation between speed and electromotive force was determined for the condition of constant current by twelve series with currents varying between 9.7 and 2.0 amp. using a Wood series dynamo. The speed was gotten by an automatic chronograph record. The line determined with a constant current of 9.7 amp. is given in fig. 9. When the machine 124 Trans. Acad. Sci. of St. Louis. has been stopped (i. e., n= 0) the current is maintained by an impressed e. m. f . which is equal to the intercept of the line on the e. m. f. axis.* The resistance of the machine is then this value divided by the current. 120 2 4 6 8 10 12 14 Fig. 10 Series Dynamo Test. E. M. Fs. Computed from those Observed. From the observed line e + ri 55.3 + ri 1.6 + ri nc 1159 c 454 c ri = 32.9 since i = 9.7 r = 3.39 • S. P. Thomson in " Dynamo-Electric Machinery, pp. 88 and 206, fifth edition, apparently disregards the fall of potential through the machine and so finds that the e. m. f. is not proportional to the speed. The correction is made by subtracting from the speed that gotten from the intercept on the n axis. He calls these rev. pr. mio. the " dead turns." Kinsley — Discussion of Series Dynamo- Electric Machines. 125 The resistance determined by the fall of potential method was r = 3.38 In the twelve cases observed. The resistance computed from the line with the machine in operation was r = 3.135 The resistance gotten by fall of potential method was r = 3.15 The difference is accounted for by a better brush contact while the machine is running. 1600 i=6.62 j° 1200 0 i=10.74>^€f r. p. m. n 800 y/6 400 o En=e— r i 20 Fig. 11 40 60 Series Motor Test. Effect of Varying the Speed when the Resistance is Constant. The fact that the lines are always straight shows that with widely varying speeds * the fall of potential through the ma- chine remains a constant. * S. P. Thomson, " Dynamo-Electric Machinery," p. 89, indorses the theory of " an apparent increase of resistance proportional to the speed," which was " first pointed out by Cabanellas in Comptes Eendus, Jan., 1882." In the above careful tests no evidence of a change of resistance can be found. 126 Trails. Acad. Sci. of St. Louis. This is also true for the motor as is seen from fig. 11. The poiuts do not lie as close to the line as in the case of the dynamo since a tachometer was used to determine the speeds. It will thus be evident that these straight lines will define the dynamo and motor surfaces. It will therefore be necessary to determine merely the characteristic of the machine. 60 O Ea=e-ri 40 Q-- 20 S6° fO i 10 12 2 4 6 8 Fig 12 Series Motor Test. Characteristic Determined with Speed Equal to 1500. 14 The motor surface was also determined. See fi^. 11 and fig. 12. The surface is shown in fig. 18 — the lines DB, C'B' and C"B" being the ones determined. ni' If we plot -g and i in order to get the constants of the surface in accordance with the method given on page 113, we have the lines ab, be, and cd of fig. 13. The changes in the direction of the line seem to indicate the points of saturation of different parts of the magnetic circuit. This is also seen to be even more pronounced in the case of the motor in lines ef and fo; of fio-. 13. For a limited range the constants of another machine were de- termined as is explained in connection with figs. 14, 15 and 16.* * The internal characteristic of a 1500 K. W. power generator was also determined, and it was found that the relations could be expressed in the above way with complete accuracy. Kinsley — Discussion of Series Dynamo-Electric Machines. 127 Throughout the range of current covered by a straight line the constants of the characteristic equation are absolutely correct. When the current varies outside of that range then the constants change. E« ynamo Current Fig. 13 Series Dynamo and Motor Tests. Determination of Constants when the Speed does not Change. The departure of the points from the straight lines show with what degree of accuracy the equation assumed fulfills the conditions found in the actual machine. It will therefore be possible in all machines to show by a surface the relations between E, i and n throughout the com- plete range of operation. Sometimes one and never more than three equations of the same form but with slightly different constants will give the algebraic relation between the variables under all conditions.* It has already been pointed out that the brushes must not be moved. 128 Trans. Acad. Sci. of St. Louis. lib. The Dynamo as a Speed Indicator. In a great deal of experimental work it is essential that we determine the instantaneous values of the speed of the appa- ratus. This can be very easily done by using a dynamo electric machine attached to the revolving shaft. 22.0 21.0 n 20.0 19.0 18.0 2.60 3.00 Fig. 14 3.50 4.00 Series Dynamo Test. Relation between the Currentand Speed when the Resistance was Constant. (1) If it is possible to use a separately excited dynamo with an electrostatic voltmeter connected to its brushes, then the speed is directly proportional to the potential shown by the voltmeter. Care must, of course, be taken to keep the exciting current constant. Equation (1) p. Ill gives this relation. E = 4 AHnj = nx multiplied by a constant. (2) This can also be more simply obtained by using a self- excited series dynamo and either an ammeter in the circuit or a voltmeter across the brushes. A very convenient machine for this purpose will be a Kinsley — Discussion of Series Dynamo-Electric Machines. 129 shunt dynamo, with a very strong field, with the outside cir- cuit left open and hence used as a series machine. A small Wood dynamo used in this way gave the results shown in the figures. The speed was gotten automatically by means of a chronograph. 3.00 3.50 4.00 Fig 15 Series Dynamo Test. Computation of Constants when the Speed does not Change. Figs. 14, 15, and 16 gave the data as taken. The exces- sively large scale makes the points appear somewhat scat- tered. The equation of the line of fig. 14 is 1 Ri = n, see p. 118 a a l When R = a constant = 12.23 ohms. - + 2.935 R- = 18.00. a a - + 3.935 R-= 22.00 a a a = 1.945 b = 0.635 from fig. 14 130 Trans. Acad. Sci. of St. Louis. The equation of the line of fig. 15 is 1 . n,i« b +1=: E 1 When n, = a constant = 22.2 See p. 111. t + 3.00= 1.496 o a -b + 4-00 1.821 I o a = 1.915 6 = 0.622 from fig. 15. These constants differ somewhat from those gotten with the condition of R = a constant. The dotted line of fig. 15 is one whose constants are as determined in the other two cases. The percentage varia- 36.0 38.0 40.0 42.0 44.0 46.0 48.0 Fig. 16 Series Dynamo Test. Relation between the E. M. F. and Speed when the Resistance was Constant. tion is found to be very small. If the constants a and b were determined with the condition of constant speed, and a Kinsley — Discussion of Series Dynamo-Electric Machines. 131 line was then platted showing the relation between n1 and i, the values of na determined from the line, from readings of i, would differ from the true \\ by not more than \%. The equation of the line of fig. 16 is - + - E = n, a a When R = a constant = 12.23 ? + 35.85-= 18.00 b a - 4- 48.10 - = 22.00 a a See p. 118. from fig. 16. a = 1.945 6 = 0.635 1.00 2.00 4.00 3.00 Fig. 17 Series Dynamo Surface. Effect of Varying the Speed when the Kesistance is Constant. It will thus be possible to determine this relation between Dj, i, and E by any one of the ways given with as great accuracy as the instrument may be read. 132 Trans. Acad. Sci. of St. Louis. The equation of the surface is therefore E= ^ E = 1 + 6i 1.945 n,i See p. 112. 1 + 0.635 i from figs. 14 and 16. *o This surface is given on fig. 17. The element of the sur- face wiving the condition of operation for constant resistance being indicated. The surface is explained in connection with figs, 5, 6, and 7. It follows therefore from the computations just given that experimentally the relation between nl5 i, and E may be de- termined in the most convenient way and then this relation may be used to obtain instantaneous values of the speed of the apparatus to which thi3 small machine is con- nected. This method has been used practically to determine the fluctuations of speed of an engine caused by sudden changes in its load. He. The Predetermination of the Operation of a Single Series Motor as the only Load of a Series Generator. This problem was first defined by Gisbert Kapp who sums up his observations in substance as follows — there can usu- ally be found some speed and some resistance at which the motor will operate at constant speed for all loads.* He has been successful in obtaining the result desired in the installa- tions he has made in England. Eric Gerard, f an eminent French authority, indorses Kapp's solution of the problem. C. E. L. Brown, while with the Oerlikon Engineering Co., of Switzerland, has also succeeded in operating motors installed * " Electrical Transmission of Energy," 4th edition, p. 199. t " Lec;ons sur L' Electricity," Tome second, p. 189. Kinsley — Discussion of Series Dynamo- Electric Machines. 133 in that way with no more than two per cent variation of speed. Heer von Debrowolsky has done equally well in Germany I do not know of any such installation in the United States but Dr. Louis Bell* has followed Kapp in his solution of the problem. It would of course be more satisfactory to know whether or not the desired results could be obtained before installing and to know also just how to design the circuit. The correctness of the method given under la for predeter- mining the operation of a series motor driven as the only load of a series dynamo was examined by means of a small four- pole Wagner Elect. Co.'s series motor and a two-pole Wood dynamo. 6 c" 9 C 12 15 Current Fig. 18 Series Motor Surface and Dynamo Characteristic. Predetermination of Motor Operation. The characteristic surface of the motor was determined — figs. 11 and 12 give the data and fig. 18 is the surface — and * "Electric Power Transmission," p. 86. 134 Trans. Acad. Sci. of St. Louis. then independently the characteristic of the dynamo was obtained — fig. 10. The line A'B, of figs. 10 and 18 will then give the counter e. ra. f. that the motor must develop. This line was there- o 1600 c 1400 ^ )\ o \ 1200 \ r. p. m. \ 1000 \ ° X \ 300 \ \o Current i 4 6 8 10 12 Fig. 19 Series Motor Test. Comparison of Observed and Computed Results. fore projected onto the motor surface and the line on the sur- face was projected onto the ni plane giving the relation between speed and current to be expected when the motor was driven by the generator. This is seen in fig. 18 and en- larged is the full line of fig. 19. The motor was then driven by the generator and the indicated points of fig. 19 give the result. Kinsley — Discussion of Series Dynamo-Electric Machines. 135 Another and much more extensive test was made, but the speed of the dynamo was too erratic to allow a smooth curve to be drawn; the average difference between the computed and observed values of the motor speed was only 0.05%. The resistance of the circuit that will cause the motor to operate at constant speed when the dynamo runs at 1300 r. p. m. can be determined by the method given on p. 119. From the lines of fig. 13 we can get the constants for par- ticular parts of the dynamo and motor surface and then can n' R solve for the ratios — and — for that range of current : Dynamo n n 1 .6 i - + l - = =r n a a & Speed constant = 1300 L + 10 - = 0.093 n a a - + 13 -- = 0.111 n a a Motor a = 0.0233 6 = 0.1815 I ,.b' i , — + 1 — = ^r 11 a a E„ Speed constant = 1500 i + 9-, = 0.170n' a a 1 7 ' + 13 °- = 0.203 n' a a a' = 0.00702 b' = 0.0878 We have then within the same range of current a' n' a R 1 + b'i n 1 + bi n 136 Trans. Acad. Set. of St. Louis. ' ~R 0.003745- =0.00828 when i = 10 n n 0.003280- = 0.00695 when i = 13 n n "' = 2.87 - = 0.00248 n n If we wish n = 1300 ^ = 3730 R = — 3.22 With these machines it will therefore never be possible to obtain a constant motor speed. The very close coincidence between the computed and observed value of speed shows that this method of investiga- tion can be accurately used. Issued May 20, 1898. Transactions of The Academy of Science of St. Louis. VOL. VIII. No. 8. we* S^_A /^t *^» yY r% THE NORMAL TO THE CONIC SECTION. EDMUND A. ENGLER. Issued July Id, 1898. THE NORMAL TO THE CONIC SECTION. Edmund A. Engler. The determination by the usual methods of analytic geom- etry of the normals to a conic section from a point not on the curve involves the solution of a cubic or bi-quadratic equa- tion by which the co-ordinates of the points at which the normals cross the curve may be found. This method, though of no special difficulty, is generally avoided on account of its tediousness, and in most elementary books the problem is considered solved when these equations are found. The graphical constructions given in this paper are accomplished without the algebraic solution of these equations and may prove to be a useful addition to courses in geometrical draw- ing, from which the solution of this problem is generally omitted. The analysis on which the constructions depend is based on well known principles of analytic geometry and is given in the most elementary and simple form. THE NORMAL TO THE PARABOLA. The equation of the parabola referred to its axis and the tangent at its vertex as axes of co-ordinates is, in the usual notation, y2 = 2px. ( 1) The equation of the normal to the parabola is py + %Vi — pVi + ^iVi (2) in which xv yx are the co-ordinates of the point in which the normal crosses the curve. (137) 138 Trans. Acad. Sci. of St. Louis. If the normal is to be drawn through any point in the plane, P, whose co-ordinates are ?, 77, these co-ordinates must satisfy equation (2), so that pn + Eyi = in/i + x&v (3) The co-ordinates xv yl also satisfy equation (2). If we now make xv yx variables, equation (3) becomes pv + £y = py + xv- W This equation is satisfied by the co-ordinates £ , y and also by the co-ordinates x11yl; and, therefore, represents a curve which passes through the given point P and through the intersections with the parabola of the normals through P. Equation (4) represents an equilateral hyperbola. The equations of its asymptotes are x = £ — p (5) and y = 0. (6) Equation (5) represents a straight line parallel to the axis of Y and at distance £ — p from it. The distance from the point P to this asymptote is evidently £ — (£ — p)=p; (7) that is, the asymptote is for all values of g at a distance p to the left of P. The asymptote represented by equation (6) is the axis of X. The auxiliary hyperbola can be readily constructed from the well known principle of constant areas, or perhaps more easily in one of the following ways: — Since the secant intercepts equal distances between the curve and each asymptote of the hyperbola, we have only to draw any secant, as QPL (Fig. 1), and make LK equal to PQ; the locus of K will be the auxiliary hyperbola. Again, it is evident from this principle that for any secant through P the triangles QRP and KML are equal ; and that Engler — The Normal to the Conic Section. 