aE aR pists



Vol. XXXVI > JULY; 19003

Ri ca cof OR OLE oe ee ae JOHN M. COULTER anp CHARLES R. BARNES,



Cleanliness of body was ever esteemed to proceed from a due reverence to God, to society and to ourselves.”


From the end of the 18 Century to the beginning of the 20”


has been popularly recognised as the clean and cleansing soap.

ETE rede esi ek







Pu ae University. ney) of Bonn. CASIMIR sama OLLE VOLNEY M. Spa

Geneva. Un niverity ‘of Michigan. J. B. DeTo ROLAND THAXTE

ee of Padua. Harvard University. ADOLF ENGLER, WILLIAM ——.

University of Berlin. Missouri "Botévieal Garden, LEON GUIGNARD, H. Mansman Wa

L’Ecole de Pharmacie. Uni oe me ‘Cambridge.

Rosert A. Ha EUGEN. Want

Un nversity rok Wisconsin. Uni ety of Copenhagen. JInzO MaTsuM VEIT WITTROCK,

Im cipievial University, Tokyo. Royal Academy of Sciences,





Mo. Bot Garden 1904. .

PRINTED AT The University of Chicago Press CHICAGO


PAGE On the grange and embryo of Taxodium. Contribu- tions from the Botanical of the opkins saat No. 1 (with plates I-x1) - W.C. Coker 1,114 Mitosis in Pellia. Contributions from the Hull Botanical Laboratory. XLIX (with plates x1I-xIv) - - C.J. Chamberlain 29 New Western plants. I - - - - - . = AyD BaBimer —52 Studies in spindle formation (with plates xv, Xv1) - - A. A, Lawson 81 The embryo sac of Casuarina stricta. Contributions from the Hull Botanical Laboratory. L (with plate xvi) - i. CG. Frye” Toi The vegetation of the Bay of Fundy salt and diked marshes: an ecological study. Contributions to the ecological gs geography of the Province of New Brunswick. o. 3 (with sixteen figures and maps) - W.F. Ganong 161, 280, 349, 429 Geographic distribution of /scetes saccharata. Contributions the Hull Botanical Laboratory. LI (with map) G. H. Shull 187 A. sketch of the flora of southern California - . - - S.B. Parish 203, 259 écological SS of ie flora of mountainous North Caro- i lina - - . y eS Harshterger 241, 368 _ Qdontoschisma Macounii and its North American allies (with plates xvI-xx) .W. Evans 321 On pa hence distribution and ecological relations of g plant societies of northern North America i dee maps) - - £. N. Transeau 401 Aralia in American paleobotany (with diagram) - - . £. W. Berry = 421 Notes on Garrya with descriptions of new speciesand key - Alice Eastwood 456 BRIEFER ARTICLES Positive geotropism in the ite of the isi alias (with one figure) - - £. 8B. Copeland 62 Contributions to the biology of Rhizobia - . - A. Schneider 65 The occurrence of two venters in the archegonium of Polytrichum juniperinum (with one figure) - - Mary C. Bliss 141 Polyembryony in Ginkgo > = = - - M.T. Cook 142 A gall upon a mushroom (with six figures) é - C. Thom 223


' shoot. The curv

- immensely


PAG. Selected notes. II. Liverworts (with five figures) - WC. Coker 225 A new species of Geaster (with two figures) - - G.F. Atkinson 303 Tilletia in the capsule of bryophytes - = - - B.M. Davis 306

Two megasporangia in Selaginella (with one figure) - Florence M. Lyon 308

The mitoses in the spore mother-cell of Pallavicinia

(with six figures) - A.C. Moore 384

Is Detmer’s experiment to show the need of light in starch-making reliable ? (with two figures) - - Bernice L. Haug 389

The transpiration of Sfartium junceum and other xerophytic shrubs (with two figures) - - . J. Y. Bergen 464 Geaster leptospermus: a correction - - - - G.F. Atkinson 467 CURRENT LITERATURE— - - - - - - 68, 143, 231, 309, 392, 468

For titles see index under author’s name and Reviews. Papers noticed in Notes for Students” are indexed under author’s name and subjects.

News— - - - - - - - - —- 79, 159, 238, 317, 399, 479

‘DATES OF PUBLICATION. No. 1, July 16; No. 2, August 15; No. 3, September 15; No. 4, October 15; No. 5, November 15; No. 6, December 19.


P. 62, last 2 lines for “which curved where it grew. Likewise in both root and shoot the tice se ag inaaia curved where it grew, likewise in both root and


2, line ; ta elon, for de read der.

