■ : ■ ! ._J fa- I m THE BOTANICAL GAZETTE THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS THE CAMBRIDGE UNIVERSITY PRESS LONDON AND EDINBUBGH THE MARUZEN-KABUSHIKI-KAISHA TOKYO, OSAKA, KYOTO, FUKUOKA, SENDAI THE MISSION BOOK COMPANY SHANGHAI THE BOTANICAL GAZETTE EDITOR JOHN MERLE COULTER VOLUME LXVIl JANUARY-JUNE 1919 WITH EIGHTEEN PLATES AND FIFTY-Fr\'E FIGURES THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS ,0 < «/«■' Published January, February, March, April, May, June, 1919 Composed and Printed By The University of Chicago Press Chicago, Illinois, U.S.A. TABLE OF CONTENTS Hugo De Vries I Camillo Schieider 27 Charles Drechsler 65 Bunzo Hayata 84 Dean H. Rose 105 Charles Drechsler 147 Oenothera rubrinervis, a half mutant - - - - Notes on American willows. III. A conspectus of American species and varieties of sections Reticu- latae, Herbaceae, Ovalifoliae, and Glaucae Morphology of the genus Actinomyces. I - - Protomarattia, a new genus of Marattiaceae, and Archangiopicris (with plate I and three figures) - Bhster canker of apple trees; a physiological and chemical study. Contributions from the Hull Botanical Laboratory 246 (with ten figures) - Morphology of the genus Actinomyces. II (with plates II-IX) ------- Relation of minimum moisture content of subsoil of prairies to hygroscopic coefficient - - F.J. Alway, G. R. McDole, and R. S. Trumbull 185 Notes on North American trees. IV - - - C. S. Sargent 208 Possible correlations concerning position of seeds in the pod -------- Byron D. Hoisted 243 Embryo and seedling of Dioon spinulosum (with plates X, XI) ----- Sister Helen Angela Dorety 251 Development of Stropharia epimyces (with ten figures) - - - - - - - - W. B. McDougall 258 After-ripening and germination of seeds of Tilia, Samhucus, and Rubiis. Contributions from the Hull Botanical Laboratory 247 - - - - R. C. Rose 281 Notes on American willows. IV. Species and varie- ties of section Longifoliae ----- Camillo Schneider 309 Respiration after death (with three figures) - - A. R. C. Haas 347 Effect of anesthetics upon respiration (with seven figures) - - - A. R.C. Haas 377 VI CONTENTS [volume Lxvn PAGE Basis of succulence in plants D. T. MacDougal, H. M. Richards, and H. A. Spoehr 405 A coniferous sand dune in Cape Breton Island (with eight figures) ------- Embryo sac and embryo of Pentstemon secundiflorus. Contributions from the Hull Botanical Labora- tory 248 (with plate XII) ----- Structure, development, and distribution of so-called bars or rims of Sanio (with plates XIII-XV) Apospory in Pteris sulcata L. (with plates XVI, XVII, and four figures) - - - - - Hydrogen cyanide fumigation. Contributions from the Hull Botanical Laboratory 249 (with two figures) -------- Studies of some Porto Rican fungi (with plate XVIII) -------- Briefer Articles — Byron D. Halsted (with portrait) - - - - Hybrid perennial sunflowers - - - - - Relationships within the Rhodosporeae - - - George Francis Atkinson (with portrait) - Depressed segments of oak stems (with four figures) Importance of epidermal coverings (with two figures) LeRoy H. Harvey 417 Arthur T. Evans 427 /. W. Bailey 449 W. N. Steil 469 E. E. Clayton 483 Leo R. Tehon 501 Mel. T. Cook 169 T. D.A. Cockerell 264 Geo. F. Atkinson 266 H. H. Whetzel 366 /. W. Bailey 438 R. B. Harvey 441 Current Literature 93, 171, 268, 369, 445, 512 For titles of book reviews see index under author's name and reviews Papers noticed in "Notes for Students" are in- dexed under author's name and subjects DATES OF PUBLICATION No. I, January 18; No. 2, February 15; No. 3, March 18; No. 4, April 18; No. s, May 19; No. 6, June 21. VOLUME Lxvii] CONTEXTS vii ERRATA Vol. LXVI P. 468, legend fig. i, for 9, Abies canadensis read 9, Abies balsamea P. 469, line 10 from bottom, for Abies canadensis read Abies balsamea P. 524, line 7 from bottom, for Schimper read Schiffner Vol. LXMI P. 73, line 7 from top, for Litman read Lutman P. 82, line 18 from top, for contact read contrast P. 83, line 4 from bottom, for successfully read successively P. 149, line 14 from bottom, for 20 fx read 2 .0 fj. P. 160, line 20 from top, for XVI read XVII P. 164, citation 4, for 1891 read 1892 P. 186, line 6 from bottom, for brought read drought P. 216, last word line i, for virginiana read virginica Volume LXVII Number i THE Botanical Gazette Editor: JOHN M. COULTER JANUARY 1919 Oenothera rubrinervis, a Half Mutant - - - - Hugo DeVries i Notes on American Willows. III. A Conspectus of American Species and Varieties of Sections Reticulatae, Herbaceae, Ovalifoliae, and Glaucafe -----___ Camillo Schneider 27 Morphology of the Genus Actinomyces. I - - - Charles Drechsler 65 Protomarattia, a New Genus of Marattiaceae, and Archangiopteris Bunzo Hayata 84 (With Plate I and three figures) Current Literature Book Reviews ----------- 93 Fossil plants Notes for Students ---------- Q5 The University of Chicago Press CHICAGO, ILLINOIS, U.S.A. THE CAMBRIDGE UNIVERSITY PRESS, London and Edinburgh THE MARUZEN-KABUSHIKIKAISHA, Tokyo. Osaka. Kyoto. Fukuoka. Sendai THE MISSION BOOK COMPANY, Shanghai Volume LXVII Number 1 The Botanical Gazette A MONTHLY JOURNAL EMBRACING ALL DEPARTMENTS OF BOTANICAL SCIENCE EDITED BY JOHN M. COULTER With the assistance of other members of the botanical staff of the University of Chicago Issued January 18, 1919 The Botanical Gazette is published monthly by the University of Chicago at the University Press, S7SO Ellis Avenue, Chicago, HI. 1[The subscription price is $7.00 per year; the price of single copies is 75 cents. Orders for service of less than a half-year will be charged at the single-copy rate. 1[Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawaiian Islands, Philippine Islands, Guam, Samoan Islands, Shanghai. If Postage is charged extra as follows: For Canada, 35 cents on annual subscriptions (total $7.35), on single copies, 3 cents (total 78 cents); for all other countries in the Postal Union, 84 cents on annual subscriptions (total $7.84), on single copies II cents (total 86 cents). ^[Patrons are requested to make all remittances payable to The University of Chicago Press in postal or express money orders or bank drafts. The following are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. 4. Yearly subscriptions, including postage, £1 12s. 6d. each; single copies, including postage, 35. 