139 the bases on the axis of X of all such triangles as KML, no matter how the secant may be drawn, are of the same length, ML = PR = p. Therefore, to find a point on the hyperbola draw any secant through P ; from the point where this secant intersects the axis of X lay off a distance p to the M \L Figure 1. left. At this point draw a line parallel to the axis of Y; the intersection of this line with the secant is a point on the hyperbola. This hyperbola will intersect the given parabola in one, two, or three points according to the position of the point P in the plane. The actual construction of the hyperbola for any given case will give all the real normals possible for the given position of the point. 140 Trans. Acad. Sci. of St. Louis. When the point P is so situated that only one branch of the hyperbola cuts the parabola, there will be only one normal. When the point P is so situated that both branches of the hyperbola cut the parabola, there will be three normals. When the point P is so situated that one branch of the hyperbola cuts the parabola while the other branch of the hyperbola is tangent to it, there will be two normals. Both branches of the hyperbola cannot be tangent to the parabola, because the axis of X is one of the asymptotes of the hyperbola; for the same reason there must always be at least one real normal whatever the position of the point P. The following analysis shows in what way the number of normals possible is dependent upon the position of the point in the plane: — The equation of the tangent to the parabola may be written y=y'x+y'x' (8) in which x\ y' are the co-ordinates of the point of contact on the parabola. The equation of the tangent to the hyperbola may be written V = 5 — - — ,x—e — - — -,x' + y' (9) ? — p — x f — p — x J v/ in which x', y' are the co-ordinates of the point of contact on the hyperbola. If equations (8) and (9) are made to represent the same line, which signifies that the tangents to the two curves coin- cide, and if at the same time x\ y' in (8) and (9) represent the same point, which means that the curves are tangent to each other at that point, we have as equations of condition, y ? — p — x P/X' = — ^-JL }X' + y'i (11) y ? — p — x y ' v ' Engler — The Normal to the Conic Section. 141 from which for the co-ordinates of the point of contact, *' ~ 3 (£ — JP)> (12) y - 6 prj 2 ' s—p (13) If these values are substituted in equation (1), we have v ~ 21" P (14) This, if c, t] are regarded as variables, is the equation of the evolute of the parabola, which is therefore the locus of the point P when the hyperbola and parabola are tangent to each other. i 5 ^>^ 1 ^""v l ^^^ ; \ ^^ \ W^ / i Jr ' 1 ^\ ' \r *\ ' i\ / \\! / 1 %^ ~~ —" \ x """""■"^ \ N >\ \ NT N 1 \ \\ 1 \ \ \N \ \ 1 ^Vv ^ \ \ \ 1 ^^"fc^ ^ \ ^^^_ N ^*^^\ I V Figure 2. Since the evolute of any curve is the envelope of its nor- mals, the number of normals which can be drawn through a 142 Trans. Acad. Sci. of St. Louis. point on the evoluto is always one greater than the number which can be drawn through a point on its concave side and one less than the number which can be drawn through a point on its convex side ; that is to say, that for a point on the evolute, two of the theoretically possible normals coincide, Figure 3. and for a point on its concave side, two of them are imagi- nary, while for a point on its convex side, all of them are real. From this property and by reference to the figures, it is apparent that for any point on the evolute (Fig. 2) one branch of the hyperbola is tangent to the parabola, while the other branch cuts it in a single point, and there are two normals, PNV PJV2 ; for any point in the plane on the con- vex side of the evolute (Fig. 3), both branches of the hyper- bola cut the parabola, one in one point and the other in two Engler — The Normal to the Conic Section. 143 points, and there are three normals, PNV PN2, PJV3 ; and for any point in the plane on the concave side of the evolute (Fig. 4), only one branch of the hyperbola cuts the parabola and that in a single point, and there is but one normal, PN. Figure 4. Special Gases. 1. When the point P lies on the given parabola one of the normals through P coincides with the tangent to the hyperbola at that point and is found by joining with P a point on the axis of X at a distance p to the left of the foot of the ordinate through P; which is, in fact, the usual construction for this normal to the parabola based on the principle that the sub-normal is constant. The other normals can be obtained by the construction already given. 2. When the point P lies on the axis of the parabola, the auxiliary hyperbola degenerates into its asymptotes. The intersections of these asymptotes with the parabola give the normals. It will be observed that when the point P is at the 144 Trans. Acad. Sci. of St. Louis. cusp of the evolute, one of the asymptotes is tangent to the parabola at its vertex ; this is a critical case for which the three real normals coincide. The constructions for these cases can easily be supplied by the reader. THE NORMAL TO THE ELLIPSE. The equation of the ellipse referred to its rectangular axes as axes of co-ordinates is, in the usual notation, -+- = 1. (15) The equation of the normal to the ellipse is a2xyl — b2yx1 = (a2 — 62) x1yl (16) in which xx yx are the co-ordiuates of the point in which the normal crosses the curve. If the normal is to be drawn through any point in the plane, P, whose co-ordinates are £, rj, these co-ordinates must satisfy equation (16), so that a2syl — b2r)x1 = (a2 — b2) xxyv (17) The co-ordiuates x^ y1 also satisfy equation (16). If now we make x:, yl variables, equation (17) becomes a2zy — b2r\x = (a2 — b2) xy. (18) This equation is satisfied by the co-ordinates £, rj and also by the co-ordinates xv ylt and, therefore, represents a curve which passes through the given point P and intersects the given ellipse at the intersection of the normals through P with the ellipse. Equation (18) represents an equilateral hyperbola. The equations of its asymptotes are Engler — The Normal to the Conic Section. 145 a2 x = +-f-r2Z (19) a1 — bl 2/ = -n»- <20> aDcl /"~~aT^b2''' Equation (19) represents a straight line parallel to the axis a2 a2 of Y -and at a distance -* r-9c from it. As -s To is neces- al — b2 a1 — cr a? sarily positive, the expression -, j-2£ always has the same sign as £ ; therefore, the asymptote represented by equation (19) always lies on the same side of the axis of Y as the a2 point P. And as -j — t-2 is greater than unity, the point P lies between this asymptote and the axis of Y. The distance from the point P to this asymptote is a2 b2 ^-? = -~-rJ. (21) a2 — b2 ' a2 — b2 Equation (20) represents a straight line parallel to the axis b2 of Jl and at a distance = — ^ ri from it. a2 — er b2 . b2 As -= ^ is necessarily positive, the expression — -= — r^ »? v are regarded as variables, is the equation of the evolute of the ellipse, which is therefore the locus of the point P when the hyperbola and ellipse are tangent to each other. From the property of an evolute already cited (pp. 141-2) and by reference to the figures, it is apparent that for any point on the evolute (Fig. 6) one branch of the hyperbola is tangent to the ellipse while the other branch cuts it in two points and there are three normals, PN~y, PiV2, PJV3 ; for Engler — The Normal to the Conic Section. 149 Figure 6. any point in the plane on the convex side (or within) the evolute (Fig. 7), both branches of the hyperbola cut the ellipse, each in two points, and there are four normals, PiV^, PN2, PNS, PNl ; and for any point in the plane on the con- FlGURE 7. 150 Trans. Acad. Sci. of St. Louis. cave side (or outside of) the evolute (Fig. 8) only one branch of the hyperbola cuts the ellipse, but this one cuts it in two points, and there are two normals, PNV PJSf2. Figure 8. Special Cases. 1. When the point P lies on the ellipse one of the normals through P is the tangent to the auxiliary hyperbola at that point and is readily found by joining with P a point on the asymptote parallel to the axis of X at a dis- b2 tance ~s rr} q from the toot of the ordinate through P. a1 — o* a This furnishes a convenient method of drawing a normal to the ellipse at a point on the curve because no bisection of angles is required and corresponds to the method employed for the parabola by using the property that the sub-normal is constant. Engler — The Normal to the Conic Section. 151 2. When the given point P is on either axis of the ellipse, the auxiliary hyperbola degenerates into its asymptotes, one of which becomes the axis of the ellipse through the point, and the other is found by the construction. The intersec- tions of these two asymptotes with the ellipse give the normals. It will be observed that when the point P is at one of the cusps of the evolute, one of the asymptotes is tangent to the ellipse, which means that three of the normals coincide. 3. For the case when the ellipse becomes a circle, a = b, and equation (18) becomes fy-VX = 0, (31) which is the equation of a straight line through P and the center of the circle; the normals, therefore, coincide with this line. The constructions for these cases can easily be supplied by the reader. THE NORMAL TO THE HYPERBOLA. The equation of the hyperbola referred to its rectangular axes as axes of co-ordinates is, in the usual notation, ^2_^2 = 1. (32) a2 62 V ' The equation of the normal to the hyperbola is a2xy1 4- b2yxl = (a- 4- b2) xiyl (33) in which xx, yx are the co-ordinates of the point in which the normal crosses the curve. If the normal is to be drawn through any point in the plane P whose co-ordinates are c, y, these co-ordinates must satisfy equation (33), so that alqyl 4- b2r\xx = (a2 + b2) xxyv (34) The co-ordinates xlt yx also satisfy equation (33). 152 Trans. Acad. Sci. of St. Louis. If we now make xv yl variables, equation (34) becomes a2hj + &V = (a2 + b2) xy. (35) This equation is satisfied by the co-ordinates £, 77 and also by the co-ordinates xv yx\ and, therefore, represents a curve which passes through the given point P and through the in- tersections with the hyperbola of the normals through P. Equation (35) represents an equilateral hyperbola, which will be designated in what follows as the auxiliary hyperbola. The equations of its asymptotes are a2 x = a2 + b2 and (3«) y = ^+l?ri- (37) Equation (36) represents a straight line parallel to the axis a? of Y and at a distance ■ 0 , ,., c from it. a2-\- bl a2 a2 As 2 . 7 2 is necessarily positive, the expression 2 . ,2 ? always has the same sign as £; therefore, the asymptote rep- resented by equation (36) lies on the same side of the axis a2 of Y as the point P. And as 2, ,2 is less than unity, the asymptote lies between the axis of Y and the point P. The distance from the point P to this asymptote is a2 b2 £ o ■ „?= o ■ to £• (38) y ~ w+h~y = 1# (47) This, if £, r] are regarded as variables, is the equation of the evolute of the hyperbola, which is, therefore, the locus of the 15G Trans. Acad. Sci. of St. Louis. point r when the auxiliary and given hyperbolae arc tangent to each oilier. From the property of the evolute already cited (pp. 141-2) and by reference to the figures, it is apparent that for an}- point on the evolute (Fig. 10) one branch of the auxiliary hyperbola is tangent to one branch of the given hyperbola and cuts the other branch of the given hyperbola in 1 1 1 '> \ ^ \ X \j£ ,/XN. • / \\ S A* S AT \\ S AT sV/' -r*-— 1 _ -N*1 A Figure 10. a single point, while the other branch of the auxiliary hyper- bola cuts one branch of the given hyperbola in a single point and there are three normals, PNX, PN2, PJVS; for any point in the plane on the convex side of either branch of the evolute (Fig. 11), one branch of the auxiliary hyperbola cuts one branch of the given hyperbola in two points and cuts the other branch of the given hyperbola in a single point, while the other branch of the auxiliary hyperbola cuts the one branch of the given hyperbola in a single point and there are four normals, PNV PN,, PJ^3, PNi; and for any point in the plane on the concave side of both branches of the evolute (Fig. 12), one branch of the auxiliary hyperbola cuts one Engler — The Normal to the Conic Section. 157 Figure 11. 158 Trans. Acad. Sci. of St. Louis. branch, and the other branch of the auxiliary hyperbola cuts the other branch of the given hyperbola each in a single point and there are two normals, PJSTiy PN2. Special Cases. 1. When the point P lies on the given hyperbola one of the normals through the point is tangent to the auxiliary hyperbola at that point and is found precisely as in the case of the ellipse, if we observe that the constant base of the triangles on the asymptote parallel to the axis of X 2. If the given point P is on either of the axes of the given hyperbola, the auxiliary hyperbola degenerates into its asymp- totes, one of which is the axis of the given hyperbola through the point, the other is obtained by the construction. The intersections of these asymptotes with the given hyperbola give the normals. It will be observed that if the given point P is at one of the cusps of the evolute the asymptote parallel to the axis of Y is tangent to the hyperbola and, therefore, three of the normals coincide. The constructions for these cases can easily be supplied by the reader. THE NORMAL TO THE CONJUGATE HYPERBOLA. The equation of the hyperbola conjugate to the one repre- sented by equation (32) is /VIA O/" * £=_1. (48) ai o2 The auxiliary hyperbola which gives the normals through P for this hyperbola is a2£y + b2rjx = (a2 + b2) xy, (49) which is identical with equation (35). From this it appears that the same auxiliary hyperbola gives the normals not only Engler — The Normal to the Conic Section. 