A a line 21 from below, for nucleus read nucleolus.

. 159, line 15 from below, for Bogoniensis read Bogoriensi

. 173, legend under fig. 4 after section insert through; wd add, Vertical scale magnified.


P. 177, legend fig. 5, add with its clapper. : 222, line 5 from below, for Pinos read P P. 253, footnote 18, line 3, for archceology eae csichaciinnre P. 306, line 14, for 4.5 read 4.5™™. P. 306, line 22, for 3.5 read 3.5 ™™, P. 327, line 6, for is read are. P. 377, line 10, for comprises re: P. 377, line 18, transfer first pies and, to beginning of line 19. : 378, line 6 from below, for atternifolia read alternifolia. P. 396, line 20, for Jurrassic read Jurassic.

Botanical Gazette

A Montbly Ayal Embracing all Departments of Botanical Science Subscription per year F le rs, 40 Cents The subaeebaiin par must be paid in sileacalae No numbers are sent after the penis

of the time paid for. FOREIGN AGENTS: Great Britain WM. WESLEY & SON, 28 Essex Continental Europe —GEBRUDER BORNTRAEGER, St., Strand, London. 18 Shillings 6 pence. Berlin SW. 46, Dessauerstr. 29. 19 Marks Vol. XXXVI, No. 1 Issued July 16, 1903





29 NEW WESTERN PLANTS. I. 4. D. £. Elmer. 52 | elascinagl ARTICLES.


4 BOOK REVIEWS - = - - 68


MINOR NOTICES - - - - - 69 NOTES FOR STUDENTS = a - - - - 70 NEWS - - - - - - - - - - - - 79

Separates, if desired, must be ordered in advance of publication. ng articles will be pri nted, of which 25 (without covers) will be furn er (and covers, ‘f desired) to be paid for by the author. i be e€

Not less than 50 separates of lead- gratis, the actual cost of the

Seaatah es of “briefer articles” (with or

th vers) w supplied at cost. The tabl tite hows the approximate cost of separates

sonsisting of plain text or text with line maselings The actual cost may vary from the figures given and will depen t f


upo) k in poets: the pages into forms, press work, paper, binding, ttc. Separates containing half-tones may be expected to cost somewhat more than the rates given, the ncrease depending upon the number of cuts and the amount of work required upon them.

4 Number of copies 50 Too 150 200 t-press, for 4 pages orless. . . . . $1.60 $2.90 $2.25 $2.50 rete. press, for 8 pages or less. eye tt 2.25 2.75 $55 3.50 r-press, for 16 pagesorless. . . . . 4.00 5.00 5.80 6.50 single plates (1 double = 2 single) . 1.00 1.35 1.70 2-00 overs, with title (paper like GAZETTE cover). 1.50 2.00 2.50 3.00

scripts. Contributors are requested to write scientific and proper names with -ioppeamsi is in p pver to follow the form shown in the pages of the Gaze ETTE. Manuscripts should be se Gazette, The Uni Chicago, Ii, Coke Pamphlets for Review should be sent to the same add

tambers will be replaced free only when claim is made within thirty days after receipt of the yumber following.

Foreign Subscribers.— The attention of foreign subscribers is called to our special rate, cessitated by the payment of extra postage. Until further notice the prices as indicated above should remitted to our foreign agen All re mittances should be made payable to the order of The gies tie oaed of epee should be aparcomt to University of Chicago Press, Chicago, Ill.

[Entered at the Post-Office at Chicago, IIl., as second-class mail matter, ]

Every Botanist

Should be familiar with the prominent works of



Die Glykoside. Chemische Monographie der Pflanzengly koside nebst systematischer Darstellung der kiinstlichen Glykoside von Dr. I. I. L. von Ryn, Director der Reichsver

hsstation in Mastricht. 8vo. Cloth, $2.50.

Das Werk giebt—wie es bisher noch nirgens geschehen—eine eingehenh chemiscthe Behandlung der Glykoside —nicht nur eine kurzgefasste Zusammen stellung der chemiscthen Eigenschaften dieser Korperklasse, sondern die Darstel lungsmethode, die Griinde, welche zur Aufstellung der Constitutionsformel gefiihrt haben etc., so dass das Buch in chemisch-pharmaceutischen wie pharm@ hologischen Kreisen sowie unter den Studirenden und sonstigen Freunden da phytochemischen Forschung sicher mit grosser Freude begriisst werden wird.