6d. each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage, Yen 15.75 each; single copies, including postage, Yen 1.75 each. For China: The Mission Book Company, 18 Peking Road, Shanghai. Yearly sub- scriptions, $7.00; single copies, 75 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 84 cents, on single copies II cents. Claims for missing numbers should be made within the month following the regular month of pubUcation. The pubUshers expect to supply missing numbers free only when losses have been sustained in transit and when the reserve stock will permit. Deliveries are not guaranteed within the war zones of Europe, nor in other coimtries where transportation is disturbed by con- ditions incident to war. Business correspondence should be addressed to The University of Chicago Press, Chicago, 111. Communications for the editor and manuscripts should be addressed to the Editor of the Botanical Gazette, The .University of Chicago, Chicago, 111. Contributors are requested to write scientific and proper names with particular care, to use the metric system of weights and measures, and in citations to follow the form shown in the pages of the Botanical Gazette. Papers in excess of thirty-two printed pages are not accepted unless the author is willing to pay the cost of the additional pages, in which case the number of the pages in the volume is correspondingly increased. Illustrations are furnished without cost to author only when suitable originals are supplied. A copy of the suggestions made in the January munber, 1907, will be sent on appUcation. It is advisable to confer with the editors as to illustrations required in any article to be offered. Separates, if desired, must be ordered in advance of publication. A table showing approxi- mate cost of separates is printed on an order blank which accompanies the proof; a copy will be sent on request. ^_^^_ Entered as second-class matter August 21, i8q6, at the Post-Office at Chicago, Illinois, under the Act of March 3, 1879. Acceptance for mailing at special rate of postage provided for in Section 1130, Act of October 3, 1917, authorized ou July IS, 1918. VOLUME LXVII NUMBER NtW Vf>RK THE 8M?TArMCAL QaKOBN Botanical Gazette JANUARY igig OENOTHERA RUBRINERVIS, A H.AXF MUTANT Hugo DeVries In the spring of 19 13 in a culture of Oenothera ruhrinervis I noticed some young plants, the leaves of which were a little broader than those of the other rosettes. Although the difference was very small, I planted them separately and saw that the deviation did not increase until the time of flowering. The spikes, however, gave proof that the aberrant specimens constituted a type of their own, since the bracts repeated the marks of the primordial leaves, being broader and more flattened than in ordinary ruhrinervis. There were 7 specimens of the new form among a culture of 25 plants, all of which flowered in August. This indicates a percentage of about 30. In the following year the seeds of the new form gave a uniform progeny, whereas those of the normal specimens repeated the split- ting. Thereupon I studied their seeds and found that about one- fourth of those of O. ruhrinervis were empty, but almost every seed of the new t^pe contained a living embryo. On account of this very small but constant difference the new form was designated as mut. deserens} Evidently it might have escaped observation in previous years, the individuals simply being taken for weaker specimens of the type. I studied the progeny of as many self -fertilized specimens of O. ruhrinervis as were available, therefore, and found the new type among all of them, and as a rule in correspondingly high num- co bers. Different strains of ruhrinervis yielded the same result. I CD ' Zeitschr. f. Ind. Abst. 16: 262. 1916. 2 BOTANICAL GAZETTE [January If we should apply the principle of Bartlett concerning mass mutation, and that of Morgan concerning lethal factors to this case, as I have made use of them in explaining the secondary mutability of 0. grandiflora and O. Lamar ckiana,^ we would conclude that O. deserens is a mass mutation of O. rubrinervis, and as such is a repetition of the initial mutation which produced the O. rubrinervis from O. Lamarckiana in my garden. This initial mutation must have occurred in a sexual cell, which, after copulation with a normal gamete of 0. Lamarckiana, gave rise to a half mutant, O. rubrinervis. In other words, O. rubrinervis arose as a half mutant between poten- tial 0. deserens and normal 0. Lamarckiana. This half mutant, after artificial self-fertilization, must have produced a splitting into three types, exactly in the same way as this splitting can be observed in the half mutants of O. gigas nanella. Of these types two must be constant, but the third must repeat the splitting. O. deserens is one of the constant ones, whereas the other is assumed to be hidden in the empty seeds, containing a lethal factor just as in O. grandi- flora and O. Lamarckiana. The third type is the continuance of O. rubrinervis, and repeats the splitting in every generation. According to my view O. Lamarckiana produces yearly two kinds of gametes in consequence of a secondary mutability into velutina. These velutina are linked to a lethal factor, which kills them in the young seeds. If we assume that the mutation into deserens took place in the typical gametes, leaving the velutina unchanged, we would conclude that O. rubrinervis consists of two types of gametes, even as O. Lamarckiana, but that both of them are in a mutated condition. One is the new deserens, without lethal factor; the other is the old velutina, linked to a lethal factor. The result of self-fertilization is now easily explained; the copulation of deserens gametes among themselves must produce this form, that of velutina must give empty seeds, and the combination of the two types must repeat the rubrinervis with its splitting capacity. On the same basis the occurrence of twin hybrids may be explained, the deserens gametes giving the laeta hybrids; but here we have a considerable advantage over other instances of twin ^ Mass mutation and twin hybrids of Oenothera grandiflora Ait. Bot. Gaz. 65 : 377-422. 1918. i9io] DeVRIES— OENOTHERA RUBRINERVIS 3 hybrids, since both the constituents are available in pure condition for controlling crosses. Any cross which gives twins with rubri- nervis may be repeated with O. deserens and with my O. mut. velutina (O. hlandina). In the first case the result must be hybrids of the type laeta, in the second case hybrids of the form velutina, and the addition of these must simply duplicate the split progeny of the corresponding cross of O. rubrinervis. I have made these crosses in a number of cases and found this deduction verified. Apart from the described secondary mutability into viable deserens and dead velutina germs, O. rubrinervis is not known to possess any noticeable degree of mutabiUty; it has, especially, never produced those mutants which are of so common occurrence in allied mutating forms. Thus we see that secondary mutability is not, in itself, to be considered as a cause of further mutations, and this seems to me to be a fact of paramount interest in the discussion concerning the probable causes of this phenomenon. The details of the following experiments will give proof of the proposed conception. I shall first give those relating to self- fertilizations and afterward deal with the crosses. Oenothera rubrinervis originated in my garden from 0. Lamarcki- ana quite regularly in a percentage of about o. i. Every time the visible characters were exactly the same. Between 1890 and 1900 the mutation was repeated 66 times among 66,000 plants.