159 for the original hyperbola but for the conjugate hyperbola as well, as is shown in Fig. 13, in which the point P is so situated Figure 13. that there are two normals, PiV2, PJ¥3, to the original hyper- bola and two normals, PJVlt P^, to the conjugate hyperbola. Issued July 15, 189S. Transactions of The Academy of Science of St. Louis. VOL.. VIII. No. 9. CONTRIBUTION TO THE FOSSIL FLORA OF FLORISSANT, COLORADO. WALTER C. G. KIRCHNER. Issued December JO, 1898, CONTRIBUTION TO THE FOSSIL FLORA OF FLORISSANT, COLORADO.* Walter C. G. Kirchner. The region about Florissant, Colorado, has become famous for its prolific beds of plants and insects, and is regarded with much interest by the paleontologist. The remarks which follow are based on the results of a trip to this region and also on the examination of an interesting collection of fossil plants which were obtained from the same locality. The collection, which includes many hundred specimens, was made by Dr. G. Hambach and is now in the possession of Washington University. A careful investigation of the material has led to the com- pilation of a catalogue of the fossil plants found at Florissant, and has shown that the collection contains several species which have not hitherto been described. The collection also contains a few plants which, if not entirely new, have not been mentioned as being found at Florissant. The first published account of the region about Florissant is found in the report of Mr. A. C. Peale on the geology of Hayden Park. The account is short and makes mention of only a few fossil plants. A better account is found in a report by Professor Samuel H. Scudder, entitled " The Tertiary Lake basin at Florissant, Colorado, between South and Hay- den Parks." This report gives a detailed account of the geology of the place and an interesting synopsis of its pale- ontology. Mr. Leo Lesquereux, who for a number of years was the government paleontologist, has done most toward contributing to a knowledge of the fossil flora of the Western States aud Territories. His works have proved an invaluable aid. * Presented in abstract before The Academy of Science of St. Louis, June 6, 1898. (161) 162 Trans. Acad. Set. of St. Louis. The plants found at Florissant belong to the Tertiary period and to that division known as the Green River group. The shales in which the fossil plants occur are composed of vol- canic sand and ash, and are mostly drab, light-gray or light- brown in color. Some of the plants have been beautifully preserved. In the enumeration of species of the Green River group, Mr. Lesquereux states that out of 228 species 152 were found at Florissant. From the present catalogue it will be seen that the list has been increased and that 213 species can now with apparent safety be included in the flora of Florissant. I am greatly indebted to our paleontologist, Dr. G. Ham- bach, for the privilege of examining his private collection and for the kind assistance he has rendered in the preparation of this work. To a number of others who have enabled me to carry on the work, I should also like to extend my sincere thanks. In the nomenclature of some of the species, I have been permitted to use the names of Dr. Gustav Hambach, Professor Edmund A. Engler, President of the Academy of Science, and Mr. D. S. Brown, who by his munificence has materially aided the department of Natural Science in Washington University. CATALOGUE OF PLANTS. CRYPTOGAMAE. Characeae. 1. Chara? glomerata, Lesqx. Rept. U. S. Geol. Surv. 8: 135. Musci. 2. Fontinalis pristina, Lesqx. Rept. U. S. Geol. Surv. 8:135. 3. Hypnum Brownii, sp. nov. Infra, p. 178. PI. XII. fig. 4, 4a. Dr. Hambach. 4. H. Haydenii, Lesqx. Rept. U. S. Geol. Surv. 7:44. Kirchner — The Fossil Flora of Florissant, Colorado. 163 Rhizocarpeae. Salvinia Alleni, Lesqx.=Tmesipteris Alleni (5). jS. cyclophytta, Lesqx.=Phyllites cyclophyllus (48). Lycopodiaceae. 5. Tmesipteris Alleni, (Lesqx.) Hollick, Bull. Torr. Bot. Club. 21: 256. Salvinia Alleni, Lesqx. Rept. U. S. Geol. Surv. 7: 65. Ophioglossum Alleni, Lesqx. Ann. Rept. U. S. Geol. Surv. 1872 : 371. Isoeteae. 6. Isoetes brevifolius, Lesqx. Rept. U. S. Geol. Surv. 8 : 136. FlLICES. 7. Adiantites gracillimus, Lesqx. Rept. U. S. Geol. Surv. 8:137. 8. Asplenium (Diplazium) Crossii, nom. nov. Knowlton, Cat. Cret. and Tert. PI. N. Am. 1898: 44. Diplazium Muelleri, Heer. Rept. U. S. Geol. Surv. 7: 55. Diplazium Muelleri, Heer.= Asplenium Crossii (8). 9. Sphenopteris Guyottii, Lesqx. Rept. U. S. Geol. • Surv. 8 : 137. Wash. Univ. Coniferae. 10. Glyptostrobus Europaeus, Heer, Ann. Rept. U. S. Geol. Surv. 1873 : 409. — Rept. U. S. Geol. Surv. 7: 74. 11. G. Ungeri? Heer, Rept. U. S. Geol. Surv. 8: 139.— Knowlton, Cat. Cret. and Tert. PI. N. Am. 1898: 113. Wash. Univ. It is probably the same as G. Europaeus Ungeri, Heer, Rept. U. S. Geol. Surv. 8:222. 12. Pimelia delicatula, Lesqx. Rept. U. S. Geol. Surv. 8: 168. 164 Trans. Acad. Sci. of St. Louis. 13. Pinus Florissanti, Lesqx. Kept. U. S. Geol. Surv. 8: 138. 14. P. Hambachi, sp. nov. Infra, p. 179. PL XIII. fig. 3. Wash. Univ. 15. P. palaeostrobus? (Ett.) Heer, Rept. U. S. Geol. Surv. 7: 83. P. polaris, Heer, Ann. Rept. U. S. Geol. Surv. 1873: 410. P. polaris, Heer. = Pinus palaeostrobus (15). 16. Podocarpds eocenica? Ung. Rept. U. S. Geol. Surv. 8: 140. 17. Sequoia affinis, Lesqx. Rept. U. S. Geol. Surv. 7: 75. 18. S. Langsdorfii, (Brgt.) Heer, Rept. U. S. Geol. Surv. 7:76. 19. Taxodium distichum miocenicum, Heer. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297. 20. Thuites callitrina, Ung. Ann. Rept. U. S. Geol. Surv. 1872: 371. 21. Widdringtonia linguaefolia, Lesqx. Rept. U. S. Geol. Surv. 8: 139. MONOCOTYLEDONES . Typhaceae. 22. Najadopsis rugulosa, Lesqx. Rept. U. S. Geol. Surv. 8: 142. 23. Potamogeton geniculatus, Al. Br. Rept. U. S. Geol. Surv. 8 : 142. 24. Potamogeton vertictllatus, Lesqx. Rept. U. S. Geol. Surv. 8: 142. 25. Typha latissima, Al. Br. Rept. U. S. Geol. Surv. 8:141. Aroideae. 26. Acorus affinis, Lesqx. Ann. Rept. U. S. Geol. Surv. 1873: 410. Not afterwards recognized. Kirchner — The Fossil Flora of Florissant, Colorado. 165 27. Acorus brachystachys, Heer. (?) Rept. U. S. Geol. Surv. 7 : 105. Lemnaceae. 28. Lemna penicillata, Lesqx. Rept. U. S. Geol. Surv. 8: 143. Palmae. 29. Palmocarpon? globosum, Lesqx. Rept. U. S. Geol. Surv. 8: 144. DICOTYLEDONES. Myricaceae. Callicoma microphylla? Ett.=Myrica Drymeja (34). 30. Myrica acuminata, Ung. Rept. U. S. Geol. Surv. 7: 130.— Ann. Rept. U. S. Geol. Surv. 1873:411. 31. M. amygdalena, Sap. Rept. U. S. Geol. Surv. 8: 147. 32. M. Bolanderi, Lesqx. Rept. U. S. Geol. Surv. 7: 133. Florissant? M. caUico?naefolia, Lesqx. = Myrica Drymeja (34). 33. M. Copeana, Lesqx. Rept. U. S. Geol. Surv. 7: 131. —Ann. Rept. U. S. Geol. Surv. 1873: 411. M. diversifolia, Lesqx. = Crataegus flavescens (181). 34. Myrica Drymeja, (Lesqx). n. comb. Knowlton, Cat. Cret. and Tert. PI. N. Am. 1898: 146. Wash. Univ. Myrica callicomaefolia, Lesqx. Rept. U. S. Geol. Surv. 8: 146. Callicoma microphylla, Ett. Rept. U. S. Geol. Surv. 7 : 246. 35. M. fallax, Lesqx. Rept. U. S. Geol. Surv. 8:147. Wash. Univ. 36. M. insignis, Lesqx. Rept. U. S. Geol. Surv. 7: 135. — Ann. Rept. U. S. Geol. Surv. 1874: 312. 37. M. latiloba, Heer, var. acutiloba, Lesqx. Rept. U. S. Geol. Surv. 7: 134. 166 Trans. Acad. Sci. of St. Louis. 38. Myrica obscura, Lesqx. Rept. U. S. Geo!. Surv. 8:145. Wash. Univ. 39. M. polymorpha, Lesqx. Rept. U. S. Geol. Surv. 8 : 146. Wash . Univ . 40. M. rigida, Lesqx. Rept. U. S. Geol. Surv. 8: 145. Wash. Univ. 41. M. Scottii, Lesqx. Rept. U. S. Geol. Surv. 8:147. Wash. Univ. 42. M. Zachariensis, Sap. Rept. U. S. Geol. Surv. 8: 146. Betulaceae. 43. Alnus cordata, Lesqx. Rept. U. S. Geol. Surv. 8: 151. 44. A. Kefersteinii, Goepp. Rept. U. S. Geol. Surv. 7 : 140. 45. Betula dryadum, Brongn. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297, 46. B. Florissanti, Lesqx. Rept. U. S. Geol. Surv. 8: 150. 47. B. truncata, Lesqx. Rept. U. S. Geol. Surv. 8: 150. 48. Phyllites cyclophyllus, (Lesqx). Hollick, Bull. Torr. Bot. Club. 21:256 (1894). Salvinia cyclophylla, Lesqx. Ann. Rept. U. S. Geol. Surv. 1873: 408. — Rept. U. S. Geol. Surv. 7: 64, 315. CUPULIFERAE. ' 49. Carpinus attenuata, Lesqx. Rept. U. S. Geol. Surv. 8 : 152. Wash. Univ. 50. C. fraterna, Lesqx. Rept. U. S. Geol. Surv. 8: 152. Wash. Univ. 51. C. grandis, Ung. Rept. U. S. Geol. Surv. 7: 143. 52. C. pyramidalis, Heer. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297. Kirchner — The Fossil Flora of Florissant, Colorado. 167 53. Castanea intermedia, Lesqx. Kept. U. S. Geol. Surv. 7 : 164. 54. Ostrya betuloides, Lesqx. Kept. U. S. Geol. Surv. 8: 151. 55. Quercus antecedens, Sap. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297. 56. Q. Drymeja, Ung. Ann. Rept. U. S. Geol. Surv. 1871: 308.— Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297. — Rept. U. S. Geol. Surv. 8 : 154. 57. Q. elaena, Ung. Rept. U. S. Geol. Surv. 8: 155. 58. Q. elaenoides, Lesqx. Mem. Mus. Comp. Zool. 6, no. 2: 4. 59. Q. Mediterranea, Ung. Rept. U. S. Geol. Surv. 8: 153. 60. Q. neriifolia, Al. Br. Rept. U. S. Geol. Surv. 7 : 150. — Ann. Rept. U. S. Geol. Surv. 1873: 411. Identi- fication said to be doubtful. 61. Q. Osbornii, Lesqx. Rept. U. S. Geol. Surv. 8:154. 62. Q. pyrifolia, Lesqx. Rept. U. S. Geol. Surv. 8: 154. 63. Q. salicina, Sap. Scudder, Bull. U. S. Geol. Surv. 6, no. 2:297. 64. Q. serra, Ung. Rept. U. S. Geol. Surv. 8: 153. Salicineae. 65. Populus arctica, Heer, Rept. U. S. Geol. Surv. 8 : 159. Wash. Univ. 66. P. balsamoides ? Goepp. var. lattfolia, Lesqx. Rept. U. S. Geol. Surv. 8:158. 67. P. Heerii, Sap. Rept. U. S. Geol. Surv. 8:157. Wash. Univ. 68. P. oxyphylla, Sap. Rept. U. S. Geol. Surv. 8 : 159. 168 Trans. Acad. Sci. of St. Louis. •69. P. Zaddachi, Heer, Kept. U. S. Geol. Surv. 8: 158. 70. P. pyrifolia sp. nov. Infra, p. 185. Plate XV. fig. 4. Wash. Univ. 71. Salix amygdala efoli a, Lesqx. Rept. U. S. Geol. Surv. 8: 156. Wash. Univ. 72. S. Integra,- Goepp. Scudder, Bull. U. S. Geol. Surv. 6, no. 2 : 297. — Rept. U. S. Geol. Surv. 7 : 167. 73. S. Lavateri, Heer. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297. 74. S. Libbeyi, Lesqx. Rept. U. S. Geol. Surv. 8 : 156. 75. S. media, Heer. Scudder, Bull. U. S. Geol. Surv. 6, no. 2 : 297. — Rept. U. S. Geol. Surv. 7: 168. 76. S. varians, Goepp. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297. Ulmaceae. 77. Celtis McCoshii, Lesqx. Rept. U. S. Geol. Surv. 8:163. 78. Planera longifolia, Lesqx. Rept. U. S. Geol. Surv. 7: 189. Wash. Univ. 79. P. longifolia, var. myricaefolia, Lesqx. Rept. U. S. Geol. Surv. 8:161. Wash. Univ. 80. P. Ungeri, Ett. Rept. U. S. Geol. Surv. 7 : 190. 81. Ulmus Braunii, Heer, Rept. U. S. Geol. Surv. 8: 161. Wash. Univ. 82. U. Brownellii, Lesqx. Rept. U. S. Geol. Surv. 8 : 160. Wash. Univ. 83. U. Fischeri, Heer. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 296. 84. U. Hilliae, Lesqx. Rept. U. S. Geol. Surv. 8: 160. Wash. Univ. 85. U. tenuinervis, Lesqx. Rept. U. S. Geol. Surv. 7: 188. Kirchner — The Fossil Flora of Florissant, Colorado. 169 MOREAE. 86. Ficus Haydenii, Lesqx. Eept. U.S. Geol. Surv. 7: 197. Wash. Univ. 87. F. lanceolata, Heer, Kept. U. S. Geol. Surv. 7: 192. — Ann. Rept. U. S. Geol. Surv. 1873: 414. Santaleae. 88. Santalum Americanum, Lesqx. Rept. U. S. Geol. Surv. 8: 164. Proteaceae. 89. Banksites lineatus, Lesqx. Rept. U. S. Geol. Surv. 8: 165. 90. Lomatia abbreviata, Lesqx. Rept. U. S. Geol. Surv. 8:167. 91. L. acutiloba, Lesqx. Rept. U. S. Geol. Surv. 8: 167. Wash. Univ. 92. L. hakeaefolia, Lesqx. Rept. U. S. Geol. Surv. 8:166. 93. L. interrupta, Lesqx. Rept. U. S. Geol. Surv. 8 : 167. Wash. Univ. 94. L. spinosa, Lesqx. Rept. U. S. Geol. Surv. 8: 166. 95. L. terminalis, Lesqx. Rept. U. S. Geol. Surv. 8 : 166. Wash. Univ. 96. L. tripartita, Lesqx. Rept. U. S. Geol. Surv. 8 : 167. PlMELEAE. 97. Pimelea delicatula, Lesqx. Rept. U. S. Geol. Surv. 8:168. Oleaceab. 98. Fraxinus abbreviata, Lesqx. Rept. U. S. Geol. Surv. 8: 170. 170 Trans. Acad. Sci. of St. Louis. 99. F. Brownellii, Lesqx. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 296. 100. F. Heerii, Lesqx. Rept. U. S. Geol. Surv. 8: 169. 101. F. Libbeyi, Lesqx. Rept. U. S. Geol. Surv. 8: 171. 102. F. mespilifolia, Lesqx. Rept. U. S. Geol. Surv. 8: 169. 103. Fraxinus? myricaefolia, Lesqx. Rept. U. S. Geol. Surv. 8 : 170. 104. F. praedicta, Heer, Rept. U. S. Geol. Surv. 8: 169. 105. F. Ungeri, Lesqx. Rept. U. S. Geol. Surv. 8: 171. 106. Olea praemissa, Lesqx. Rept. U. S. Geol. Surv. 8: 168. Apocyneae. 107. Apocynophyllum Scudderi, Lesqx. Rept. U. S. Geol. Surv. 8: 172. Dr. Hambach. CONVOLVUL ACE AE . 108. Porana Speirii, Lesqx. Rept. U. S. Geol. Surv. 8: 172. 109. P. tenuis, Lesqx. Rept. U. S. Geol. Surv. 8 : 173. Myrsineae. 110. Myrsine latifolia, Lesqx. Rept. U. S. Geol. Surv. 8 : 173. Sapotaceae. 111. Bumelia Florissanti, Lesqx. Rept. U. S. Geol. Surv. 8: 174. Ebenaceae. 112. Diospyros brachysepala, Al. Br. Rept. U. S. Geol. Surv. 8: 174. 113. D. Copeana, Lesqx. Rept. U. S. Geol. Surv. 7: 232. Kirchner — The Fossil Flora of Florissant, Colorado. 171 114. D. cuspidata, sp. nov. Infra, p. 185. Plate XII. fig. 1. Dr. Hambach. 115. Macreightia crassa, Lesqx. Kept. U. S. Geol. Surv. 8: 175. Ericaceae. 116. Andromeda rhomboidalis, Lesq. Kept. U. S. Geol. Surv. 8: 176. 117. Vaccinium reticulatum? Al. Br. Rept. U. S. Geol. Surv. 7:235. Araliaceae. 118. Aralia dissecta, Lesqx. Rept. U. S. Geol. Surv. 8: 176. 119. Hedera marginata, Lesqx. Rept. U. S. Geol. Surv. 8:177. Hamamelidae. 120. Liquid amb ar Europaeum, Al. Br. Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 296. —Rept. U. S. Geol. Surv. 8:159. Magnoliaceae. j-121. Carpites Pealei, Lesqx. Rept. U. S. Geol. Surv. 7 : 306. 122. C. milioides, Lesqx. Rept. U. S. Geol. Surv. 8: 204. Saxifrageae. 123. Weinmannia Haydenii, Lesqx. Rept. U. S. Geol. Surv. 8 : 178. Wash. Univ. Bhus Haydenii, Lesqx. Ann. Rept. U. S. Geol. Surv. 1873 : 417. — Bept. U. S. Geol. Surv. 7 : 294, 327. 124. Weinmannia integrifolia, Lesqx. Rept. U. S. Geol. Surv. 8: 178. 125. W. obtusifolia, Lesqx. Rept. U. S. Geol. Surv. 8: 178. 172 Trans. Acad. JSci. of St. Louis. Malvaceae. 126. Sterculia Engleri, sp. uov. Infra, p. 180. PI. XIV. fig. 3. Wash. Univ. 127. S. rigida, Lesqx. Eept. U. S. Geol. Surv. 8: 179. TlLIACEAE. 128. Tilia populifolia, Lesqx. Kept. U. S. Geol. Surv. 8 : 179. Wash. Univ. Aceraceae. 129. Acer Florissanti, sp. nov. Infra, p. 181. PI. XI. fig. 1. Wash. Univ. 130. A. mysticum, sp. nov. Infra, p. 181. PL XI. fig. 2. Wash. Univ. 131. A., species, Lesqx. Kept. U. S. Geol. Surv. 8: 181. Sapindaceae. 132. Dodonaea, species, Lesqx. Kept. U. S. Geol. Surv. 8: 182. Wash. Univ. 133. Sapindus angustifolius, Lesqx. Kept. U. S. Geol. Surv. 7: 265. Wash. Univ. 134. S. inflexus, Lesqx. Kept. U. S. Geol. Surv. 8: 182. 135. S. lanceofolius, Lesqx. Rept. U. S. Geol. Surv. 8:182. 136. S. obtusifolius, Lesqx. Rept. U. S. Geol. Surv. 7: 266. 8: 181, 210. 137. S. stellariaefolius, Lesqx. Rept. U. S. Geol. Surv. 7: 264. Staphyleaceae . 138. Staphyle a acuminata, Lesqx. Rept. U. S. Geol. Surv. 7 : 267, 326. Dr. Hambach. Kirchner — Tlie Fossil Flora of Florissant, Colorado. 173 Frangulaceae. 139. Celastrinites elegans, Lesqx. Eept. U. S. Geol. Surv. 8: 185. 140. Celastrus fraxinifolids, Lesqx. Rept. U. S. Geol. Surv. 8 : 184. 141. C. Greithianus, Heer, Rept. U. S. Geol. Surv. 8: 184. 142. C. Lacoei, Lesqx. Rept. U. S. Geol. Surv. 8: 184. Iliceae. 143. Ilex grandifolia, Lesqx. Rept. U. S. Geol. Surv. 8:187. 144. I. knightiaefolia, Lesqx. Rept. U. S. Geol. Surv. 8: 188. 145. I. microphylla, Lesqx. Rept. U. S. Geol. Surv. 8: 186. 146. I. pseudo-stenophylla, Lesqx. Rept. U. S. Geol. Surv. 8: 185. 147. I. quercifolia, Lesqx. Rept. U. S. Geol. Surv. 8 : 186. 148. I. rigida, sp. nov. Infra, p. 182. PI. XIV. fig. 2. Wash. Univ. 149. I. sphenophylla ? Heer. Lesqx. Ann. Rept. U. S. Geog. Surv. 1873:415. 150. I. subdenticulata, Lesqx. Rept. U. S. Geol. Surv. 7 : 271. — Ann. Rept. U. S. Geol. Surv. 1873 : 416. Rhamneae. 151. Paliurus Florissanti, Lesqx. Rept. U. S. Geol. Surv. 7:274. — Ann. Rept. U. S. Geol. Surv. 1873:416. 152. P. orbiculatds, Sap. Rept. U. S. Geol. Surv. 8 : 188. 153. Rhamnus ellipticds, sp. nov. Infra, p. 183. PI. XV. fig. 3. Wasli. Univ. 174 Trans. Acad. Sci. of St. Louis. 154. R. notatus? Sap. Rept. U. S. Geol. Surv. 8: 189. 155. R. oleaefolius, Lesqx. Rept. U. S. Geol. Surv. 8: 188. 156. Zizyphus obtdsa, sp. nov. Infra, p. 182. PI. XIII. fig. 1. Wash. Univ. JUGLANDEAE. Carya foYnuca, ( Ung. )= Hicoria juglandiformis (158). C. Bruckmanni? Heer.=Hicoria Bruckmanni (157, a). C. rostrata, (Goepp.) Schp. = Hicoria rostrata (159). 157. Engelhardtia oxyptera, Sap. Rept. U. S. Geol. Surv. 8: 192. 157. a. Hicoria Bruckmanni, (Heer.) n. couib. Knowl- ton, Cat. Cret. and Tert. PL N. Am. 1898: 117. Carya Bruckmanni'} Heer. Lesqx. Rept. U. S. Geol. Surv. 8: 191. 158. H. juglandiformis, (Sternb.) n. comb. Kuowlton, Cat. Cret. and Tert. PL N. Am. 1898 : 117. Carya bilinica, (Ung.) Ett. Lesqx. Rept. U. S. Geol. Surv. 8: 191. 159. H. rostrata, (Goepp.) n. comb. Kuowlton, Cat. Cret. and Tert. PL Am. 1898: 118. Carya rostrata, (Goepp.) Schimp. Lesqx. Rept* U. S. Geol. Surv. 8: 191. 160. Juglans affinis, sp. nov. Infra, p. 184. PL XIII. fig. 2. Wash. Univ. ♦ 161. J. costata, Ung. Rept. U. S. Geol. Surv. 8: 190. 162. J. Crossii, nom. nov. Kuowlton, Cat. Cret. and Tert. PL N. Am. 1898: 122. — Infra, p. 183. PL XIV. fig. 1. Wash. Univ. Juglans denticulata, Heer. Lesqx. Rept. U. S. Geol. Surv. 7:289.— Ann. Rept. U. S. Geol. Surv. 1871: 298. J. denticulata, Heer, 1869, preoccupied by J. denticulata, O. Web. 1852. J. denticulata, Heer. = Juglans Crossii (162). 163. J. Florissanti, Lesqx. Rept. U. S. Geol. Surv. 8:190. Kirchner — The Fossil Flora of Florissant, Colorado. 175 164. J. thermalis, Lesqx. Rept. U. S. Geol. Surv. 7 : 287, 327.— Scudder, Bull.U. S. Geol. Surv. 6, no. 2: 297. 165. Pterocarya Americana, Lesqx. Rept. U. S. Geol. Surv. 7: 290, 327. Anacardiaceae. 166. Rhus acuminata, Lesqx. Rept. U. S. Geol. Surv. 8 : 194. Wash. Univ. 167. R. cassioides, Lesqx. Rept. U. S. Geol. Surv. 8 : 193. 168. R. coriarioides, Lesqx. Rept. U. S. Geol. Surv. 8: 193. 169. R. Evansii, Lesqx. Ann. Rept. U. S. Geol. Surv. 1871: 293. 1872:402.— Rept. U. S. Geol. Surv. 7: 291, 327. 170. R. fraterna, Lesqx. Rept. U. S. Geol. Surv. 8 : 192. Wash. Univ. 171. R. Hilliae, Lesqx. Rept. U. S. Geol. Surv. 8: 194. Wash. Univ. JR. Haydenii, Lesqx. = Weinraannia Haydenii (123). 172. R. rosaefolia, Lesqx. Rept. U. S. Geol. Surv. 7 : 293. 173. R. rotundifolia, sp. nov. Infra, p. 184. PI. XII. fig. 2. Dr. Hambach. 174. R. subrhomboidalis, Lesqx. Rept. U. S. Geol. Surv. 8:195. 175. R. trifolioides, Lesqx. Rept. U. S. Geol. Surv. 8:196. 176. R. vexans, Lesqx. Rept. U. S. Geol. Surv. 8: 195. Z ANTHOXYLE AE . 177. Zanthoxylon spireaefolium, Lesqx. Rept. U. S. Geol. Surv. 8: 196. 176 Trans. Acad. Sci. of St. Louis. KOSIFLORAE. 178. Amelanchier typica, Lesqx. Kept. U. S.Geol. Surv. 8: 198. 179. Amygdalus gracilis, Lesqx. Rept. U. S. Geol. Surv. 8: 199. 180. Crataegus acerifoli a, Lesqx. Rept. U. S. Geol. Surv. 8:198. 181. C. flavescens, Newby. Proc. U. S. Nat. Mus. 5 : 507. Myrica diver sifolia, Lesqx. Rept. U. S.Geol. Surv. 8: 148. 182. Rosa Hilliae, Lesqx. Rept. U. S. Geol. Surv. 8 : 199. Leguminosae. 183. Acacia septentrionalis, Lesqx. Rept. U. S. Geol. Surv. 7: 299. Caesalpina? linearis, Lesqx. =Mimosites linearis (191). 184. Cassia Fischeri, Heer, Rept. U. S. Geol. Surv. 8: 202. 185. Cercis parvifolia, Lesqx. Rept. U. S. Geol. Surv. 8 : 201. 186. Cytisus Florissantianus, Lesqx. Rept. U. S. Geol. Surv. 8 : 200. 187. C. modestus, Lesqx. Rept. U. S. Geol. Surv. 8: 200. 188. Dalbergia cuneifolia, Heer, Rept. U. S. Geol. Surv. 8 : 200. 189. Leguminosites, species, Lesqx. Rept. U. S. Geol. Surv. 8: 203. 190. L. serrulatus, Lesqx. Rept. U. S. Geol. Surv. 8: 202. Mimosites linearifolius, Lesxq. = Mimosites linearis (191). Kirchner — The Fossil Flora of Florissant, Colorado. 177 191. Mimosites linearis, (Lesqx.) n. comb. Knowlton, Cat. Cret. and Tert. PI. N. Am. 1898 : 144. Wash. Univ. Mimosites linearifolius, Lesqx. Rept. U. S. Geol. Surv. 7 : 300. Caesalpinia? linearis, Lesqx. Ann. Rept. U. S. Geol. Surv. 1873: 417. 192. Podogonium acuminatum, Lesqx. Rept. U. S. Geol. Suit. 8: 201. 193. P. americanum, Lesqx. Rept. U. S. Geol. Surv. 7 : 298. 8: 212. Podogonium, sp., Lesqx. Ann. Rept. U. S. Geol. Surv. 1878: 417. Incertae Sedis. 194. Ailanthus, species, Scudder, Bull. U. S. Geol. Suit. 6, no. 2: 296. 195. Antholithes amoenus, Lesqx. Rept. U. S. Geol. Surv. 8: 203. 196. A. improbus, Lesqx. Rept. U. S. Geol. Surv. 8: 204. Florissant? 197. A. obtusilobus, Lesqx. Rept. U. S. Geol. Suit. 8: 203. 198. Asplenium tenerum, Lesqx. Rept. U. S. Geol. Surv. 8: 221. Dr. Hambach. 199. Bombax, species, Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 296. 200. Carpites gemmaceus, Lesqx. Rept. U. S. Geol. Surv. 8 : 204. 201. Carpolithes, species, Lesqx. Ann. Rept. U. S. Geol. Surv. 1873:418. 202. Catalpa, species, Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 296. 203. Colutea, species, Scudder, Bull. U. S. Geol. Surv. 6, no. 2:296. 204. Convulvulaceae ? sp. nov. Infra, p. 187. PI. XV. fig. 2. 178 Trans. Acad. Sci. of St. Louis. 205. Cokylus McQuarryi, Heer, Ann. Rept. U. S. Geol. Surv. 1871: 308. 206. Fagus Antipofii, Heer, Ann. Rept. U. S. Geol. Surv. 1871: 308. 207. Iris, species, Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297. 208. Onagraceae? sp. nov. Infra, p. 186. PI. XV. tig. 1. 209. Palaeocarya, species, Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 297. 210. Prunus, species, Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 296. 211. Robinia, species, Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 296. 212. Sabal, species, Scudder, Bull. U. S. Geol. Surv. 6: 297. 213. Spiraea, species, Scudder, Bull. U. S. Geol. Surv. 6, no. 2: 296. DESCRIPTIONS OF NEW SPECIES. MUSCI. Hypnum. 1. Hypnum Brownii, sp. nov. (Plate XII. figs. 4, 4a). Stems creeping, forked or divided into nearly opposite branches ; leaves ovate lanceolate, acuminate, concave. This specimen is figured here to show the general habit of the plant. The leaves in most cases are indistinct and only the more solid stems are discernible. The plant seems to be analogous to the recent species H. populeum, Sw. The stems are slightly curved and the leaves on some portions are faintly visible. The leaves do not appear to be very closely imbricated. The tissues of mosses are very delicate, which explains why the fossil remains are so exceedingly rare. It is only in the later formation that the fossil forms are found. Kirchner — The Fossil Flora of Florissant, Colorado. 179 CONIFERAE. PlNUS. 2. Pinds Hambachi, sp. nov. (Plate XIII. fig. 3). Leaves in threes, long, narrow, pointed; stem rather thick; nerves obscure. The specimen shows the end of a pine-branch with about 45 needles closely fascicled. The stem is 7 centimeters long and 3 to 4 millimeters in thickness and has a roughish ap- pearance. The leaves average about 7 centimeters in length and a little less than a millimeter in width ; their nervation is obscure. At the end of the stem are about 36 leaves so closely crowded together as to make all but the tips indistinct; but about oue centimeter lower down there is a distinct bun- dle of nine leaves which spring from the surface of the stem. Upon closer examination it is shown that the leaves are fascicled ; and that there appear to be three leaves to each sheath. I have not been able to find a description nor a figure of a pine which would characterize this specimen. The branch with the leaves has been well preserved, although from the nature of evergreens most specimens show only a few leaves. It is of course to be regretted that the fruit is not present. Two other species have been found at Floris- sant, P. palaeostrobus, Ett., and P. Florissanti, Lesqx., the latter having been determined by the cone. MOREAE. Ficus. 3. Ficus Haydenii, Lesqx. (Plate XII. fig. 3). U. S. Geol. Kept. 7: 197. Fl. XXX. fig. 1. — Ann. Rep. U. S. Geol. . Surv. 1872: 394. Leaf subcoriaceous, entire, broadly lanceolate with a cor- date base, tapering upward to along acumen; petiole long; primary nerve strong near the base ; secondary nerves thinner, curved in passing to the borders, camptodrome. This leaf answers well to the description given by Les- quereux. The form of the leaf is well preserved with the 180 Trans. Acad. Set. of St. Louis. exception of the apex, but enough is shown to indicate the acuminate nature. The blade was probably 7.5-8.5 centimeters long, and has a width of 45 centimeters at the broadest portion near the base. The petiole is nearly 6 centimeters long. In the figures and description given by Lesquereux, the base curves slightly downward to the petiole. The present leaf is distinctly cordate. The secondary nerves alternate and are confluent to the mid- rib. They pass to the borders in gentle curves and anasto- mose in simple bows. Their angle of divergence is 40-50 degrees. The leaf has the form of a Po2mlus but the vena- tion is that of a Ficus. The description which Lesquereux gives was from a single specimen, and, even if the specimen here figured has a cordate base, the other characters are in favor of F. Haydenii. The specimen submitted to Les- quereux is from Black Buttes, Wyoming, and the species is considered very rare. MALVACEAE. Sterculia. 4. Sterculia Engleri, sp. nov. (Plate XIV. fig. 3). Leaf coriaceous, comparatively large, palmately trilobate, triple-nerved; lobes cut nearly to the base, linear-oblong, entire, apparently acuminate; the middle lobe narrower than the lateral ; base rounded ; primary nerves distinct. The specimen is a fragment, the upper portions of the lobes having been destroyed. The lateral lobes are 10 to 13 millimeters broad and the central lobe a little more than half as wide. Their apparent length was 7 or 8 centimeters. The lobes are almost straight and slightly narrowed toward the base which is rounded and decurrent to a thick petiole. The primary nerves, diverging at an angle of about 40°, arise from the top of the petiole. The secondary nerves are not visible. Only one other species has been found at Florissant. By the facies of this leaf, it might be compared with 8. Labrusca, Ung. The main points of difference are, however, that the middle lobe of this leaf is the narrowest and that the lateral lobes are more nearly straight and longer than those of Kirchner — The Fossil Flora of Florissant, Colorado. 181 the leaves described and figured by Unger. The leaves of 5. lugubris, described by Lesquereux in the " Flora of the Dakota Group," are also comparable with this leaf, but differ from it in their greater size, cuneate base and scythe-shaped lateral lobes. ACERACEAE. Acer. 5. Acer Florissanti, sp. nov. (Plate XI. fig. 1). Leaf comparatively large, five-lobed, outline broadly oval; petiole long; middle lobe longest, broad and oblong; lateral lobes lanceolate ; margin incised ; base broad ; basal nerves five, straight; secondary nerves distinct and nearly straight to the borders; veinlets anastomose, forming a fine net-work. The markings of this leaf have been beautifully preserved. The blade is 10.5 centimeters long and about 8 centimeters broad. The middle lobe is oblong and dissected at the top. Of the lateral lobes, the upper are the largest. The central basal nerves, diverging at an angle of 30°-40°, are the most prominent. The secondary nerves are quite straight, and enter the points of the teeth on the margin. About forty- six species from the Tertiary formations of Europe have been described and referred to the different types of this genus, whereas in this country comparatively few fossil species have been found. This leaf by the facies and character of the venation is comparable with the recent species, A. dasycar- pu?n, which it resembles in many respects. 