-Die Harze und die Harzbehalter. Historiscr kritische und expermentelle, in Gemeinschaft mit zahlreichet Mitarbeitern ausgefiihrte Untersuchungen von Prorrssor DR. A. Tscuircu, Director des pharmaceutischen Institutes der Uni versitat Bern. Mit 6 Tafeln. 8vo. Half calf, $5.00. |

Das Werk stellt zum ersten Mal das gesammte Material dieser wichtig@® Gruppe von Phlanzenproducten kritisch durchgearbeitet dar. Die streng wissen sthaftlichen Undersuchungen werden auch fiir die Praktiker, besonders fir ai aie sich mit Harzen und Harzproducten beschiiftigen, von Interesse sein, da jeu

rationelle Technik ja auf wissenschaftlicher Grundlage ruht.

Write for free Catalogue; postpaid. Address:

Gebrider Borntraeger, Publishers, | BERLIN, S. W. |


2 2s.

Tus, 4to, cloth,


onyms, and pl c “Ba shiner, Ww. Soma 0- 1866



BOLUS pnt , ICONES ORCHIDEARUM AUSTRO agg a de ae EXTRA TROPICARUM; or fgeeet s sig h,

ptions of extra tropical South African orchids, Vol. JACKSON _ GUIDE TO ag ail caieatae OF BOTANY, including nearly 6,000 titles not given in Pritzels’

sigh cag V lbagemeightie ger a Spaces te towards a bibliography of economic botany, with a compre-

raed nsive subject-index, 4to, cloth, 1882, MILLS Sole = eee DIATOMACEAE OF THE HULL DISTRICT with illustrations of 600 species, SANDER’ s ‘oncmD ‘ure, all the best known species, varieties, and s of orchids in cultivation, h the names and parentage of all the known hybrid orchids up to fg 5 ah I, 190t, roy. 8vo, half bes i gt, SIM se Ph ban ERNS OF SOUTH AFRICA, containing descriptions and figures of the ferns and fern allies 1S.

outh Ainicn with localities, cultural agi etc., with rsq plates, 8vo, half bound, 1892, £1

life i collection, cultivation and JB beta of diatoms, and a iexieive and figure ty pical o

known genus, as well as a description and figure of a erg fou = in hoy North $

it, Translated by W. E. Baxter, 36 plates and engra s (2 gu

FOR nto tse bt ) and J. E. SMITH—-ENGLISH BOTA ; OF aces owth, wi oa ral indexes and supplement (in five vols.) by sh enggay the by J. de C. Sowerby and os alter, 2,998 colored plates, 41 57 15S.

roo colored and plain plates, 8vo, clot

E, containing introductory remarks on the every é

ering &

s) imp. ira, ate, 1896, £2.

lored fi f British plants, with their essential


28 Essex Street, Strand, London


Light Waves and Cheir Uses

Eight Lectures Delivered Before the Lowell Institute in 1899

segigiel oe by 108 Len coy 2 Colored Plat

By ace A. MICHELSON Professor and Head of the Department of ioc Director of sina Physical Laborato

Issued as Volume III of the Second Series of the Decennial Publications of the University of Chic: cago, pp. xii = 8vo, cloth » net, $2.00; postpaid, $2.15

Che University of Chicago Press




New Apgar Fo a fk ny we Ne Dissecting Mic pe, Apla $4.00 Complete Studert’s tie Tose oe gion a en to $22 00; Beck’s New Laboratory Micro te vn uced t © $36. 00. ‘Special (duty free) pe chools.

nieal Supplie

, Ma: ee eae Sides Agents for E. Leitz R. & J. Beck’s IAMS, BROWN & EARLE,

WILL Dept. 17, 918 Chestnut Street, Philadelphia,

Che University of Chicago Press

ee printing _ for academic work, including theses reports of educational bodies and ieties.

Educational and i works printed in eer German oe Hebrew, and pared on nguages

Byte mat es Woes ts tee

Ghe ee ef C eee alate @TH ST. aNd Ets Ave., CHIC

Insured for One Million Dollars


From the

New-Dork Daily Cribune.

e Prudential Insurance Company, through its Philadelphia agents, has insured L. Rodman Wana-

> WwW one ily insured priv ate individu —, in The world, as he carries doors 90 for $2,000,000. His insurance ‘is said to be e ee only by that of King Edward VII. John Wa aker. his fog earries policies ag- eregating "31 300. 000. and John M. Mack carries $1.-

250.000. of which $1,000.000 win placed recently.