^ In 1905 I introduced new rosettes of O. Lamarckiana from the original locality near Hilversum into my garden, and among their offspring I observed also repeated mutations into rubrinervis. The charac- ters were always the same, namely, a pale reddish tinge, narrow and longitudinally folded leaves, a hairy epidermis, cup-shaped flowers, but above all the brittleness of the stems, branches, and petioles, due to the incomplete development of the cell walls in the fibers of the bark and wood. Until now I cultivated mainly two strains, derived from two different mutants of 1895. One of them has given the material for all my crosses, and I shall designate it as the main line. The second line of 1895 was originally destined for control experiments only, but in 1 913 it produced the first observed ^The mutation theory, English ed. 1:331. 1909. 4 BOTANICAL GAZETTE [january case of O. deserens, as previously mentioned, and since then it has been studied carefully in this respect. After repeated cultures of pure O. deserens had been made and compared with O. rubrinervis , the characteristic marks of the two forms became quite clear and reliable, although very small. In mixed cultures the types may even be separated when very young, but some dubious specimens may remain. At the time of flowering these have almost always been shown to belong to the new type. Among the very young rosettes, with only 3-5 leaves, those of deserens are broader and more flattened and of a deeper and purer green, resembling therein young plants of 0. Lamarckiana. These differences increase slowly until the time when the rosettes must be planted out from the boxes into the garden, about the middle of April. The leaves of O. deserens have now a broader base and a less pointed top than those of 0. rubrinervis, besides the marks already given. In July the differences remain very small, the two types reaching the same height at the same period, but the rubriner- vis begin to flower one or two weeks earlier than the deserens. Seen from above, the spikes show narrow, folded bracts in the first type and flat, broad ones in the second type. This character is easily appreciated and wholly reliable; no dubious cases trouble the counting in mixed cultures. The beginning of August, when the deserens have opened only a few of their first flowers, is the best time to separate them. A difference in the color has now become clear, the red tinge of the parent type failing in the deserens. Here the leaves, bracts, and flower buds are green, and the flowers are also of a purer yellow. A number of smaller marks, which are helpful in the distinction, almost escape description, as, for example, the form of the flowers, their grouping at the top of the spike, and the more erect position of the buds before opening. The main character of 0. rubrinervis is the brittleness of all its parts, as already mentioned. It is exactly the same in 0. deserens. I have broken the stems of all the plants to be mentioned in this article at the time of sorting them out in August or after harvesting their seeds, but no exception has been found to this rule.'' Hybrids -t I have, moreover, injected all the seeds for the sowings of the later years under a pressure of 8 atmospheres during 48 hours, this being the only reliable means of making the germination as complete as possible. I9I9I de vries—oexo thera r ubriner vis which possess the other characters of ruhrinervis, but lack the brittleness, are easily recognizable as rosettes of radical leaves as well as during the growth of the stems. They will be designated as subrobusta .^ The determination of the percentage of mutants in self-fertilized seeds was made in 19 16 in the following manner. The specimens were counted in the boxes at the time of planting out in April and the most undoubtful specimens of rubrinervis were counted and destroyed. All the others were planted out and tried at the time of flowering in August. By this means the space required for the cultures was reduced to about one-half of what would have been necessary if all the plants had been set out. Some losses were unavoidable and the percentage figures may be a little too small in ordinary cases. Only under favorable conditions do they come up to the amount of the theoretical expectation, namely, one-third of all the indi\'iduals. Since my question, however, was mainly to decide whether all specimens of rubrinervis split into this form and deserens, or whether there are also plants with a uniform progeny, I shall give the figures as I found them. Percentage of O. deserens among cultltres of O. rubrinervis Seeds of Parent number Number of specimens Percentage of deserens Mean Third generation in 1910. . . . I9I3--- 1913- ••• 1914.... 1914 1914 1914 Fourth generation in 1915 . . . . I9I5--- I9I5--- 1915 ••■■ I 2 3 4 5 6 7 I 2 3 4 60 59 60 59 60 59 60 89 60 79 89 10 20 22 19 25 19 15J i6' 25 13 19J . 19 18 In the first place, I shall now describe the main line of 0. rubri- nervis. Part of the seeds of the original mutant of 1895 had been preserved until 1905; they germinated sufficiently and gave the second generation. From this a third generation was derived in 1910, 1913, and 1914, and a fourth in 1913, 1914, and 1915. All the parents of these generations have been artificially fertilized by s Gruppenweise Artbildung, p. 143. 1913. 6 BOTANICAL GAZETTE [JANUARY myself. In 1914 I observed for the first time a specimen of deserens among these cultures, and this in a third generation. Thereupon I sowed in 19 16 the self-fertilized seeds of 11 specimens in order to decide whether all of them would repeat the splitting and to deter- mine roughly the percentage of the new type. The results are given on page 5. From these figures we see that all the speci- mens tried show the same splitting, and that this is always amass mutation. The countings for this table were partly made in the stadium of the young rosettes and partly at the time of flowering. In order to prove the correctness of this process, I repeated the sowings in 19 1 7 for those of the parents of which sufficient seed had been preserved, planted out all of their seedlings, and counted them in August, when they were ripening their first fruits. The results are as follows: Seeds of Parent number Number of specimens Percentage of deserens Mean Third generation in 19 13 ... . 1914 1914 1914 Fourth generation in 1915 . . . . 1915.... 