6. Acer mysticum, sp. nov. (Plate XI. fig. 2). The specimen which is here figured represents the fruit of an Acer. The fruit, oblong in shape, is nearly two centi- meters long and six centimeters wide, and appears to contain an ovate seed. Along the back of the wins: are four or five strong nerves. These give off nervilles which are more or less forked, and which cross the wing transversely nearly to the margin. The wing is somewhat wider below the middle. Since the seed was found by itself it cannot be safely classed with any of the known species. 182 Trans. Acad. Sci. of St. Louis. ILICEAE. Ilex. 7. Ilex eigida, sp. nov. (Plate XIV. fig. 2). Leaves coriaceous, short-petioled, oblong; margin irreg- ularly and acutely dentate; teeth beset with sharp spines; primary nerve very prominent; secondary nerves nearly straight, camptodrome; veinlets obscure. The leaf is unlike those described by Lesquereux. It is about seven centimeters long and nearly two centimeters broad. The petiole is thick and short. The teeth are very sharp and larger in the middle of the leaf than at the ends ; in some cases the spines can be distinctly seen. The midrib is the most conspicuous, and in the living species must have been very prominent and strong, giving rigidity to the leaf. The secondary nerves are firm and alternating, and branch out nearly at right angles from the primary. This species shows the main characteristics of the Iliceae. It differs from the known species in its oblong form and in the long and sharply pointed spines which project out nearly at right angles from the margin of the leaf. RHAMNACEAE. ZlZYPHUS. 8. Zizyphus obtusa, sp. nov. (Plate XIII. fig. 1). Leaf simple, small, subcoriaceous, ovate, somewhat un- equal, triple-nerved ; margin evenly serrated, the teeth fine and sharp, and smaller toward the apex; base round; apex obtusely pointed; petiole short and rather thick; basal nerves strong. This leaf, although rather small, preseuts the character- istic nervation of the genus. The blade is 2 centimeters long and 1.3 centimeters broad. The middle nerve is thick and runs straight to the apex. The lateral, diverging at an angle of about 35°, curve gently along the margin. The sec- ondary nerves which branch from the middle nerve diverge at a greater angle and anastomose in simple bows; those of Kirchner — The Fossil Flora of Florissant, Colorado. 183 the lateral nerves, toward the outer side of the leaf, seem to form a fine net-work which runs parallel with the margin of the leaf and sends off minute branches into the points of the teeth. The tertiary nerves are scarcely discernible. This leaf does not answer to the description of any of the five species enumerated by Lesquereux. It seems more closely allied to some form of Z. Ungeri of Heer. Rhamnus. 9. Rhamnus ellipticus, sp. nov. (Plate XV. fig. 3). Leaf simple, subcoriaceous, elliptical ; margin entire ; primary nerve thick and straight; secondary nerves close, numerous, nearly parallel, camptodrome ; areolation quadrate. The base of this leaf is wanting. The leaf is 2 centimeters broad and about 5 centimeters long. The secondary nerves, given off at an angle of 30°-40°, are nearly straight and some- times incomplete. The leaf is analogous to that of R. inter- medins, Lesqx., but the midrib is thinner and the secondary nerves in astomosing near the margin are more looped than those of the leaf described by Lesquereux. JUGLANDACEAE. JUGLANS. 10. Juglans Crossii, Knowlton. (Plate XIV. fig. 1). Juglans denticulata, Heer. Lesqx. Ann. Rep. U. S. Geol. Surv. 1871: 298.— Tert. Fl. 7 : 289. PI. L VIII. Jig. 1 . Leaves long-lanceolate, narrowed to a point and denticulate upwards ; either rounded to the petiole or gradually attenuated to it (Lesquereux). The specimen is fragmentary, but enough of the plant is present to show the necessary characteristics. Portions of two leaflets attached to the stem are shown in the fragment. One leaflet which seems to be terminal has a petiole whose length is three centimeters; most of this leaflet is wanting; its base is unequal and attenuated to the petiole. About half of the second leaflet is present. It was probably 10-12 centi- meters long and 4 centimeters wide, lanceolate or elliptical, 184 Trans. Acad. Sci. of St. Louis. and narrowed to the point; it is denticulate and becomes nar- rower as it approaches the petiole which is four millimeters long. The base is unequal and round. The primary nerve is very strong ; the secondary nerves are prominent, alternate, nearly straight, curving near the border, camptodrome, and are connected with the points of the teeth by distinct veinlets ; the tertiary nerves are very oblique and, as in many recent species, nearly at right angles to the secondary ; nervilles dis- tinct. These leaves belong, undoubtedly, to the Juglandaceae. In the general character of the venation they agree with all the figures with which they were compared. 11. Juglans affinis, sp. nov. (Plate XIII. fig. 2). Leaves lanceolate, acuminate, narrowed in a curve to a short petiole ; border serrulate ; lateral veins distant, alternate, parallel, curved in passing to the borders, ascending high along them in simple festoons, separated by short intermediate tertiary veins; areolation irregularly quadrate. The above description corresponds almost entirely to that of J. alkalina, Lesqx., the only essential difference being in the kind of border. From the figure of J. alkalina, Lesqx. (Hayden's Rept. 7. PL LXII. figs. 6-9), it appears that the leaves must have had uniformly crenate margins, while the present species shows distinct serrations. The figure repre- sents the only leaf of the kind that came to my notice. The leaf is nearly ten centimeters long and two and a half centimet- ers broad. The nervation is distinct. In consequence of the serrated border, the bows along the margin are connected with the teeth by fine nervilles. ANACARDIACEAE. Rhus. 12. Rhus rotundifolia, sp. nov. (Plate XII. fig. 2). Leaf trifoliolate (or odd-pinnate); leaflets orbiculate, ses- sile; nervation looped; primary nerve strong; secondary nerves curved in passing to the borders, camptodrome. Most of the plants belonging to this genus are characterized by a strong nervation which varies much according to the Kirchner — The Fossil Flora of Florissant, Colorado. 185 character of the leaf. In many cases where the margins of the leaves are entire the nervation is camptodrome. Plants with compound leaves are not uncommon. The specimen shows three leaflets, and although the leaf was probably trifo- liolute, since the petiole has been broken off there remains the possibility that it might have been odd-pinnate. The leaflets are a little less than one centimeter long. The primary nerve is strong, especially toward the base. The secondary nerves, making an angle of 35°-65°, are confluent with the primary, mostly curved in passing to the borders, and camptodrome. The tertiary nerves are somewhat curved and form a polygo- nal net-work. This leaf is comparable with R. villosa, of Linnaeus. EBENACEAE. DlOSPYROS. 13. Diospyros cuspidata, sp. nov. (Plate XII. fig. 1). Calyx thick, coriaceous, four-lobed ; lobes deeply cut, ovate-lanceolate, concave ; peduncle comparatively long. The lobes of the calyx, about one centimeter long and less than half as broad, are cut nearly to the base. Their nerva- tion is indistinct. On account of the overlapping of one of the lobes by the opposite one, there appear to be but three divisions; on closer examination, however, the remains of the four sepals can be distinctly made out. In the general char- acter the calyx might be compared with Macreightia crassa> Lesqx., although this species is described as having only three lobes. Two other species, D. brachysepala, Al. Br., and D. Copeana, Lesqx., represented by their leaves, have been found at Florissant. SALICACEAE. POPULUS. 14. Populus pyrifolia, sp. nov. (Plate XV. fig. 4). Leaf membranaceous, ovate-lanceolate, very obtuse at the base, (apparently) crenulate, palmately seven-nerved; cen- tral, and upper lateral primary nerves strong and straight, at 1S6 Trans. Acad. Sci. of St. Louis. an acute angle of divergence, ascending high up along the borders and anastomosing in gentle curves. This leaf was membranaceous in texture, and in the speci- men the margin is poorly defined. Only at one place is the crenulate nature discernible. The margin near the base, however, is entire. The blade was probably 8 centimeters long and 5.3 centimeters broad at the widest part. The petiole is wanting. The primary nerves, diverging from each other at an angle of 25°-30°, are straight and strong toward the base, becoming much thinner as they approach the margin of the leaf. The secondary nerves are nearly parallel in their course through the blade. On the lower side of the lateral primary nerves, gently curving branches are given off which anastomose with other branches near the margin. The nervilles are scarcely visible and are nearly at right angles to the principal nerves. The leaves of the poplar vary as to size, shape, margin and num- ber of primary nerves, which accounts in great part for the many species that have been ascribed to the genus. From Florissant alone we have five other species, some of them also representatives of the European fossil flora. Flores. (Plate XV. figs. 1, 2.) While looking over the collection of fossils, two flowers came to my notice, which I have not been able to classify with certainty. Plate XV. figure 1, shows the remains of a flower that seems to have affinities with some of the Onagraceae. The calyx-tube prolonged beyond the ovary is 3.3 centimeters Ions: and inflated above. Its divisions cannot be made out. The ovary is 1.3 centimeters long and three millimeters broad at the middle. The petals, five in number, are membrana- ceous, lanceolate and marked by a central nerve. The stigma is cylindrical and the course of the style through the tube can be traced nearly to the ovary. Most of the living Onagraceae with tubular calyces are 4-merous ; a variation, however, might not be improbable. Kirchner — The Fossil Flora of Florissant, Colorado. 187 Another flower, shown at Plate XV. fig. 2, has some of the characteristics of the Convolvulaceae. The corolla, appar- ently funnel-form or campanulate, has a five-sided border divided bv slight clefts into five lobes. The sides of the borders are about two centimeters long. The venation has been beautifully preserved. Each lobe has a central straight nerve which passes to the apex. On either side of this are nerves which curve gently from the apex along the entire length of the lobe. Other prominent nerves mark the divisions between the lobes. These nerves before reaching the cleft in the margin generally divide and send branches toward the apices of the lobes on either side. The principal nerves generally anastomose near the margins. The nervilles are at right angles and the areolation is mostly quadrate. BIBLIOGRAPHY. Balfour, John Hutlon. Introduction to the Study of Palaeon- tological Botany. Edinburgh, 1872. von Eltingshausen, O. Die Blatt-Skelete der Dikotyle- donen. Wien, 1861. von Etlingshausen, C. und A. Pokorny. Die Gef&sspflanzen Oesterreichs in Naturselbstdruck. Wien, 1.873. Geyler, H. Th. Ueber fossile Pflanzen von Borneo. Goeppert, H. R. Die Gattungen der Fossilen Pflanzen. Hollick, A. Fossil Salvinias. Bull. Torr. Bot. Club. 21 : 253. Knowlton, F. H. The Flora of the Dakota Group, a posthu- mous work by Leo Lesquereux. (Rept. U. S. Geol. Surv., Powell, 1891). — A Catalogue of the Cretaceous and Terti- ary Plants of North America. (Bull. U. S. Geol. Surv. No. 152. 1898). Lesquereux, Leo. Fossil Flora. (Rept. U. S. Geol. Surv., Hayden, 1871: 281-373). — Cretaceous Flora. (Ibid. 6). — The Tertiary Flora. (Ibid. 7, Part 2. 1878).— The Cretaceous and Tertiary Flora. (Ibid. 8, Part 3. 1883). 188 Trans. Acad. Sci. of St. Louis. Peale, A. 0. — Ann. Kept. U. S. Geol. Surv. Terr. 1873: 210. Washington, 1874. Saporta el Marion. — Recherches sur les Vegetaux Fossiles de Meximieux. 1876. Schimper, W. P. et A. Mougeot. Monographic des Plantes Fossiles des Gres Bigarre' de la Chaine des Vosges. Leip- zig, 1844. Scudder, S. H. The Tertiary Lake Basin of Florissant, Colorado, between South and Hayden Parks. (Bull. U. S. Geol. and Geog. Surv. Terr. 6, No. 2: 279-300. 1881). Seward, A. G. Fossil Plants for Students of Botany and Geology. 1. Cambridge, 1898. Solms-Laubach. Fossil Botany. Oxford, 1891. linger, F. Die fossile Flora von Sotzka. 1850. — Beitrage zur Flora der Vorwelt. 1847. United States. Ann. Repts. of Geol. Surv. Mont, and Adj. Terr., Hayden, 1871 (1872); Geol. Surv. Terr., Hayden, 1872 (1873) ; Geol. and Geog. Surv. Terr., Hayden, 1873 (1874); Geol. and Geog. Surv. Terr., Hayden, 1875 (1877); Geol. and Geog. Surv. Terr., Hayden, 1877 (1879). Velenovshy, J. — Dis Flora aus den Ausgebrannten Tertiaren Letten von Vrsovic bei Laun. (Abhand. d. k. bohm. Gesellsch. d. Wissensch. 6: 11. Prag, 1881-1882). Zittel, K. A. Handbuch der Palaeontologie. Abt. 2. 1890. EXPLANATION OF ILLUSTRATIONS. Plates XI-XV. (All of the figures slightly reduced, unless otherwise noted.) Plate XI. — 1, Acer Florissanti. 2, A. mysticum. Fla.teXIl. — l,Diospyros cuspidata. 2, Rhus rotundifolia. S, Ficus Hay- denii. 4-4a, Ilypnum Brownii, the latter enlarged. Plate XIII. — 1, Zizyphus obtusa. 2, Juglans affinis. 3, Pinns Hambachi. Plate XIV. — 1, Juglans Crossii. 2, Ilex rigida. 3. Sterculia Engleri. Plate XV. — 1, Onagraceous? flower. 2, Convolvulaceous? flower. 3, Bhavinus ellipticus. 4, Populus pyrifolia. Issued December 10, 1S9S. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate XI. FLORISSANT PLANTS. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate XII. FLORISSANT PLANTS. Trans, acad. Sci. of St. Louis, Vol. VIII. Platk XIII. FLORISSANT PLANTS. Trans, acad. Sci. of St. Louis, Vol. VIII. Plate XIV. FLORISSANT PLANTS. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate XV. FLORISSANT PLANTS. Transactions of The Academy of Science of St. Louis. VOL. VIII. No. 10. ON THE MODE OF DISSEMINATION OF USNEA BARB ATA. HERMANN VON SCHRENK. Issued December 17, 1898. ON THE MODE OF DISSEMINATION OF USNEA BARB AT A.