Mr. Wanamaker’s Selection of The Prudential is a Striking Demonstra- tion of —The High Esteem in which This Company is held by Farsighted Business Men, and Their Great Confidence in The Prudential’s— Permanent Strength, and Able and Conservative Manage- ent,—Both Conducing to Profitable Dividend Returns on its Policies. Whether the Policy be Large or Small, the Same Progressive and

Conservative Management is Back of All Send Coupon for Specimen Policy, Suitable to Your Means


myself to any action, I shall be glad to receive, free, particulars and rates of policies for $


Oceupation “uke ¢ ¢HAS THE °


THE PRUDENTIALA GIBRALTAR } Insurance Co. of America #@ N

John F. Dryden, prest. Home Office, Newark, N. J.



JULY, 1903


In spite of the recent great increase in our knowledge of spermatogenesis in many groups of gymnosperms, this part of the life history of the Taxodieae remained, at the time this work was undertaken, almost unknown. A short contribution by Shaw (’97) on Sequoia had appeared in 1897, and Arnoldi (’99”>) has recently added two papers on the development of the reproductive organs in Sequoia. These observers have cleared up many salient points in the development of this genus, but the group as a whole is still to be studied.

Taxodium itself, probably on account of its limited geograph- ical distribution, has been greatly neglected by investigators. Coulter on the histology of the leaf, Masters on the seedling, Lotsy and Meehan on the knees, and Von Schrenk on the dis- ease called ‘‘peckiness’’ are among the few papers that have been devoted, in whole or in part, to the study of Taxodium, and none is concerned with the development of the seed.

The present work was suggested by Dr. D. S. Johnson, to whom I wish to express my gratitude for his unfailing kindness and helpful advice throughout its prosecution. I also wish

‘A dissertation submitted to the Board of University Studies of the Johns Hop- __ kins University, June 1901, for the degree of Doctor of Philosophy.

: I


publicly to thank my brothers for their assistance in sending me material at frequent intervals.


Collections of Taxodium distichum Richard, the only species studied, have been made for about three years, chiefly from Hartsville, S. C., but also from Baltimore, Md., and New Berne, N.C. Fixing has been done at the tree in all critical stages, but fresh material, sent in tight boxes from Hartsville to Balti- more, has frequently given good results. Flemming’s strong solution, chrom-acetic acid solution, alcoholic solution of picric acid, saturated solution of corrosive sublimate in 95 per cent. alcohol have all been used to some extent; but a saturated aqueous solution of corrosive sublimate (95 or 90 parts) and glacial acetic acid (5 or 10 parts) has been generally used. The latter gives results that are scarcely, if at all, inferior to those obtained with the Flemming solution, while it is more satisfac- tory than any of the other fluids mentioned. In searching for protoplasmic connections between cells, Gardner’s (83) methods were used, but only with fixed material. Potassium iodid and chlor-zinc-iodid were useful in determining the presence of starch, and have been used throughout for this purpose. A number of stains have been tried, but Flemming’s triple has been most used. Young cones were split, or the scales removed entire. In older cones the ovule was removed and the nucellus exposed by breaking off the lignified tip of the integument, or the whole prothallium was take from the seed. Sections 5-10 # thick were made by the usual paraffin method.


The staminate flowers are born on short branches which are either simple or compound. If simple, these branches are usually longer and more numerous than if compound. They appear in the fall from near the tips of the branches of the same year, and at the beginning of October or even earlier the young staminate flowers may be seen in the axils of their scale-like leaves. A longitudinal section of a sporophyll at this time shows no distinction between primary archesporium and other


tissue, all the cells of the lower part of the sporophyll being of about the same size, and having dense contents. Soon, however, cer- tain centrally placed hypodermal cells begin to divide by peri- clinal walls and give rise to rows of cells as shown in fig. 7. The outermost cells of these rows, by a periclinal division, form the one-layered tapetum and the inner layer of the sporangial wall. By division of adjoining cells the tapetal layer is extended completely around the sporogenous tissue (fig. 2), and by Janu- ary, or earlier, the microspore mother-cells are formed and ready for their division in early spring. Chamberlain (’98) has reported a similar stage during winter in the microsporangia of Pinus, Cupressus, and Taxus. The cells of the whole sporophyll, with the exception of the tapetum and the sporogenous tissue, con- tain starch through the late fall and winter until renewed growth in spring alters its arrangement. In the middle of November the nuclei of the tapetal layer show a peculiar structure not found at other times. They havea very coarse and wide-meshed reticulum, upon which the chromatin is distributed in large granules of very unequal size. There is no nucleolus. The nuclei of the sporogenous tissue have several nucleoli and a thinner reticulum than at a later stage.