2 4 6 7 I 4 48 55 59 53 58 57 15 24 13 13 14 19J • 16 Although the cultures were but small, they show that the devia- tions from the theoretically expected result (25 per cent) do not depend upon the method of counting as used in 19 16. In this race I self-fertiHzed the first mutant deserens observed in 19 14 and derived from it a second and a third generation in 19 15 and 19 16. The second generation consisted of 95 plants, of which 50 flowered ; the third was derived from two parents and embraced 77 and 140 specimens, among which 60 and 60 were left to flower. All of these cultures were wholly uniform at the time of planting out as well as during the flowering period. No rubrinervis and no new mutants occurred among them. Thus 0. deserens is seen to constitute a pure and uniform race. The percentage of empty grains among the seeds has been given elsewhere for this race of O. rubrinervis.^ The determination was made in the harvest of 5 plants of the third generation grown in ^ Zeitschr. f. Ind. Abst. 16:262. 1916. I9I91 DeVRIES— OENOTHERA RUBRINERVIS 1 9 10 and 19 1 5, and in that of two specimens of the fourth genera- tion of 19 1 5. I found 53-68 per cent of germs, with a mean of 60 per cent. Among the specimens of deserens, quoted in the same table, 5 belonged to this race; their seeds contained 96, 99, 94, 83, and 58 per cent of good germs. Thus we see that the empty grains, which are a character of 0. rubrinervis, have disappeared almost wholly in the new mutant. My second strain of O. rubrinervis was derived from another mutant of 1895. It has not been used for any crosses except those mentioned in this article, and which served as control for the experi- ments in the main Hne. Part of the seeds of 1895 were sowed in 1907 and yielded a second generation from which a third has been derived in 19 13 and a fourth in 19 14. I counted the deserens for three parent plants as previously described and found the per- centages as follows: Percentage of 0. deserens in cultures of 0. rubrinervis Seeds of Culture Number of specimens Percentage of deserens Mean Second generation in 1910. Third generation in 19 13. Third generation in 1913. 1913 1914 (A) 1914 (B) 25 70 70 19 The results agree exactly with those deduced from the previous table. The suspicion, however, that in the two last cases the per- centage figures were found too low, on account of losses of speci- mens of deserens at the time of planting out, induced me to repeat these sowings in 1916 from preserved seeds, giving them all the care which the previous cultures and the first ones of 19 16 had shown to be necessary. Moreover, in 19 13 I had self-fertilized a third plant, besides the two mentioned in the table, and also sowed its seeds. In this way I got in 191 6 the following percentages for the seeds of the three self-fertilized plants of 1913: Percentages of 0. deserens in cultures of 0. rubrinervis, strain B Seeds of third generation Number of specimens Number of deserens Percentage of deserens Mean Plant A 84 98 32 25 22 38] 28 [ 2SJ B 30 C 8 BOTANICAL GAZETTE [January The result confirms the expectation and shows that the figures given in the former table, although they give proof of the occurrence of mass mutation among the offspring of every plant of rubrinervis, are too low for the appreciation of the exact percentage of deserens. This must be estimated at about 30 per cent, or almost one-third of the whole progeny. In this race I self-fertilized three mutants in 191 3 and two specimens of deserens among their offspring in 191 5. I cultivated 150+84+180 specimens of the first group and left about one-half of them to flower, and 70+89 plants of the second group, all of which flowered in 1916. I had 573 plants in all, among which 319 bore flowers and fruits. They were all uniformly deserens, showing the marks of the type as previously described. No specimens of rubrinervis and no new mutants were observed among them. For one mutant of 19 13 and for two plants of the second generation in 191 5 I determined the amount of germs in the seeds and found 97-96 and 98 per cent, or an almost total absence of empty grains. Besides the two described families of O. rubrinervis I have con- trolled the seeds of some mutants in order to know whether all of them contained specimens of O. deserens and in percentages point- ing to mass mutation. I found the following figures: Percentage of deserens among the offspring of MUTANTS Mutant rubrinervis from Number of offspring Percentage of deserens 0. Lamarckiana, 1910 0. Lamarckiana, 1910 0. pallescens, 191 1 25 60 25 12 16 The strain of 0. Lamarckiana was derived from a rosette found in the original station near Hilversum in 1905, and the pallescens had been a mutant from this same strain. ^ Although the cultures were small, they prove the existence of mass mutation. I sowed the seeds of a specimen of deserens from the first culture in 1914, cultivated 25 flowering plants, and found these uniform with the type of their parent. ' New dimorphic mutants of the Oenotheras. Bot. Gaz. 62:262. 1916. 1919] DeVRIES— OENOTHERA RUBRINERVIS 9 Summing up the results of all the tables, we may conclude that all specimens of 0. rubrinervis, derived from various sources, and the mutants as well as their offspring, show mass mutation into O. deserens, besides a considerable number of empty seeds. Taking into consideration the unavoidable losses in the numerical estima- tions, we may further conclude that O. rubrinervis produces about one-fourth empty seeds, and among the living offspring about one- third 0. deserens, which are constant in their progeny and have no empty grains, or almost none. This points to a relation of i : 2 : i for the whole harvest. The phenomenon is thus shown to be parallel to the sphtting of the hybrid mutant of O. gigas nanella and to the mass mutability of O. grandiflora and of 0. Lamarckiana itself. Crosses between O. rubrinervis and O. deserens. — If this explanation is true, it may be confirmed by means of crossing O. rubrinerois with its mass mutant. The sexual cells of the first are about one-half deserens without a lethal factor, and the rest velutina provided with such a factor; those of 0. deserens, however, are uniformly so. We must expect, therefore, a splitting into almost equal parts of deserens Xdeserens = 0. deserens, and of velutina X deserens = 0. rubrinervis. I made both the reciprocal crosses in 191 5, cultivated 58 and 50 specimens of their offspring in 1916, and counted them at the beginning of the flowering period in July, finding as follows: O. rubrinervis XO. deserens. . O. deserens XO. rubriner\'is. . Percentage of j Percentage of rubrinervis deserens 48 78 52 22 The two types of hybrids resembled their parents exactly, and the figures point to numerical equality of the two groups, although the cultures were only small. Thus we see that the expectation from our formula is confirmed by the experiment. Twin hybrids of O. rubrinervis. — The twin hybrids of O. Lamarckiana and 0. grandiflora are now explained as the result of the mass mutation of these species, but the experimental proof is not complete as yet because neither of these species is known to occur without that form of mutation. In this respect the case of lO BOTANICAL GAZETTE [JANUARY O. ruhrinervis is far stronger, since its two constituents are both represented in my cultures. This fact makes a complete analysis possible, as I have already pointed out. If O. ruhrinervis is split by some cross into laeta and velutina on account of its composition of gametes of deserens and velutina, then the corresponding cross with O. deserens must evidently give the same laeta and that with O. Lamarckiana mut. velutina the same velutina. Thus the split progeny can be duplicated by the addition of its components. I have described the splitting crosses in Gruppenweise Artbild- ung (pp. 122, 196-200, 1913) and repeated some of them so as to have the dimorphic progeny together with the cultures of the pre- sumed constituents, in order to be able to identify their characters during the whole time of their development. The percentage figures given in my book are as follows: Twin hybrids of 0 . ruhrinervis Cross Percentage of laeta Percentage of velutina 0. biennis X ruhrinervis 0. ruhrinervis XO. biennis Chicago. . 