* Hermann von Schrenk. In many places along the North Atlantic coast one comes across tracts of forest in which a large number of the trees are covered with the common gray lichen, Usnea barbata and its varieties. This lichen often covers the trees entirely, giving them a sort of hoary appearance ; where many such trees stand close together, especially if they happen to be conifers, the effect is truly a gloomy and somber one. The question has frequently occured why so many of the trees should be dead, and so universally covered with large masses of the lichen, and the notes following are to record some of the observations made in the Middle and North Atlantic States. The species of Usnea are found all over the world, grow- ing in nearly all cases on trees. Two forms may be distin- guished. In the more common and well known, a central main stem of the thallus is firmly attached to the wood or bark, and forms a small clump or bush. The other type is that represented by such forms as Usnea barbata dasypoga, Usnea longissima, etc., in which the thallus as a unit has little or no connection with the branch or tree upon which it rests. These forms are truly epiphytic, and free from any suspicion of absorbing even small and insignificant quantities of food substances from their substratum. It is with this type that the following deals. The thallus consists of long threads, some as much as 20 inches in length, branching freely both * Presented before the Academy of Science of St. Louis in preliminary form Dec. 21, 1896 (Trans. 7: lxi.), and, by title, Dec. 6, 1898. (189) 190 Trans. Acad. Sci. of St. Louis. dichotomously and by adventive branches, the latter standing usually at an angle of about 80°. Examined more minutely the thallus is found to consist of three layers, an external cortical layer, in which the hyphae are intricately entangled into a network of greater or less density, a middle layer with the gonidia, very much looser in texture, and a central or medullary layer in which the hyphae are parallel to the long axis of the thallus. Towards the tips both the main trunk and the branches rapidly decrease in diameter. As the thal- lus grows older, the cortical layer weakens and breaks in many places, exposing the medullary layer, with rings of cortex about it. This is specially marked when the thallus is wet and stretched. On the long hanging forms of Us7iea, apothecia are very rarely found. On Long Island and in Connecticut Usnea barbata (L.) var. plicata, Fr.,* is very common particularly on Juniperus Vir- giniana and Pinus rigida, while farther north in Massachu- setts and Maine I have found IT. barbata (L.) var. dasypoga, Fr., the more prevalent form, growing on the white spruce, JPicea alba, and the balsam fir, Abies balsamea. This apparent predilection for the short-leaved Coniferae is probably due to the fact that the long threads are easily caught and held by branches with short leaves extending like brushes in all direc- tions. Deciduous-leaved trees and the longer-leaved pines do not present such favorable supports; pieces of the lichen placed on such trees rarely become fixed, while those placed on Juniperus and Abies remained fast as long as they were kept under observation. If one examines a tree covered by the lichen, one will find pieces varying from an inch in length to great masses on one and the same branch. The threads are rarely straight, but are bent and twisted among themselves, and wind in be- tween and around the leaves and stem in a tangle which it is almost impossible to unravel. (See PI. XVI.) This cannot be the result of growth, as the lichen grows so very slowly, and such tangles I have seen form in a few weeks. Nor is * I am indebted to Miss Clara E. Cummings for the determination of these varieties. von Sclirerik — Dissemination of Usnea barbata. 191 it conceivable how the wind alone could cause the branches of the thallus to coil about the leaves and stem, and in- tertwine in the way they do. The explanation I believe lies in the fact that the branches are capable of a good deal of motion when moistened unequally. When a fila- ment is torn away, its branches are always more or less bent and coiled. When moistened, the various branches at first either straighten out or coil more, depending upon the side which is wet first; after a few moments, the re- verse motion takes place. The bending is due to the threads of the medullary layer which absorb water readily and in- crease some 10-15% in length, in the younger branches. If then a filament is caught by a branch like that of the spruce, it is first held simply by the diverging leaves. As it dries the various branches curve about in almost as many directions as there are branches, the tips describing arcs of 180° in many cases. When equal drying has taken place the branches en- deavor to get back to their original position. Oftentimes they do, but far more frequently some other branch has crossed over, or a leaf has been pushed aside by a contracting filament, and in that way the branches are forced to assume different positions from the one first occupied. Repeated wetting and drying brings about more coiling and uncoiling, and ultimately such a tangle as is represented in the figure results. Many a filament lodges on a branch, coils and un- coils and is blown away again after a time, but on the whole more filaments are firmly secured shortly after being caught. I experimented with numerous pieces, which were placed on spruce branches, and carefully watched. After about four weeks the branches of the thallus were found firmly coiled about the leaves of the spruce. A few days later the whole piece might be off on another branch. (This was true only of single filaments, and did not occur with larger masses.) The manner in which the lichen is disseminated is of inter- est. The wind is the chief factor concerned, although I have once seen birds using it for building their nest, and the lichen seemed to be in good condition. It is extremely probable that it is often carried by birds, but I have not seen any observations 192 Trans. Acad. Sci. of St. Louis. to this effect.* The wind sweeping through a tree catches hold of the long strands hanging from the branches and tears off pieces, which are carried to other trees. To find how and where the wind can tear the thallus various tests were made.f Several masses were weighed dry, and again after being soaked in distilled water. The latter was removed as com- pletely as possible from the lichen by means of filter paper. Two trials showed: — Weight of mass dry = 3.8 grs. 3.97 grs. Weight soaked 12 hrs. =13.3 grs. 11.5 grs., that is, the mass increased about three times in weight, which is probably below what it would be during a heavy rain. The threads are very elastic when wet, stretching 30-50% of their original length before breaking. The next test was to de- termine the longitudinal breaking strength. It was somewhat difficult to attach weights to the thread, and after some trials, wire hooks were inserted into loops tied into the ends of the threads, to one of which a scale pan with weights was attached. A number of threads were selected, as nearly as possible of the same diameter, but as this method is of course not very exact, the results can have only a relative value. When dry the filaments broke at: When soaked 12 hours : 198 gr. 28 gr. 109 gr. 23 gr. 144 gr. 27 gr. 130 gr. 25 gr. Threads of llsnea barbata, var. ceratina gave these results : — When dry : When soaked 30 minutes : 224.6 gr. 53.5 gr. 266.0 gr. 43.0 gr. * Schimper (Die epiphytische Vegetation Amerikas: 31) says that birds frequently use the strands of Tillandtia usneoides for building nests and thus spread the plant. t Tests made with Bamalina reticulata are omitted as they have already been given by G. J. Peirce, On the mode of Dissemination and on the reticulations of Bamalina reticulata. Bot. Gazette 25 : 404. 1898. von Schrenk — Dissemination of Usnea barbata. 193 These figures show that the individual threads are very much weaker when wet than when dry, and at the same time they are much heavier and more elastic. The action of the wind is liable to be most effective in times of storm, when it is also raining, in other words, when the thallus is wet and conse- quently most readily torn. It remained to be seen what wind velocities would suffice to bring about the tearing. So far as I know no tests have ever been m;ide to determine experimentally the strength or resisting power of any plant member in a stream of moving air. It is claimed that a hur- ricane may tear the leaf lamina, and we know it breaks branches and timber, but we do not know the actual strength of these organs. This has been due no doubt to an inability to measure the velocities of and pressure exerted by any given wind with any degree of accuracy. A piece of apparatus recently devised * enables one to measure with great exactness the velocity of, and pressure exerted by a moving current of air. Unfortunately for the tests made with Usnea, all the conditions existing during a rain-storm could not be brought about, so the results are somewhat higher than had been expected. Nevertheless they are given, as they bring out some interesting connections between longitudinal strength and the tearing force of the wind. A pipe 8 inches in diameter and some 20 ft. long was attached to the exhaust of a fan inclosed in a box, and capa- ble of several thousand revolutions per minute. By throt- tling the engine which turned the fan, the speed of the latter, and, correspondingly, the velocity of the air current produced by it, could be varied at will. Into the pipe a window was cut four feet from the end, and into this the cup for measuring the air current, as well as the lichen, was placed. The dry mass of Usnea barbata was first tried. In all the trials the long threads were tied to a vertical post in the pipe by means of silk threads. Repeated trials showed that wind velocities of 77 miles per hour were not sufficient to cause any pieces to fly off. The thallus was then soaked in water 30 minutes to 1 hour * Nipher, F. E. Trans. Acad. Sci. St. Louis 8: 1. 1898. 194 Trans. Acad. Sci. of St. Louis. and tested similarly. The first pieces came flying off at velo- cities of about 50 miles per hour, and from then on large masses were torn away. By calculations it was found that the wind velocities represent the following pressures per square centimeter: — 49 miles per hour = 2.738 gr. per sq. cm. 52 " " " =3.425 " " " " 56 «' " " = 3.930 " " " " 70 " " " =6.105 " " " " It appears from this that a wind velocity of at least 50 miles an hour would be required to tear away pieces of the thallus. That this figure is much too high will be evident to one who knows how frequently the lichen is torn, and an inspec- tion of the accompanying table recording wind velocities along the Atlantic and Pacific coast shows that the mean velocities nowhere approach this. The seeming error lies here. In the first place when the moist thallus was introduced into a stream of air, the drying action of the latter was felt im- mediately, so quickly, in fact, that after one minute's expo- sure the thallus was almost dry, i. e., no longer in its weakest condition. In its native habitat no drying takes place, even with high wind velocity, because of the atmospheric moisture. Again, granting that a velocity of 50 miles tore the mass, it is not by any means necessary to assume such velocities for winds acting for periods longer than one minute. In this respect the experiment was a failure, that it gave only an extreme wind velocity. The relation found between tension and pressure was unexpected. The wind pressure per square centimeter at the period of greatest velocity, i. e., 70 miles, was 6.1 gr., while at the tearing point it was but 3.+ grs. per sq. cm. The lowest breaking strength of the thallus threads however was found to be at least 23 gr. von Schrenk — Dissemination of Usnea barbata. 195 o o > 8 B CD ■a +3 u o V XI o o r— I a s X3 -. 3 CJO CC a> O. Q- S .a H y CD Q Vi * Si ^© O . oo °?co "3*1 =?* f 00 l-H XS 00* 05* CO liS m O O CD CD t» >o* 1- CO co iH e- 1 i— i CN i-H > o to . 00 . OO i— i ^W ^* ten °?o to toi CM * «o co oo CO i-H* ■* us co lO co * »o CO t>. CO eoi-H f-H i-H r^ i-H f-H o °?* ^* -*co oo CO eo i— 1 i— i I-H i-H I-H 43 Pi «*! °°«P t»i5 • CO oo . 00 . 00 ^* . US =?* CD 30 ©9 co ej t- CO O CO Mt* CN CO -H T* ia CO l-H CO CO iH OS »-H f-H i-H f-H ^o • CO • CO «0B *o "t Ol °?oi • o ^o *« t~ > . GO . C3 00 r_ to ^o • CO Ins to . oo ^* 13 COOl co 01 t-(M -. CO ^H CO o* f-H* w> CO >oeo CO "H 1-3 i-H |-H 1—1 1—1 t© "3 • • GO . r» . -* • o . o . © ?H S3 f-H * oo * •* o t>C0 C7^ CO coOl l-H f-H f— 1 i-H -O en °?0 t © "? fr- ^O °?o t CO ^O i-H . l-H • 03 V ■* t» l- CO C5 CO ee * t- C* C5 O CO* iO CO t^CO eo p^ »— » pH i-H .-H a *?o t* to to t* to CO . o to "3* . 00 09 -a f— i f-H i-H I-H • j co 1 B GO "3 . . 93 o p^ a > i a > -4-J > oo CO o3 p i to S ' o" o GO S3 a go" 43* *H c 'O !3 2 S3 oo c3 i-H GO r— 1 o sT '3 a "3 s Cm S3 .a. O •c M fid P«( Pammel — Caryopsis and Endosperm of some Grasses. 213 Moeller,1 Koernicke,2 Hoehnel,3 Marek,4 Wittraack,5 and Bessey.6 Harz has given us an excellent account of the different forms of Triticum: T. vulgare, T. polonicum, 1. turgidum, and T. durum, as well as the different forms of Triticum Spelta . Triticum Spelta, L., and T. monococcum. These differ anatomically in a very marked degree, according to Harz, chiefly in the epidermal cells of the caryopsis, and the nucellus. In T. Spelta the epidermis consists of tangentially elon- gated cells, the internal walls greatly thickened. The under- lying cells are somewhat thinner-walled. The pericarp is much more simple in its structure than in T. vulgare. My studies do not entirely agree with those given by Harz. The testa consists of elongated cells not markedly different from those of the pericarp. The aleurone consists of a single layer of large cells densely filled with aleurone grains. The remainder of the endosperm usually of larger cells containing the spherical or elliptical starch grains. The cells of the first row of the starch layer are smaller. The remnants of the nucellus evident only in the groove of the seed. Here a dozen or more rows of small thick-walled, colorless cells occur, which soon disappear on either side (PI. XVIII. 