No trace of an indusium-like outgrowth from the sporophyll is present for the protection of the sporangia, such as occurs in Cupressus, Thuja, and species of Juniperus. During early stages of development the cells of the upper part of the sporophyll are completely filled with a peculiar homogeneous substance staining bluish with gentian, which, as its subsequent history shows, is either a form of starch or an intermediate product in the forma- tion of starch. It is not stained blue by iodin. At the stage of fig. z, this substance is being replaced by starch grains of the usual kind, and a direct relation in amount between the two is evident, the starch appearing in proportion as the amorphous substance disappears. The cells on the line of transformation contain both starch and amorphous substances in proportionately smaller quantities.

Before their division in the spring, the pollen mother-cells become filled with starch, while the grains in other parts of the


sporophyll are being rapidly corroded. The persistence of this starch in the mother-cells during division and its disappearance as the exine is formed in the pollen grain agrees with what is already known in cycads and conifers. The ripe pollen grain contains no starch, nor is any found in the pollen tube until it appears in the protoplasm of the central cell shortly before the formation of the sperm cells. The number of micro- sporangia ona sporophyll may be as many as nine, seven being a common number. The wall of the mature microsporangium consists of but two layers of cells on the exposed surface, and in this respect Taxodium differs from the Abieteae, Taxeae, Cycadales, and Ginkgo, and agrees with the e Cupresseae and Gnetales. The cells of the outer layer of the wall have the sides and inner faces strengthened by bands of cellulose, while those of the inner layer are very much flattened and poor in contents. The cells in the tapetum have very dense contents and are shorter and thicker than those of the inner wall. They disorganize at about the time that the division occurs in the pollen grains.

The division of the pollen mother-cells took place this year (1901) in South Carolina on March 6th, and both divisions were found on the same day, even in the same cone, but the stages found in the same sporangium are not quite so different as Coulter and Chamberlain (’or) figure for Pinus Laricio. Changes of the nucleus leading up to the first division were not present in my material, but good preparations of all stages during and subsequent to the metaphase of the first division show that the phenomena are similar in all essential respects to those described in detail by Strasburger (’00) for Larix.

The chromosomes, as arranged on the nuclear plate, are short and thick (fig. 3). They stand at right angles to the axis of the spindle, the fibers being attached to the inner ends. The splitting begins at the point of attachment and in favorable cases the line may be seen between the two halves in the as yet unseparated outer limb. Very soon after the splitting is com- pleted and the daughter chromosomes begin to move to the poles, the fibers are seen to be attached to the middle of the

Ce ae ee ee Te eee te et Cael Bee eee |e ee ae rene


bent chromosomes and the inner ends of the latter are com- posed of four arms, lying side by side, and generally of the same length (jigs. 4-6). The second splitting has evidently occurred and the arrangement is now just as in Larix as described by Strasburger ('00). The chromosomes remain thick and short as they approach the poles, and their number can be easily determined to be either eleven or twelve. Eleven are shown in fig. 6 in polar view, and at least this number could be distinctly made out in other cases. Sometimes there seem to be twelve, but on account of the crowding in such cases I have never been sure of this number. Twelve chromosomes have been found by Belajeff (’94) and Strasburger (’92) in the pollen mother-cells of Larix europaea, by Blackmann (’98) in pollen mother-cells and oosphere of Pinus sylvestris, by Juel (00) in the megaspore mother-cell of Larix sidirica, and by Chamberlain (799) in the pollen mother-cells, endosperm, and jacket cells of Pinis laricio. It would thus seem from analogy that the number of chromosomes in the pollen mother-cells of Taxodium is also twelve rather than eleven.

The daughter nuclei (fig. 7) before the next division enter into a fairly well-developed resting stage. There is a distinct reticulum, if indeed a rather coarse one, and the chromatin is grouped in larger masses than in the reticulum of many resting cells, approaching more nearly the condition already described in the nuclei of the tapetal layer of the microsporangium in Novem- ber. Strasburger (’00) describes such a condition in Larix, but tries to bring it in harmony with other cases by considering the network as spun out from the chromosomes. His distinction is

not clear to me, and I think it must be acknowledged that the

daughter nuclei of the first division may, at least in some cases, reach before their next division a relatively well advanced rest- ing stage. From fig. 7 it will be seen that the cell walls of the mother-cell have not disappeared at the time of tetrad-formation. In places the walls have begun to go to pieces, but in others fFemain entire and in close contact with their neighbors. No case was found where the final divisions were bilateral, as is some- times the case in Pinus Laricio (Coulter and Chamberlain, ’01).