0. ruhrinervis XO. Cockerelli Mean 30-49 39-44 49 42 51-70 56-61 51 58 In the second and third generation of the two latter crosses the laeta have split off brittle ruhrinervis in about one-third of the cul- tures, whereas the velutina remained constant. I repeated the two first named crosses in 1915, but not the third one. On the other hand, I have repeated the cross with O. Hookeri, in the progeny of which I had previously not been able to distin- guish the twin types. I had the following cultures in 191 6. Most of these plants flowered in August. Twin hybrids of 0. ruhrinervis; cultures of 191 6 Cross Number of specimens Percentage of laeta Percentage of velutina 0. biennis X ruhrinervis 0. ruhrinervis XO. biennis Chicago 0. Hookeri X ruhrinervis 59 60 60 46 53 20 40 54 47 80 Mean 60 igig] DeVRIES— OENOTHERA RUBRINERVIS II The results coincide with the previous ones as nearly as might be expected. The types of the twins were the same as those in the older cultures. For these twins I have determined the amount of empty grains, self-fertilizing two specimens of each of them in August 191 6 and counting out 100 seeds for each parent. Percentage of germs in seeds of laela and vehitina Cross Parent number laela velutina 0. biennis Xrubrinervis I 2 I 2 I 2 93 96 94 95 94 95 3 4 28 0. rubrinervis X Chicago 0. Hookeri Xrubrinervis 31 59 71 As in other cases, the seeds of the laeta hardly contain any empty grains, whereas those of the velutina are often badly developed. My task was now to repeat these crosses, substituting O. mut. deserens and 0. mut. velutina for O. rubrinervis. The latter group of crosses have already been described elsewhere;^ they yielded pure cultures of velutina which in every case were exactly like the velutina of the corresponding cross with 0. Lamarckiana. The crosses with 0. deserens were made in 191 5 and their progeny studied in 191 6; it was in every case wholly uniform. Crosses of 0. mut. deserens Cross Number of specimens Type 0. biennis X deserens 0. svrticolaX deserens 60 70 60 70 0. (bien.XLam.) laeta 0. (syrtic.XLam.) laeta 0. Hookeri X deserens 0. (Hook. X Lam.) laeta 0. deserens X biennis Chicago 0. (Lam. X Chic.) laeta About one-half of each culture flowered and developed their fruits. For each culture a control parcel was cultivated with the laeta from the corresponding cross with O. Lamarckiana, and they were compared during all the time of their development. I could ^ Kreuzungen von Oenothera Lamarckiana mut. velutina. Zeitschr. f. Ind. Abst. 17:1917. 12 BOTANICAL GAZETTE [january not find any differences. The descriptions for the deserens laeta are exactly the same as those given previously for the Lamarckiana laeta. Although I have not made the cross O. syrticolaXO. rubri- nervis, I have added the second experiment of the table. From this and the result of O. syrticolaXO. blandina (mut. velutina) described in my former article the result of the cross O. syrticola X rubrinervis may be predicted, and so it would be in other cases also. Summing up the results of these experiments, we see that in pro- ducing twin hybrids O. rubrinervis is split in exactly the same way as an artificial mixture of about equal parts of gametes of O. deserens and O. mut. velutina would be. The conclusion that its gametes really possess this dimorphy is thereby as clearly proven as might be expected. Crosses of O. rubrinervis with O. Lamarckiana and its DERIVATIVES. — In Gruppenweise Arthildung I have described the first generation of these crosses as consisting of two types, O. La- marckiana and 0. hybr. subrobusta. The latter is a rubrinervis in which the brittleness fails, and thereby very similar to our new mut. erythrina;'^ but this similarity is only an external one, since after self-fertilization the hybrid subrobusta splits off, as a rule, brittle rubrinervis plants, whereas the erythrina produces the decipiens, which is not brittle. Shortly after publishing my book, however, I discovered in the summer of 1 913, among the progeny of a cross of O. rubrinervis and O. Lamarckiana, a slight difference among the Lamar ckiana-like plants. Some of them were stouter and had broader and less crinkled leaves than the others. I self-fertilized them and got a culture, which, although not uniform, repeated the deviating marks of the parental type in the majority of the indi- viduals. I shall call this hybrid t}^e lucida. Moreover, in making a large number of crosses of individuals of the same family of rub- rinervis with Lamarckiana plants from various sources, as well as different mutant strains, I discovered that the second hybrid type is not always the solid subrobusta, but sometimes the brittle rubri- nervis. I have not as yet discovered why this should be so. We should expect the brittleness to be recessive to the production of 'Zeitschr. f. Ind. Abst. 16:262. 1916. iQig] DeVRIES— OENOTHERA RUBRINERVIS 13 Strong fibers, and as a rule it is so, but not always. The two con- trasting cases have occurred mainly in strains derived from different initial plants, and some hidden mutation might be responsible for the dominance of the brittleness. This seems to be the case at least in O. nanella, but the number of crosses in each of my different families of dwarfs is too small to decide whether this is the real cause. Crosses of O. rubrinervis with other mutants than the dwarfs have also given sometimes the brittle form and sometimes the subrobusta for the second hybrid. If we keep in mind that the hybrid rubrinervis is only a brittle form of the hybrid subrobusta, and that the one may be substituted for the other for unknown reasons, the following descriptions will easily be understood. I might add, however, that from a single cross between two individual parents both types never arise simultane- ously in the first generation. In the succeeding generations the rubrinervis as a rule are constant, whereas the subrobusta may split off the brittle form. If we assume the