1. — Triticum sativum, PI. XIX. 9). Triticum monococcum, L. This does not differ so essen- tially from T. Spelta except in the structure of the epidermal cells of the caryopsis. The exterior walls of the epidermal cells are greatly thickened and larger than in T. Spelta. The caryopsis consists of two rows of cells. The cells of the testa are longer, thick-walled and colored brown. The nucellus of this species as in Triticum Spelta. The aleurone of one row of large cells, and occasionally a second row of smaller cells, longer than broad, densely filled with 1 l. c. 2 Koernicke and Werner, Handbuch des Getreidebaues. Bonn. 1885. 3 Vergleichende Untersuchungen d. Epidermis d. Gramineen Spelzen u. deren Beziehung zum Hypoderma. Haberlandt's Wiss. Prakt. Untersuch- ungen auf d. Gebiete d. Pflanzenbaues 1: 168. 4 Das Saatgut u. dessen Einfluss auf Menge u. Giite d. Ernte. Wien. 193. /. 174. 1875. 6 Ueber die Erkennung d. Verfalschung von Roggenmehl mit Weizen- mehl. Sitzb. Ver. Prov. Brandenburg 24: 4-8. (Abstract Hanausek, Bot. Centralbl. 13: 91-92. 1883.) « 1. c. 214 Trans. Acad. Sci. of St. Louis. aleurone grains. The starch layer variable; as a rule, how- ever, the starch cells are much larger than the aleurone. In some cases the first row of cells smaller, and these carry more protein than the cells below. The starch grains are large, elliptical or spherical, extremely variable in size. These cells also have an abundance of aleurone (PI. XVIII. 2). Secale cereale, L. An excellent account of the structure of this fruit will be found in the works of Harz, Tschirch & Oesterle,1 Tietschert,2 Gregory3 and other writers on economic food products. The pericarp consists of tangentially elongated epidermal cells with large cell cavities. The outer wall is thickened as well as the inner, the lateral walls thinner. Harz states that there is an important distinction between Secale and Triticum; a somewhat analogous structure, however, occurs in the spelts. The epidermal layer is followed by smaller thin-walled parenchyma cells. The layer next to these parenchyma cells is frequently composed of thick-walled porous cells with pore canals. These cells not evident except in mature fruits. The testa is but slightly developed and consists of compara- tively small cells frequently colored brown. The uucellus occurs as a remnant especially in the groove, where the cells are thick-walled and somewhat gelatinous. The endosperm resembles that of wheat. The aleurone layer consists of a single layer of cells. The exterior walls are greatly thick- ened. The cells of the starch layer large, containing a large number of round or elliptical starch grains, extremely vari- able in size. The starch grains on the whole are larger than those of the genus Triticum (PI. XVIII. 6). Hordeum vulgare, L. The structure of Hordeum has been given by Tschirch & Oesterle,4 Morris & Brown,5 Harz,6 Ku- delka,7 and others. 1 1. c. a Keimungsversuche mit Roggen unci Raps. 104./. 1. s Die Membranverdickungen der sogenannten Querzellen in der Frucht. wand des Roggens. FUnfstiick, Beitrage zur wiss. Bot. 2: 165-168. 4 Anatomischer Atlas der Pharmakognosie und Nahrungsmittelkunde. 9: 181. pi. 41. 5 Jour. Roy. Chem. Soc. 57: 458. 6 1. c. 1159./. 136. 7 Ueber d. Entw. u.d. Bau d. Frucht- u. Samenschalen unserer Cerealien Inaug. Diss. Berlin. 1875. Pammel — Caryopsis and Endosperm of some Grasses. 215 The adhering palet consists of several rows of thick- walled cells. The epidermal cells are longer and somewhat thicker-walled. The cells below are also thick-walled, pro- vided with pore canals. In places the epidermal cells develop into a short trichome. Underneath the thick-walled cells occur several rows of thin-walled irregular parenchyma cells. The pericarp follows. It consists of several rows of thick- walled translucent cells, the cavity being very much reduced. The testa is colored brown. The cells are tangentially elongated and the cell-walls are thinner. The nucellus is very much reduced excepting in the groove, where it occurs as thick-walled cells. The endosperm differs in a very marked degree from either that of Triticum or Secede, especially the aleurone, which consists of three to four or exceptionally more rows of cells. The starch cells are much larger and contain large spherical or elliptical starch grains, accompanied by numerous smaller ones. Small protein grains abundant (PI. XVIII. 9). Hordeum jubaCam,'L. The grain in Hordeum jubatum is adherent to the palet. The epidermis consists of thick-walled tangentially elongated cells, at least most of the cells are longer than broad. In part the cells are developed into short conical trichomes. The underlying cells are thick- walled with prominent pore canals. The remainder of the adhering palet consists of thin-walled cells much longer than broad. The pericarp as well as the testa is but slightly developed. In some cases the underlying parenchyma cells are not clearly defined, or at least not evident until the specimens are treated with chloral hydrate. The blackish pigment is largely found in the internal part of the palet. Some of the pig- ment also occurs in the pericarp and the aleurone layer. The pericarp consists of one or two rows of rather thin-walled tangentially elongated cells. The testa is reduced to a single layer of cells somewhat longer than broad. The nucellus is nearly absent except in the groove. It consists of a single row of thin-walled, colorless, compressed cells. In the groove several rows of cells occur. On either side of the groove these soon disappear. The aleurone in the specimens studied is made up of a single row of cells. The cells of the starch 216 Trans. Acad. Sci. of St Louis. layer are large and filled with round or elliptical grains. The starch cells are larger than those of the aleurone. The starch grains are mostly round or somewhat elliptical (PI. XVII. 6). Elymus canadensis, L. In Elymus canadensis and other species of the genus the grain is adherent to the glume. The outer part does not differ essentially from that of the genus Hordeum except that the walls of the epidermal cells are not so thick. The underlying cells are thick-walled as in the genus Hordeum. The pericarp consists of thick- walled tangentially elongated cells with a very small cell-cavity. The inner portion of the pericarp consists of a row of thinner- walled cells. The testa consists of a layer of thin-walled cells. The remnants of the nucellus much compressed, and on the addition of chloral hydrate appear as a translucent colorless layer with greatly thickened walls; the cell-cavity much reduced. The aleurone consists of a single layer of cells. The cells of the starch layer are much larger than the aleurone and filled with spherical or elliptical starch grains (PI. XVIII. 4). BAMBUSEAE. Arundinaria macrosperma , Michx. The large grains of this species are free. The epidermal cells of the peri- carp are thick- walled and roughened. The lateral walls are much thinner than the exterior. The external wall is plainly stratified. The cells are marked with prominent pore canals. The underlying layer consists of two to three rows of very similar cells except that the walls are less thickened. The internal part of the pericarp consists of one to two rows of thick-walled cells with a very narrow cell- cavity. The testa consists of thinner-walled cells of slightly brownish color. The aleurone consists of several rows of thick-walled cells. The large starch cells contain relatively large compound grains. The individual elements are five- to six-sided (PI. XVII. 5). Pammel — Garyopsis and Endosperm of some Grasses. 217 SYNOPSIS OF ANATOMICAL CHARACTERS FOR TRIBES. The following synopsis of the tribes studied by myself, including those studied by Harz, gives the chief anatomical characters of the fruit and seed. Oryzeae. Starch grains compound. Aleurone cells small, one or two rows. Testa but slightly developed. Oryza, Zizania, Leersia. Maydeae. Starch grains simple, solidly packed. Aleurone cells a single row. Testa and pericarp but slightly developed in Euchlaena, Coix. In Zea, peri- carp greatly developed, testa less so. Andropogoneae. Starch grains simple. Andropogon Ischaemum, Sorghum vulgare. Aleu- rone a single layer of cells. Testa with large cells, especially in our common forms of Sorghum. Paniceae. Starch grains simple, solidly packed in Panicum, Setaria; more loosely in Cenchrus. Aleurone cells a single row. Testa and pericarp slightly developed. Phalarideae. Starch grains compound. Aleurone cells a single row. Testa and peri- carp but slightly developed. Phalaris, Anthoxanthum, Hierochloe. Agrostideae. Starch grains compound in Agrostis, Galamagrostis, Ammophila, Alopecu- rus, Phleum? Stipa1? in Aristida, simple. Pericarp and testa slightly de- veloped, except Ammophila. Aleurone cells a single row. Aveneae. Compound starch grains, Avena, Arrhenatherum, Aim, Holcus, Deschamp- sia, Trisetum. Aleurone cells one or more rows. Pericarp and testa vari- able, usually not greatly developed. Chlorideae. Compound starch grains, Cynodon. Bouteloua? Aleurone cells one or more rows. Testa and pericarp but slightly developed. Festuceae. Compound starch grains, Phragmites, Arundo, Melica, Poa, Glyceria, Festuca. — Scholochloa, Koeleria, in part. Simple, Dactylis, Bromus, Brachy- podium, Festuca, Scholochloa. Aleurone cells one or more rows. Testa and pericarp not greatly developed. Nucellus pronounced in Bromus, Festuca, Brachypodium. Hordeae. Simple grains, Hordeum, Secale, Triticum, Elymus, Agropyron, Aegilops. Compound, Lolium, Nardus. Aleurone usually a single row. Nucellus very evident in Lolium, Elymus. Testa and pericarp not well developed. Bambziseae. Starch grains compound. Aleurone layer of more than one row of cells. Pericarp well developed. Testa less. Arundinaria. 218 Trans. Acad. Sci. of St. Louis. SUMMARY. Structurally there are wide differences between the tribes of the order, very marked in some closely allied genera. The pericarp is well developed in such genera as Zea, Arundi- naria, and fairly well in Trilicum, Secale, Hordeum and Avena. The testa is but slightly developed in most cases, notably so in Fesluca, Panicum glabrum, Aristida and Oryza saliva, the protective features being provided for by the glumes surrounding the fruit, or the wall of the ovary. The nucellus is never entirely absent, especially in the groove. It is usually much compressed. In the genera Fesluca and Bromus, the cells are large, thick-walled and mucilaginous, and no doubt act as reserve food. The aleurone layer is variable. It is never absent. Of one row of cells in Trilicum, Zea, Zizania. The cells are very small in Panicum Crus-galli, Aristida, Setaria italica. Of more than one row of cells in Avena, Arrhena- therum, Fesluca, Hordeum vulgare. The starch cells differ in size and contain small spherical or elliptical grains in Sor- ghum vulgare and Cenchrus tribuloides. Large spherical or somewhat elliptical grains occur in Trilicum and Hordeum, accompanied by numerous smaller ones. Small five or six- sided grains in Panicum Crus-galli, Zea Mays, Euchlaena mexicana. This applies in general to the tribe Maydeae and Paniceae. Cenchrus is, however, an exception to the rule. Compound starch grains occur in Zizania, Oryza, Avena, Arrhenalherum, Glyceria, Poa, Phalaris and Arundina- ria; most grasses appear to have compound grains. The endosperm always contains protein, though much reduced in the starch cells, except in the aleurone layer, where no starch occurs. The starch cells next to the endosperm contain more protein than the interior of the endosperm. Fat is also present in small amounts. The compound starch grains of Nardus in the tribe Hordeae are somewhat anomalous. Pammel — Caryopsis and Endosperm of some Grasses. 219 EXPLANATION OF ILLUSTRATIONS. PLATES XVII-XIX. Sections of Phragmiles communis, Oryza sativa, Glyceria fluitans, Arundinaria mawosperma, Arrhenalherum, Stipa ro- busla, Phleum pratense, and Dactylis glomerata, were drawn to the same scale with Grunow ocular and Schrauer No. 7 objective. All other figures drawn to the same scale with Abbe camera, Leitz objective No. 7 and ocular No. 3. The two systems nearly correspond. Different parts of the caryopsis lettered uniformly unless otherwise stated. P = pericarp, t = testa. a=aleurone. s = endosperm with starch grains, cc = cuticle, n = nucel- lus. gl = glume, scl = sclerotic cells. en = endosperm. e = embryo, ep = epidermis. All figures original wash drawings made from author's sketches by Miss King and uniformly reduced in engraving. Plate XVII. — i, Zizania aquatica: large compound starch grains in en- dosperm; aleurone layer a single row of cells. — n, Arrhenatherum avenaceum: aleurone layer of two rows of cells; starch grains compound. — m, Euchlaena mexicana: starch grains solidly packed together. — iv, Oryza sativa: compound starch grains; aleurone layer a single row of cells. — v, Arundinaria macro- sperma: aleurone layer of more than one row of cells; compound starch grains in endosperm. — vi, Hordeum jubatum : thick-walled sclerotic cells of glume with short trichomes above the caryopsis; c, the nucellus only a remnant. — vn, Glyceria fluitans: with compound starch grains. — vm, G. aquatica: with compound starch grains. — ix, Phalaris arundinacea: testa and pericarp but slightly developed; large compound starch grains. — x, Poa pratensis: large compound starch grains in endosperm. — xi, Dactylis glomerata: aleurone layer a single row of cells; starch grains simple. — xn, Phleum pratense : lower figure, general view cross section of seed with embryo in position. — xiii, Stipa robusta: thick-walled cells x above testa and several rows of thinner-walled cells underneath the fertile glume; longitudinal view of sclerotic cells of glume shown in upper figure ; cross section to the left with an enlarged view at I'. Plate XVIII. — i, Triticum Spelta: simple round or elliptical starch grains In endosperm. —ii, T. monococcum: remnants of nucellus present. — in, Bromus ciliatus var. purgans: nucellus with thick-walled stratified cells. — iv, Elymus canadensis: cells of glume thick- walled, also those of nucellus, but the latter not stratified. — v, Setaria italica: starch grains compressed ; aleurone cells small. — vi, Secale cereale: large spherical or somewhat ellip- tical starch grains; nucellus a single row of cells; p', porous thickened walls of pericarp. — vn, Panicum glabrum : aleurone cells small ; starch grains solidly packed. — vm, Aristida ramosisaima: compound starch grains in en- 220 Trans. Acad. Sci. of St. Louis. dosperm.