The connecting fibers of the first spindle produce a distinct cell-plate, extending entirely across the cell before the nuclei have begun to divide a second time. -On each side of the plate the starch grains are densely crowded. The chromosomes of the second division are single slightly curved rods, and are evi- dently of about the same size as the halves of the double chro- mosomes of the first division. The starch begins to disappear during the second division of the pollen mother-cell, and is com- pletely used up during the formation of the exine of the pollen grains, which becomes quite evident in about three days after the last division. The nucleus of the fully formed but yet undivided pollen grain is evenly and coarsely granular and gen- erally without a nucleolus (fg. 8).

About ten days after its formation the pollen grain divides. The spindle is very small and the chromosomes are propor- tionately longer than in the reducing division (figs. g and 70). This is the only division of the pollen grain, no sterile prothal- lial cell being formed, and it separates at once the generative cell from the tube cell. The former is flattened lens-shaped, concave toward the inside, and furnished with a distinct //aut- schicht ( fig. 11). This division occurs a few days before the pollen is shed, and it is in this condition that the ripe pollen reaches the nucellus (fig. 72). In the absence of any sterile prothallial cells, Taxodium agrees with the Capresseat and Taxus, and differs from all other conifers and’ “cycads. The number of sterile prothallial cells in the pollen grain of gym- nosperms has been determined in the following cases: two in Ginkgo (Strasburger, ’92), Larix europaea (Strasburger, 84), Picea vulgaris ( Belajeff, 93), Pinus silvestris (Strasburger, ’92), Pinus Pumilio (Coulter and Chamberlain, O01); one in Ceratozamia

(Juranyi, 82; he occasionally found two in C.. longifolia), Zamia (Webber, ’97), Cycas (Ikeno, ’99); none in Biota, Cupressus, Juniperus (Strasburger, ‘02h Taxus baccata and Juniperus ( Bela- jeff, ’93).

The great importance of correctly determining the number of divisions in the pollen grain has not been overlooked, and repeated sections, at all stages of the development of the pollen

eee ee rs pretence

se a Bia

eer ity CAPa Wy =O ae dE Ey


from the mother-cell stage to the sprouting of the pollen tube, have been made from collections obtained in both 1900 and 1901, and I think it certain that there is but one division of the pollen cell in Taxodium. |


The first indication of sprouting is given by the swelling up of the generative cell into the tube cell, and by an increase in size of both nuclei (fig. 73). The exine is usually thrown off at an early stage, as shown in fig. zg. In this figure the nuclei of the pollen tube have not changed their position, the tube nucleus lying immediately above the generative cell. The pol- len tube contains no starch, either now or during its course to the prothallium. As the tube advances, the tube nucleus moves from its position over the generative cell and passes slowly down toward the tip. Indications of branching are soon seen in the pollen tube (jigs. 17, 29, 22, 23). In fig. 76 the generative cell seems by its position to be bounded by an actual membrane, but no indication of a cellulose wall was obtained, and if one is present it is exceedingly thin and quickly dissolved. By com- paring figs. 75 and 76 it will be seen that the sprouting does not take place at any definite point in reference to the position of the generative cell.

The division of the generative cell does not occur until sev- eral weeks after the sprouting of the grain (figs. 19-21). The stalk nucleus soon loses its definite hold upon the protoplasm around it, although immediately after the division (fig. 79) it is

" still bounded by a distinct protoplasmic sheath. The central cell

retains the characteristics which mark the generative cell before division. It is furnished with a distinct Hautschicht and has the Shape of a double convex lens. It will be noticed that imme- diately after the division the stalk cell is larger than the central cell. Belajeff (’91, ’93) describes these two cells as being of €qual size in Taxus. In Juniperus communis he (’93) finds the outer cell to B& smaller, while in Picea vulgaris the opposite is true. There is not much difference in size in Pinus Laricto, as figured by Coulter and Chamberlain (grt). It will thus be seen that Taxodium agrees with Juniperus in the relative size of the stalk and central cells immediately after their formation.