gametes of O. Lamarckiana to consist of equal parts of typical ones and of velutina, and those of O. rubrinervis to consist of deserens and velutina, O. LamarckianaXO. rubrinervis must yield 25 per cent typicaXdeserens, 25 per cent typica Xvelutina, 25 per cent velutinaXdeserens, and 25 per cent velutina Xvelutina. The last combination will produce empty grains, since the same lethal factor comes in from both sides; on the other hand, the three first named combinations must give viable seeds. Typica Xvelutina is the formula for O. Lamarckiana, and velutinaXdeserens that for O. rubrinervis and subrobusta, and so the occurrence of these hybrid types is easily explained. The remaining combination typicaX deserens must then be assumed to give the new hybrid lucida, and this can be verified by crossing O. deserens with Q. Lamarckiana. All these deductions are, of course, the same for the reciprocal crosses. If these deductions are reliable, they show that the poly- morphy of the first generation of hybrids between the two older forms is due to the combination of their capacities to produce twins in other crosses. In other words, it is a natural sequence of their secondary mutability. I shall now describe the experiments which seem to me to justify these deductions. 14 BOTANICAL GAZETTE [JANUARY 0. LamarckianaXO. rubrinervis. — According to the deduc- tions just given the expectation for this cross is O. Lamarckianay.0. rubrinervis = (typica+velutina) X (deserens+velutina) = typ.Xdes.+vel.Xdes.+tjrp.Xvel.+vel.Xvel. = lucida subrobusta Lamarckiana empty grains or rubrinervis In the first place, I determined the amount of empty grains, using the same method as in previous cases. Cross Parent number Cross Percentage of germs 0. Lamarckiana X rubrinervis I 2 3 I 2 1913 1913 21 0. rubrinervis X Lamarckiana 29 33 48 60 The presence of empty grains is thereby proven, although the percentages of germs are much smaller than would be expected; but this may be due to quite different causes, as has been shown else- where. In the second place, I studied the living progeny for the fourth cross of this table and for a cross made in 1907 with another strain of O. Lamarckiana. Each of these cultures was trimorphous, con- taining the types Lamarckiana and lucida, and besides these either subrobusta or rubrinervis (that is, tough or brittle). Number of Specimens Percentage of Cross Lamarckiana lucida subrobusta or rubrinervis 0. rubrinervis X Lamarckiana 1913 1907 Mean 60 69 32 40 36 20 6 13 48 rubrinervis 54 subrobusta 151 The expectation would be for equal parts, but for some unknown reason the lucida almost always fall short of this. Apart from this difficulty the results of these cultures coincide with the theoretical igig] de vries—oeno thera r ubriner vis IS deductions from our formula. I have made quite a number of further crosses between these two forms, partly in 1905 and partly in 1913, using always the same family of rubrinervis, and taking the combinations in both reciprocal directions. Six of them have given for the third hybrid t>'pe rubrinervis and three of them subrobusta; but since I have not determined the amount of lucida among them, it is of no use to give the percentage figures. The exactness of the identification of the types in the formula can be controlled by direct crosses with the constituents mut. deserens and mut. velutina. The latter has been described under the synonym O. blandina. I made the following combinations: Year Number of plants Percentage of Cross lucida subrobusta velutina 0. deserens XLamarckiana 0. rubrinervis X blandina 0. blandina X rubrinervis 0 deserens X blandina 1915 1915 1913 191S 1915 49 70 70 70 49 18 0 0 0 0 82 50 53 100 100 0 50 47 0 0. blandina X deserens 0 The expectation for these crosses was: O. deserensXLamarckiana = 0. deserensX(typ.+velutina) = lucida+subrobusta O. blandina X rubrinervis = O. blandinaX (deserens4- velutina) = subrobusta+ velutina O. blandinaX deserens = 0. blandina X (deserens) = subrobusta Apart from the figure for lucida, which is too small, the results of the experiments directly confirm the expectation. I have deter- mined the amount of empty seeds for the four last named crosses, and found almost none: Percentage of Cross germs in seeds O. rubrinervis X blandina 97 O. blandinaX rubrinervis 91 O. deserens X blandina 100 O. blandinaX deserens 90 Moreover, I made the same determinations for the hybrids from the two first named of these crosses, self -fertilizing them in 1916. For the two latter crosses it was evident that the hybrids would i6 BOTANICAL GAZETTE [JANUARY hardly have any empty grams, and I did not think it necessary to control this. Cross Parent NUMBER Percentage of germs in seeds OF sttbrobusla velutina 0 rubrinervisXblandina I 2 3 4 I 2 3 96 97 99 97 96 96 96 70 0. blandina X rubrinervis 75 76 68 75 80 The laeta have hardly any empty grains, but the figures for velutina fall short of this, even as in other instances. In the last place, I counted the germs in the hybrids of the crosses with La- marckiana, self -fertilizing their specimens of each of the types: Parent NUMBER Percentage of germs in seeds of Cross lucida Lamarckiana rubrinervis 0. rubrinervis XLamarckiana 0. deserens XLamarckiana I 2 3 I 2 87 91 94 85 86 25 34 93 53 59 65 The lucida have almost no empty grains; the figures for hybr. rubrinervis are the same as those for the mutant of that name, but those for the Lamarckiana type give an unexpected result. In two cases they are the same as for the species, but in the third the empty grains have almost wholly disappeared. This latter specimen has lost all the external marks of O. deserens and O. rubrinervis, but kept the absence of the lethal factors. Its progeny splits into Lamarckiana, lucida, and rubrinervis , and the first of these forms repeats the splitting in the following generation. 0. Lamarckiana nanellaX rubrinervis. — As in so many other cases, the crosses with dwarfs can give a verification of those with the species itself. In Gruppenweise Artbildung (p. 215) I have 1919] DeVRIES—OENOTHERA RUBRINERVIS 17 described the pedigrees of two reciprocal crosses, both of which produced as a second hybrid the suhrobusta. This was seen to split off, after self-fertilization, brittle plants and dwarfs. In 191 5 I sowed some seeds of the suhrobusta plants of 1907 mentioned in those tables, in order to compare their progeny with my newer cultures. I found for two specimens of O. {nanellaXruhrinerois) suhrobusta 38 and 45 per cent of dwarfs among 82 and 60 plants, and for two parents O. {rubrinervisXnanella) suhrobusta 20 and 13 per cent of dwarfs among 60 and 46 individuals. The number of brittle plants, however, was very small, being two specimens for the first and one for the reciprocal group. It is possible that the germs of this t>pe are weaker, and that some of them had died during the 7 years of their preservation. I self-fertilized one brittle specimen