— ix, Hordeum vidgare: thick-walled cells of glume with pore canals; alcurone cells of more than one row. Plate XIX. — i, Cenchrus tribuloides: aleurone layer of a single row of cells; starch grains spherical or elliptical. — n, Panicum Crus-galli: glume adjacent to pericarp with lower layer of cells thick -walled, showing pore canals; single row of cells to aleurone layer; starch grains solidly packed. — ill, Festuca Shortii: aleurone layer of one or more rows, cells of the nucellus thick-walled ; starch grains apparently compound, showing individual com- ponent parts separated. — iv, Sorghum vulgare var. communis (Andropogon Sorghum var.) : small aleurone cells ; simple spherical or elliptical starch grains. — v, Bouteloua hirsula: aleurone layer of several rows of cells; starch grains solidly packed. — vi, Avena sativa: pericarp with irregular cells on surface; the minute down of " seed; " Aa single row of cells to aleurone layer; in B more than one row; starch grains compound. — vn, Sporobolus cryptandrus : pericarp, c, mucilaginous; aleurone cells small, of a siDgle row; U, a more magnified portion of the pericarp and testa after the addition of chloral hydrate. — vin, Phragmites communis : aleurone layer of a single row of cells, starch grains solidly packed. — ix, Triticum vulgare: aleurone layer a single row of cells, cell-walls of testa and pericarp thick- walled; longitudinal view of pericarp cells shown at b; starch grains large, spherical. — x, Koeleria sp. : nucellus a single row of cells, also the aleurone layer; starch grains large, compound, testa of a single row of cells. — xi, Zea Mays: nucellus a single row of cells; pericarp much more developed than testa. Issued December 29, 1898. Trans, acad. Sci. of St. Louis, Vol. VIII. Plate XVII. CARYOPSIS AND ENDOSPERM OP GRASSES. Trans. Acad. Sci. of St. Louis, Vol. VIII. Plate XVIII. Trans. Acad. Sci of St. Louis, Vol. VIII. Plate XIX. CARYOPSIS AND ENDOSPERM OF GRASSES. Transactions of The Academy of Science of St. Louis. VOL. VIII. No. 12. TITLE-PAGE, PREFATORY MATTER AND INDEX RECORD FROM JAN. 1 TO DEC. 31, 1898. Issued January 30, 1899. List of Axithors. 221 LIST OF AUTHORS. Baker, F. C. 71 Bernays, A C. xix Carter, H. xxiii Engler, E. A. xxvi, 137 Harabach, G. xxv Hitchcock, A. S. xix, 55 Holman, M. L. xxii Johnson, J. B. xxi Kinealy, J. H. xxii Kinsley, C. xxii, xxiii, 107 Kirchner, W. C. G. xxiv, 161 Kodis, T. xxv Cipher, F. E. 1 Pammel, L. H. xix, xxv, 199 Ravold, A. N. xviii, xx, xxiii Robertson, C. xix, 43 Seddon, J. A. xxiii, xxiv Terry, R. J. xx Thompson, C. H. xxiv Von Schrenk, H. xxv, 25, 189 Wood, 0. M. xxvi Woodward, C. M. xx, 95 222 Trans. Acad. Sci. of St. Loxiis. GENERAL INDEX. Agar- gelatine xxi Alkali plants 66 Alternating currents xxiii Anatomy of seeds xix, xxv, 199 Annual rings 28 Aquatics 57 Bacteria cultures xx Bark injuries 26 Bark scorching 36, 40 Bees xix, 43 Blastoideae xxv Bogs 59 Buds, adventitious 34 Buds after tornado 27 Buildings, air-pressure on 1 Buffalo wallows 59 Caryopsis xxv, 199 Cements xxi Cervical rib xx Chlorine xviii Currents, electric xxlii Collector for wind-pressure 3 Conic section 137 Conjugate hyperbola 158 Constitution amended xvii Consumption xxiii Convolvulaceae? 177, 187-8, pi. 15 Curators xvii Defoliation and growth 29 Disinfectants xviii Disk collector 3, 9, 21 Dissemination xxv, 189 Drying of wood 38, 40 Dynamos xxii, 107 Ecology xix, xxiv, xxv, 55, 189 Efficiency of gearing xx Electric currents xxiii Ellipse 144 Endosperm xxv, 199 Eplcycloidal teeth 97 Exchanges xv, xxx Field plants 67 Filtration xxii Floods xxiii Florissant fossils xxiv, 161 Flow in hydraulics xxiv Forests 67 Formaldehyde xviii Fossil plants xxiv, 161 Friction of gearing xx, 95 Frost injury 29 Gauge for wind pressure 6, 10, pl.l Gearing efficiency xx, 95 Gelatine-Agar xxi Geography, ecological xix, 55 Grass caryopsis xxv, 199 Gray, M. L., resolution xviii. Growth of twigs after tornado 28 Halophytes 57, 66 Histology, vegetable xxv, xix, History of Academy xiii Hydraulics xxiv Hydrophytes 57 Hygiene xxiii Hyperbola 151 Involute teeth 102 Kansas plants xix, 55 Leaves affected by wind 25-6 Librarian, report xvii, xxx Man in nature xix Marantaceae xxiv Marsh plants 66 Meadow plants 66 Mesophytes 57, 66 Meteorites xxviii Milk xxiii Mississippi river xxiii Mollusks of New York 71 Morphology xxv Motors, series 118, 132 199 General Index. 223 Negro sociology xxv New York mollusks 71 Normal to conic section 137 Nucellus of grasses 200 Oedematous wood 34 Officers xxvi Onagraceae? 178, 186, 188, pi. 15 Ontogeny xix Overcooling tissues xxv Paleontology xxiv, 161 Parabola 137 Pathology xx Pollination xxiv Portland cement xxi Prairies 63 President, address xvii, xxvi Pressure board 2, 7, 14, 17, pi. 2 Pressure gauges 6, 10, pi. 1 Pressure of wind 1 Resistance in hydraulics xxiv Rib, cervical xx Rings of wood 28 Rock plants 60 Root activity 33 Salt marshes 66 Sand hills 62 Scorching of bark 36, 40 Seed anatomy xix, xxv, 199 Series dynamos 107 Shunt dynamo, 108,/. 1 Sociology xxvi Stylar movements xxiv Sulphur dioxide xviii Sun-scald 36, 40 Surfaces, dynamo and motor 108 Swamps 57-8 Temperature of trees 36 Tissues, overcooling xxv Tornado injuries to trees 25 Transpiration 33, 37 Treasurer, report xvii, xxix Trees and lichens 197 Trees and tornado 25 Trees of Kansas 68 Trees, temperature of 36 Tuberculosis xxiii Turgor 34 Typhoid bacillus xx Ventilation xxii Water filtration xxii Water plants 57 Weeds 67 Wet leaves injured 26 Wind pressure effects 1, 25 Wood, abnormal 33-4 Wood drying 38, 40 Xerophytes 57, 60, 65 224 Trans. Acad. Sci. of St. Louis. INDEX TO GENERA. Abies 37, 190 Abronia 63 Acacia 176 Acer 26-7, 34, 39, 68-9, 172, 181, 188, pi. 4, 6, 8,9,11 Acorus 73, 164-5 Actinella 59, 61 Actinomeris 69 Adiantites 163 Aegilops 217 Aesculus 68 Agrimonia 69 Agropyrum 66, 217 Agrostis 205, 207, 217 Ailanthus 177 Aira208, 217 Alasraodonta 75 Alisraa 59 Allium 64 Alnus 166 Alopecurus 205, 207, 217 Amarauthus 62 Ambrosia 63-4, 66, 69 Amelanchier 176 Ammannia 59 Ammophila 62, 205/.207, 217 Amnicola 93 Amorpha 63, 68 Amphiacbyris 60-1 Amphicarpaea 68 Amygdalus 176 Ancylus 90 Andrena 46-7 Andromeda 171 Andropogon 63-4, 200 (203), 217 (220) Androsace 66 Anemone 65 Anodonta 75 Anodontopsis 76 Anteunaria 69 Antholithes 177 Anthorauthum 205, 217 Apios 68 Aplexa 91 Aplopappus 62, 64 Apocynophyllum 170 Aralia 171 Arisaema 69 Aristida 64, 205, 217-9, pi. 18 Arrhenatherum 200, 208, 217, 219, pi. 17 Artemisia 62-4 Arundinaria 216-9, pi. 17 Arundo 209, 217 Asarum 69 Asclepias 62-5 Asclepiodora 65 Asimina 68 Asplenium 163, 177 Aster 60-1, 66, 69 Astragalus 60-1, 64-5 Atriplex 66 Avena 208, 217-8, 220, pi. 19 Bacillus xx Banksites 169 Baptisia 65 Berula 59 Betula 166 Bidens 69 Bombax 177 Bouteloua 60, 64-5, 200, 208, 217, 220, pi. 19 Brachypodium 200, 209, 212, 217 Bromus 200-1, 209, 212, 217, 21», pi. 18 Buchloe 64-5 Bumelia 170 Bythinia 92, pi. 10 Cacalia 69 Caesalpinia (176-7) Calamagrostis 205, 207, 217 Calathea xxiv Callicoma (165) Calliopsis 48 (49, 50) Callirrhoe 65 Callitriche 58 Camassia 60-1 Index to Genera. 225 Campanula 69 Campeloma 94 Camptosorus 60 Cardamine 69 (or 59; Carex 59, 66 Carpinus 166 Carpites 171, 177 Carpolithes 177 Carya 68-9 (174) Carychium 88 Cassia 62, 176 Castanea 167 Catalpa 177 Ceanothus 61 Celastrinites 173 Celastrus 68, 173 Celtis 68, 168 Cenchrus 62, 200, 204, 218, 220, pi. 19. Cephalanthus 60 Ceratophyllum 57 Cercis 68, 176 Cereus 64 Chaerophyllum 69 Chamae&aracha 64-5 Chara 162 Chenopodium 64-6, 69 Chrysopogon 64 Chrysopsis 62 Cincinnatia 93 Cinna 69 Circinaria 79 Cladothrix 64-5 Claytonia 69 Clematis 61 Cleome 64 Cleomella 63 Cnlcus 64-5 Coix, 200, 217 Colletes 43 Colutea 177 Comraelina 63 Conobea 59 Conulus 81 Coreopsis 65 Corispermum 63 Cornus 61, 68 Corylus 68, 178 Crataegus 176 Croton 60-1 Cryptotaenia 69 Cucurbita 64 Cuphea 66 Cycloloma 62 Cynodon 209, 217 Cyperus 59, 62 Cytisus 176 Dactylis 200, 210, 217, 219, pi. 17 Dalbergia 176 Dalea 61 Delphinium 65 Dentaria 69 Deschampsia 207-8, 217 Desmanthus 67 Desmodium 62, 65 Dicentra 69 Diodia 66 Diospyros 170-1, 185, 188, pi. 12. Diplazium (163) Distichlis 66 Dodonaea 172 Draba 65 Eatonia 64 Echinacea 60 Echinocystis 68 Echinospermum 64-5, 69 Eleocharis 59 Eliraia 91-2 Ellisia 69 Elodea 58 Elymus 64, 200, 216-7, 219, pi. 1 8 Emphor (47) Engelmannia 64 Entechnia (47) Epeolus 51-2 Epilobium 59 Eragrostis 59, 62-3 Erigeron 63, 66, 68 Eriochloa 59 Eriogonum 61, 63 Erysimum 64 Erythronium 69 Eucera 47 Euchlaena 202, 217, 219, pi. 17 Eupatorium 59, 69 Euphorbia 60-3 Evax 64 Evernia 197 Evolvulus 64 Fagus 29, 178 226 Trans. Acad. Sci. of St. Louis. Ferussacia 87 Festuca 200, 209, 210, 217-8, 220, pi. 19 Ficus 169, 179, 188, pi. 12 Flaveria 66 Fontinalis 162 Fraxinus 69, 169-170 Froelichia 62 Gaillardia 63 Gastrodonta 81 Gaura 63 Gerardia 59 Geum 69 Gilia 61, 63 Gillia93 Gleditschia 68 Glyceria 209, 211, 217-9. pi. 1 7 Glyptostrobu8 163 Gnaphalium 62 Goniobasis (91) Grindelia 65 Gutierrezia 64 Gymnocladus 68 Gymnopogon 63 Halictoides 50 Halictus 44-5 Hedeoma 66 Hedera 171 Helianthus 60, 62, 65-6, 69 Heliotropium 60-1, 63 Hemihalictus (45) Herpestis 58 Heterotheca 6 Hibiscus 60 Hicoria 174 Hierochloe 217 Holcus 217 Hordeum 200-1, 212, 214-220, pi. 17,18 Houstonia 60 Hyalina (80) Hymenopappus 60, 63 Hypericum 65 Hypnum 162, 178, 188, pi. 12 Ilex 173, 182, 188, pi. 14 Ilysanthes 59 Indigofera 63 Ipomoea 62-4 Iris 178 Isoetes 163 Isopyrum 69 Iva 66 Juglans 68, 174-5, 183-4, 188, pi. 13, 14 Juniperus 61, 190 Jussiaea 58 Koeleria64, 209, 217, 220, pi. 19. Krynitzkia 63-5 Kuhnia 64 Lactuca 68-9 Lampsilis 77 Laportea 69 Leersia 59, 202, 217 Leguminosites 176 Lemna 58, 165 Lepachys 64 Lespedeza, 62, 65 Lesquerella 61 Liatris 62, 64-5 Limax 81-2 Limnaea 88-9 Linum 64-5 Lippia 59 Liquidambar 171 Lithospermum 60, 62, 65 Lolium 217 Lomatia 169 Ludwigia 59 Lycopus 59 Lythrum 59 Macreight a 171 Macrocera (53) Macropis 47 Malvastrum 64 Maranta xxiv Megachile 48 Melampodium 61 Melica 209, 217 Melissodes 52-4 Menispermum 68 Mentzelia 60-1, 63 Mimosites (176) 177 Mimulus 59 Monarda 62, 69 Morus 68 Index to Genera. 227 Muhlenbergia 69 Munroa 64-5 Myrica 165-6 Myriophyllum 58 Myrsine 170 Najadopsis 164 Nardus 209, 217-8 Nasturtium 58 Negundo 39, 68 Nelumbo 58 Nomada 61 Notholaena 60 Nuphar 58 Nymphaea 58, 73 Oenothera 59-61 Olea 170 Omphalina 80 Ophioglossum (163) Opuntia 64-5 Oryza201, 217-9, pi. 11 Osmorrhiza69 Ostrya, 69, 167 Oxybaphus 63-4 Palaeocarya 178 Paliurus 173 Palmocarpon 165 Panicum 59, 65, 200, 204, 217-220, pi. 18, 19 Panurginus 48-50 Panurgus (45) 50 Parandrena 47 Paronychia 61 Paspalum 62 Patula (82) Pellaea 60 Pentstemon 60-2 Petalostemon 60-1, 65 Peucedanum 60 Phalaris 205, 217-9, pi. 1 7 Philomycus 82 Phleura 205, 207, 217, 219, pi. 17 Phragmites 59, 209, 217, 219-220, pi. 19 Phryma 69 Phyllites 163, 166 Physa 90-1 Physalis 62-3 Pi lea 69 Pimelea 163, 169 Pinus 29, 164, 179, 188, \90,pl 13 Pisidium 79 Planera 168 Planorbis 89-90 Plantago 66 Platanus 26, 30,/. 1, 68, pi. 3, 5 Pleurocera 91 Poa 69, 209, 210, 217, 219, pi. 17 Podalirius 47 Podocarpus 164 Podogonium 177 Podophyllum 69 Polanisia 62 Polygala 64 Polygonatum Polygonum 66, 68-9 Polygyra 83-5, pi. 10 Polypteris 63 Populus 68, 167-8, 185, 188, pi. 15 Porana 170 Potamogeton 58, 164 Prosopis 43 Prunus 61, 63, 68, 178 Pseudopanurgus (50) Psoralea 61-2, 64-5 Pterocarya 175 Punctum 85 Pupa 86-7 Pycuanthemum 46, 69 Pyramidula 82 Pyrus 68 Quadrula 76-7 Qucrcus 29, 66, 68-9, 167 Ramalina 192, 196 Ranunculus 58, 69 Redfleldia 62-3 Rhamnus 68, 173, 183, 188, pi. 15 Rhus 61, 68 (171) 175, 184, 188, pi. 12 Ribes 68 Riddellia 61 Robinia 178, pi. 8 Rosa 68, 176 Rubus 68 Rudbeckia 69 Ruppia 58 Sabal 178 Sagittaria 69 228 Trans. Acad. Sci. of St. Louis. Salix 60; 68, 168, pi. 7 Salvia 65 Salvinia (163, 166) Sambucus 68 Sanguinaria 69 Sanicula 69 Santalum 169 Sapindus 172 Schedonnardus 64-5 SchraDkia 65 Scirpus 59, 66 Scolocloa 209, 217 Scrapta (47) Scrophularia 69 Scutellaria 61 Secale 200, 212, 214, 217-9, pi. 18 Segmentiua 90 Selenites (79) Senecio 60,64 Sequoia 164 Setaria 200, 204, 217-9. pi. 18 Sicyos 68 Silene 65 Silphium 65 Smilacina 69 Smilax 68 Solarium 64-5 Solidago, 64-6, 68-9 Somatogyrus 93 Sophora 64 Sorbus 29 Sorghum 203, 217-8, 220, pi. 19 Sparganium 59 Spartina 67 Sphaerium 78-9 Sphecodes 45 Sphenopterie 163 Spiraea 178 Spirodela 57 Spirogyra 90 Sporobolus 62, 64, 200, 207, 220, pi. 19 Stanleya 61 Staphylea68, 172 Stelis 48 Stenosiphon 60 Sterculia 172, 180, 188, pi. 14 Stillingia 63 Stipa G3, 65, 205-6, 217, 219, pi. 11 Strigula 197 Strobila (85) Strobilops 85 Strophitus 76 Strophostyles 62 Stylosanthes 65 Suaeda 66 Succinea 87 Symphoricarpus 61, 68 Tachea 85 Talinum 63 Taxodium 25, 164 Teucrium 69 Thalia xxiv Thaspium 43 Thelesperma 64 Thuites 164 Tilia 27-8, 39, 68, 172, pi. 9 Tillandsia 192, 196 Tmesipteris 163 Tragia 60 Triodia 62 Triosteum 69 Trisetum 217 Triticum 200, 201, 212-3, 217-220, pi. 18,19 Typha 59, 73, 164 Ulmus68, 168, pi. 8 Unio 76 (77), 78 Uuiola 69 U.-nea xxv, 189, pi. 16 Utricularia 67 Uvularia 69 Vaccinium 171 Vallonia 83 Valvata 93-4 Verbena 64-5, 69 Vernonia 66 Veronica 69 Vertigo 87 Viola 69 Vitis 68 Vitrea 80 Vitrina 82 Vivipara 94 Weiomannia 171 Widdringtonia 164 Wolffia 58 Index to Genera. 229 Xanthoria 197 Xanthoxylum 68, 175 Yucca 61 Zannichellia 58 Zanthoxylon 68, 175 Zea 200-202, 217-8, 220, pi. 19 Zinnia 61 Zizania 202, 217-9, pi. 17 Zizyphus 174, 182, 188, pi. 13 Zonites (80) Zygadenus 60-1 Issued January 30, 1899. imamSSJtSP^ Garden Librar 3 5185 00240 7144