Belajeff (’91, ’93) describes the stalk nucleus as passing the central cell as it wanders down the tube. Such a description could hardly be applied in Taxodium when the tube is at right angles to the axis of the spindle of the generative cell. The stalk nucleus is as near the tip of the tube as is the central cell, and they both wander down together until they reach the tube nucleus (fig. 22). It will be seen from fig. 26 that the three nuclei of the pollen tube can easily be distinguished at this stage. The stalk nucleus is smaller than the tube nucleus, while the protoplasm of the central cell is distinct from that of the pollen tube. The stalk and tube nuclei now advance slightly ahead of the central cell (fig. 23), and this relative position is retained by the three nuclei throughout the subsequent history of the pollen tube. In fig. 23 the stalk nucleus is still slightly smaller than the tube nucleus, but the structure of the two is the same. The male nucleus is very like the other two, its nucleo- lus being slightly smaller. _ The pollen tube proceeds to the prothallium without inter- ruption; the growth, however, is much slower in the upper part of the nucellus than in the lower. No particular tissue of the nucellus tip is set apart to nourish or guide the pollentube. All of its cells contain more or less starch, but there is no grouping _of starch in definite areas. The pollen tube may reach. the megaspore before the formation of a cellular prothallium (fg. 25). So early an approach of the pollen tube to the sprouting megaspore has not been described in any other case, so far as I am aware. Jager (’99) gives one figure of Taxus baccata showing a pollen tube almost in contact with a young prothal- lium, and I have found that in Taxus baccata canadensis the pollen tube may reach the level of the megaspore before the latter has divided even once. One case was found in this plant where the pollen tube has grown against and badly compressed -the megaspore before the latter had advanced far beyond the sixteen-cell stage. It was so completely crushed that the stage could not be exactly determined.

Fig. 26 gives the structure of the contents of the pollen-tube at a slightly later stage than fig. 25. The two free nuclei are

feo ae


now exactly similar and lie side by side immediately beneath the central cell. The latter has increased greatly in size, as has also its nucleus, and the protoplasm is seen to possess a radiate structure. We find in the nucleus of the central cell a distinct peripheral network, and a nucleolus, irregular in outline and evi- dently of a compound nature. This kind of nucleolus, which we here meet for the first time, will be found to occur also in the nucleus of the central cell of the archegonium. In one case the central cell was at the tip of the pollen tube, with the two free nuclei behind it. One of the latter was pressed so closely to the protoplasm of the central cell as to indent it slightly. Such an abnormal relation between the generative and free nuclei has been noted in Pinus Laricio by Coulter (’97).

The further changes in the central cell before its final division into the sperm cells are so remarkable and have been so neglec- ted in other conifers studied that I shall go into them with some detail. Fig. 27 represents the central cell after it has reached its full size. It is no longer spherical, but has become elliptical in section, the long axis being perpendicular to the axis of the tube. The protoplasm is seen to be radiating from the two poles of the long axis. At these poles are sometimes to be dis- tinguished slightly more granular areas, from which the radia- tions seem to diverge. The protoplasm is very dense, finely granular and in thin sections can be seen to have a reticulate Structure. The faint areas at the poles of the cells will at once Suggest in position the blepharoplasts of Ginkgo, Zamia, and Cycas. In reality, the resemblance is entirely confined to their position, Dr. Webber has kindly shown me his preparations of blepharoplasts in Zamia, and their intense staining and large size make further comparison impossible.

The nucleus, which is about half of the diameter of the cell, has rather abundant reticulum and a fragmented nucleolus. In addition to these, there has appeared a finely granular material Which does not seem to differ in any respect from the linin Material in the egg nucieus to be described below. It is most abundant around the nucleolus, but extends to all parts of the nucleus. In jig. 27 one of the free nuclei is seen closely appressed


to the Hautschicht of the central cell; the other free nucleus does not appear in the section. This is about the latest stage at which these free nuclei retain their normal structure. They _ very soon begin to go to pieces, and the protoplasm of the pollen tube at the same time begins to disorganize. It becomes more homogeneous and retains more tenaciously the safranin stain. The nucleoli and chromatin of the free nuclei become more or less broken up and collected into masses of different size, a process which we shall see corresponds exactly to what occurs in the nuclei of the jacket cells of the archegonium shortly before its final division. Concomitantly with the disor- ganization of the nuclei and cytoplasm of the pollen tube, there becomes evident in the cytoplasm of the central cell a number of bodies staining a deep red in safranin. They resemble exactly the plastin granules that we have seen to appear at the disor- ganization of the free nuclei, and that they are actually transferred from the latter into the central cell seems possible. ig. 28 isa central cell after the appearance of these granules. They are arranged in a circular manner at some distance from the nucleus, and it may be that this distribution is connected in some way with the concentric arrangement of the fibers. At the time of the appearance of the plastin granules in the protoplasm of the central cell, there seems to be a distinct connection at the base of the cell between its protoplasm and that of the disorganizing material beneath it. Hirase (’95) describes large bodies lying in the protoplasm of the central cell of Ginkgo between the nucleus and the blepharoplasts. Webber (’97) confirms this and says that in addition to the two large bodies smaller masses of similar material were observed in other localities of the cell. He speaks of them as extra-nuclear nuclein. It is easy to com pare these bodies with those of Taxodium. They stain deeply with safranin in both cases, the principal difference being that in Ginkgo they are generally fused into two large masses which © occupy a definite position in the cell.