in each of the two main groups and had in 191 6 two lots of 45 and 60 flowering plants, all of which were brittle and like their parents. They contained 9 and 8 per cent of dwarfs, the stems of which were likewise brittle at the time of flowering.'" Other races of O. nanella or other conditions may produce in the corresponding crosses brittle hybrids instead of suhrobusta. I made the cross O. rubrinervisXnanella in 1905 with a dwarf mutant race of 1895 and the reciprocal one with the progeny of a dwarf which had arisen in 191 1 from Lamarckiana, using in both cases the same family of rubrinervis as in all previous crosses. The first named cross gave 35 per cent Lamarckiana, 3 per cent lucida, and 62 per cent brittle rubrinervis among 68 specimens in 1913. The second cross yielded the same three t^'pes, but the percentage figures devi- ated widely. I had only 6 per cent Lamarckiana and 2 per cent lucida, but 92 per cent brittle rubrinervis among 140 plants, most of which flowered in August. The main result, however, is clear, namely, that the crosses between O. rubrinervis and O. nanella give three types of viable hybrids, one of which carries the visible marks of O. rubrinervis, but may be either brittle or tough. I have made only one cross between O. deserens and a dwarf, taking this latter from the first of the two last mentioned families. " By means of this the gap left in the second pedigree of p. 215 of my book is filled up, and both pedigrees are completed by the production of dwarfs from the rubri- nervis specimens. i8 BOTANICAL GAZETTE [JANUARY I crossed them in 1915 and had in 1916 a culture of 60 plants, among which 3 per cent were lucida and 97 per cent brittle rubrinervis. Other types failed, as was to be expected. The seeds of the two lucida plants contained 89 and 95 per cent of good germs. Summing up the results of the crosses between O. rubrinervis and O. nanella, we see that they yield exactly the same hybrid types as those with O. Lamarckiana and in corresponding percentages. Crosses of O. rubrinervis with heterogamic mutants.— Crosses with the pollen of these forms must simply confirm those with O. Lamarckiana, since their pollen carries mainly the same hereditary qualities as that of the parent species. I fertilized in 1913 two plants of my main race of O. rubrinervis with O. cana, two with the pollen of the Lamarckiana-like offspring of self-fertilized scintillans, and added the reciprocal cross of the latter combination. In the following table I shall call these offspring scintillans- Lamarckiana. Crosses of 0. rubrinervis with heterogamic mutants Cross O. rubrinervis X cana O. rubrinervis X cana ._ . O. rubrinervis X scintillans-Lamarcki- ana • • O. rubrinervis Xscintillans-Lamarcki- ana O. scintillans-Lamarckiana X rubriner- vis O. scintillans-Lamarckiana X rubriner- vis Number of specimens 60 57 60 60 84 34 Percentage of Lamarckiana 32 53 13 7 7 IS lucida 38 32 3 10 4 9 subrobusta 30 15 84 83 89 76 About one-half of each group flowered in August. No brittle specimens occurred. The types were exactly the same as those derived from the cross between O. rubrinervis and O. Lamarckiana. If the heterogamic types are used as female parents, the splitting of course will be more complicated. I fertilized a strong biennial specimen of O. scintillans with the pollen of a plant of O. rubrinervis and had in 191 6 a culture of only 23 plants, all of which flowered in August. There were 5 types: 11 Lamarckiana, 2 lucida, i sub- zgig] DeVRIES—OENOTHERA rubrinervis 19 rohusta, 9 scintillans, and 8 oblonga. The first three were the same as in previous crosses and confirm their result; the last two named t>'pes are the same as are always seen in the first generation of crosses of O. scintillans when this is used as the seed parent. Moreover, in 191 5 I fertilized 4 plants of my race of O. lata with 0. rubrinervis, counted the lata and alhida in their progeny in May 191 6, and for want of space planted out only a part of the others, in order to distinguish the types, but without trying to determine percentage figures. Altogether I had 434 seedhngs, among which 7 per cent were lata and 6 per cent albida. At the time of flowering I counted 23 Lamarckiana, i lucida, 20 brittle rubrinervis , besides II mutants (5 oblonga, 5 obovata, and i scintillans). No subrobusta occurred in these cultures. These results confirm those previously given. Second and later generations. — Brittleness and dwarfish stature are recessive characters, and as such may be expected to be split off in the succeeding generations. For the crosses between O. rubrinervis and O. nanella this splitting has already been dealt with. For the other crosses our analytical formula for O. Lamarck- ianaXrubrinervis shows that the types lucida and subrobusta may be expected to produce a splitting, whereas the Lamarckiana-like hybrids cannot contain the necessary factors. The production of brittle plants from subrobusta had been observed in the case of the dwarfs, and so I studied in 1916 the progeny of three specimens of lucida from previous crosses. Splitting progeny of O. hybr. lucida; cultures of 19 16 Lucida from Number of specimens Percentage of Tall plants deserens 0. rubrinervis X scintillans 67 106 52 80 53 62 48 20 0. rubrinervis X scintillans 0. rubrinervis X Lamarckiana Mean 47 38 Moreover, in 19 16 I self -fertilized some specimens of lucida taken in the first generations of the crosses mentioned, sowed their seeds in 191 7, cultivated all the seedlings until the time of ripening their first fruits, and counted them repeatedly during the summer. The 20 BOTANICAL GAZETTE [JANUARY- difference between the deserens and the lucida was very striking,. the first reaching only half the height of the latter. I have broken the stems of all the plants in August, at the time of the last counting, and found all the deserens brittle and all the tall ones tough. The first were evidently deserens and not rubrinervis, as seen by the characters described for these two types. Among the tall ones, however, I have not succeeded in finding any difference, the type of hicida prevailing to the apparent exclusion of that of O. Lamarcki- ana. For each of the crosses mentioned in the following table I had 58-60 flowering specimens in August. Splitting progeny of hybrid lucida; culture of 191 7 Parent number Percentage of Lucida from Tall plants deserens 0. rubrinervis XO. Lamarckiana 0. deserens XO. Lamarckiana I 2 3 I 2 I 2 50 60 48 55 39 51 41 49 50 40 52 45 61 49 59 51 0. deserensXO. nanella Mean Oenothera Lamarckiana mut. oblonga and mut. nanella Our conception of Oenothera rubrinervis as a half mutant may be applied to O. oblonga, and explain its behavior in crosses in an analogous way. The main difference, as I have pointed out in Gruppenweise Artbildung, is that some types of hybrids, as we might expect, are constantly absent or suppressed, as I called it. If we assume this suppression