The disorganized mass of nuclei and protoplasm at the tip of the tube never completely disappears before fertilization, and it may appear in the tip of the archegonium above the protoplasm

Re ee Fg et eo ee a


of the egg after fertilization. In fig. 28 a number of scattered Starch grains have made their appearance in the cytoplasm of the central cell. They retain their scattered position until finally arranged into the dense starch sheath immediately surrounding the nucleus of the sperm-cell.

Changes in the nucleus preparatory to the final division of the central cell had already begun at the stage represented in fig. 27. In fig. 28 these changes have proceeded still further. The chromosomes are being prepared from the few thick con- spicuous threads that are present. The linin granules have become organized into a reticulum, and this reticulum seems to be arranging itself as if in the preparatory stages of spindle formation.

No attempt was made to study in detail all stages of spindle formation in the division of the central cell, but in fig. 29, which shows an oblique view of the spindle, the formation of its fibers from the nuclear reticulum and the granular nature of the more peripheral fibers seems evident. This formation of the spindle from the fibers of the nucleus will be described in more detail in the division of the ventral canal cell. Fig. 30 shows a late telophase in the division of the central cell, the connecting fibers still being evident. A clear area is noticed on each side of the cell plate, and this area later extends entirely around the sperm cells, The starch and plastin material are collected at the distal ends of the spindle, but after the separation of the two daughter cells the starch becomes arranged in a dense sheath immediately surrounding the nucleus (fig. 37). Just outside of this sheath the plastin granules form a more or less complete layer. Beyond them is found the clear area previously mentioned, and the surface is composed of a distinct membrane which sharply defines the sperm cells from the protoplasm of the tube. After the formation of the daughter nuclei, they again begin to fill with the linin granules or reticulum (the so-called metaplasmic

substance of Strasburger ) until, at the time of maturity, they are

So dense as to make any distinction between the granular materia] and chromatin reticulum very difficult. A small nucleolus, however, can be dimly discerned (fig. 37).


The sperm cells are now mature, and fertilization almost immediately takes place. I think it probable that the sperm cells do not round themselves off completely until after the bursting of the pollen tube, for although sometimes separated as much as a quarter of their diameter from each other, I have never seen them while still in the tube without a flat face on the inner side. This remarkably complex structure of the sperm cell dis- tinguishes Taxodium from any phanerogam hitherto described, with the exception of Ginkgo and the cycads.

Recent work on the conifers, which in the structure of the male gametophyte approach Taxodium, gives very little detail as to the structure of the sperm cells. In Taxus Jager (’99) mentions radial striae in the periphery of both the sperm cell and its smaller, functionless sister cell, but gives no further details of the protoplasmic structure. Arnoldi (’98) says that in Cephalotaxus the protoplasm of the sperm cells, which are here of equal size, is densest around the nucleus. In his work on Sequoia (’99) he gives no details of the protoplasmic struc- ture of the sperm cells, but says they resemble those of the Cupresseae. Blackman (’98) makes some interesting observa- tions on the sperm cells of Pinus stlvestris. Hesays: “It cannot be doubted that cytoplasm also passes over into the oosphere, for each generative nucleus in the pollen tube is clearly sur- rounded by its own layer of cytoplasm, as can be observed in the stage when the tube is already in contact with the oosphere.” Also, “it may here be noticed that small bodies staining deeply with fuchsin S may be observed in the generative cell proto- plasm.’ These, he says, resemble leucoplasts. ‘If leucoplasts are really present in the cytoplasm belonging to the generative cells, the general view that the male cell brings over no plastids to the egg appears to be directly contradicted.’’ This is the only mention I have seen made of a distinctive protoplasm belonging to the male cells in any of the Abieteae.

- Neither Belajeff (’91,’93) nor Strasburger (’79, ’84) describe the structure of the sperm cells of the Cupresseae in detail, but in gross structure they seem almost identical with those of Taxodium. Strasburger (°84) says that the pollen tube of Juni-


perus contains very little starch at time of fertilization, and thinks it therefore the more remarkable that the fusion nucleus should be surrounded by so much starch. From comparison with Taxodium, however, it seems probable that starch is also