to take place in the pollen before fecunda- tion, the remaining phenomena are easily explained on this basis. It will be sufficient to review the facts given in my book, and to combine them with the results of some determinations of the amount of barren grains in the seeds of self-fertilized and crossed individuals. The amount of empty seeds is about the same in O. oblonga as in 0. Lamarckiana. For the cultures of 191 1, mentioned in my book, I found among the seeds of two self-fertilized individuals 25 and 33 per cent of germs. Seeds of biennial plants collected in 1 9 13 contained 30-18 and 17 per cent of germs; but seeds of annual igig] DeVRIES— OENOTHERA RUBRINERVIS 21 plants, saved in 1914 on two new mutants from O. Lamarckiana and on one from O. cana, gave only 6-5 and 6 per cent of germs. Annual specimens are always much weaker than biennial ones, and their fruits are often thin instead of club-shaped. These figures evidently point to a complete analogy with O. Lamarckiana. The question whether the lethal factors are the same as in O. Lamarckiana may be answered by crosses with this species. I tried the seeds of a cross O. oblonga X Lamarckiana , of one of O. oblongaXnanella, both made in 191 1, and of a cross of 1913 of O. oblonga XO. cana. I found 53-40 and 34 per cent of good germs. The figures do not essentially differ from those found for self- fertilized Lamarckiana, and thereby show that the lethal factors must be the same and simply inherited by O. oblonga from its parent species without change. The ovules which produce empty grains after self-fertilization may develop into normal seeds after crosses with other species, even as in the case of O. Lamarckiana itself. O. oUongaXbiennis gave 92 per cent of germs, O. oblongaXatrovirens (cruciaia) 87 per cent, O. oblongaXHookeri 90 per cent, and O. syrlicola {muricata)X oblonga 90 per cent. Thus we see that in this respect also the lethal factors are the same as in O. Lamarckiana. Our assumption is that O. oblonga arises by means of a mutation in the Lamarckiana gametes of our species, leaving the velutina gametes unchanged. The formula for self-fertilization, assuming the oblonga gametes to be suppressed in the pollen before fecunda- tion, is as follows: (obl.-fvelu.) Xvelu. =obl. Xvelu.+velu.Xvelu. This explains the constancy of the mutant, since the velutinaX velutina germs contain the same lethal factor on both sides and thus produce the empty grains. If we compare this formula with the results of the crosses described in my book (pp. 266-267), we find a complete harmony, as I shall now try to show. Fertilized by O. Lamarckiana and analogous mutants, O. oblonga must give (obi. + velu. ) X (Lam. -fvelu.) = obi. XLam.-f obi. Xvelu. -j-velu. X Lam. -f-velu. Xvelu. = 25 per cent empty grains-f25 per cent oblonga + 2$ per cent Lamarckiana-\-2$ per cent empty grains. The expectation is therefore for two t>pes and these in equal pro- portions. The two t>pes always appeared, and no others besides 22 BOTANICAL GAZETTE [JANUARY them, but the percentage figures are very variable. I found them as follows: Cross Parent number Percentage of oblonga Lamarckiana O. oblongaXLamarckiana I 2 3 I 2 I 4 4 14 IS 46 81 93 96 85 80 O. oblongaXLamarckiana O. oblongaXLamarckiana O. oblongaXnanella O. oblongaXnanella 14 17 O. oblongaXscintillans The reciprocal crosses cannot produce any oblonga, since this is assumed to be suppressed in the pollen. The only exception is O. scintillans, which gives rise to a high amount of oblonga after self-fertilization, and therefore may produce the same mutant after a cross. I found as follows: Cross Percentage of oblonga Lamarckiana 0. Lamarckiana X oblonga. . . 0. nanellaX oblonga 0. lataXoblonga 0. scintillans X oblonga 0 0 0 18 100 100 100 82 The pollen of O. oblonga must produce, after crosses with differ- ent species, only velutina, as is easily seen from our formula. No oblonga and no laeta are to be expected. O. biennis, O. syrticola, O. Cockerelli, and O. Hookeri fecundated with O. oblonga uniformly gave this result. The reciprocal crosses, however, must give a splitting, the laeta hybrids assuming the characters of O. oblonga. The percentages should be about 50, but in my experiments there was much fluctuation in this respect. I found as follows: Cross Parent number Percentage of oblonga velutina 0. oblonga X biennis Chicago I 2 3 I I 2 8 16 38 41 II 24 02 0. oblonga X biennis Chicago 84 0. oblonga X biennis Chicago 0. oblonga X Cockerelli 62 0. oblonga X Hookeri 0. oblonga X Hookeri 89 7=^ 1919] DeVRIES— OENOTHERA RUBRINERVIS 25 Crosses with O. ruhrinervis also yield the expected result. This would be in one direction (obl.+velu.) X(deserens+velu.) =obl. X des.+obl. Xvelu.+velu. Xdeserens+velu. Xvelu. = 25 per cent {obi. Xdes.)-]r2$ per cent oblonga-\-2^ per cent ruhrinervis -\- 2^ per cent empty grains. I have not as yet tried the cross between O. ohlonga and O. deserens, however, and thus must leave undecided the ques- tion as to which characters will dominate in this hybrid. As a matter of fact, I found 20 per cent ohlonga and 80 per cent ruhriner- vis and no other types. The reciprocal cross must give (des.-f- velu.)Xvelutina = des. Xvelu.+velu. Xvelu. = 50 per cent ruhriner- vis-\-$o per cent empty grains. Only ruhrinervis have been observed in this culture. Crosses with the pollen of O. biennis must give ohlonga Xhiejtnis -\-velutinaX biennis. The former is intermediate between the par- ents, whereas the second is the same as the hybrid type Lamarckiana Xbiennis. I found in one cross 65 per cent ohlonga (partly dwarfish)' and 35 per cent hybrids of the second type. In another instance^ however, the ohlonga failed from some unknown reason. With those species which ordinarily produce the twins densa and laxa the pollen of O. ohlonga must evidently give only the latter type. This has been the case in three trials with O. biennis Chicago> and in one with O. atrovirens {cruciata). For further details and for the constancy or splitting in the second generation I must refer the reader to the pages of my book already quoted. These results, however, show clearly that all the facts hitherto ascertained confirm the formula assumed for the self- fertihzation, and thereby the analogy with the phenomena observed in O. ruhrinervis. Summing up this discussion we may say, therefore, that O. ohlonga arises through a mutation of the typical sexual cells of O. Lamarckiana, leaving the velutina gametes and also the lethal factors unchanged, but producing, besides the externally visible marks of the mutant, a suppression of the mutated pollen grains. On the other hand, O. mut. nanella seems to arise through mutations in the velutina gametes of O. Lamarckiana, as is shown by the fact that the laeta do not split off dwarfs, whereas the velutina regularly do so. The figures given in my book for the crosses with O. nanella may be calculated in the same way, and 24 BOTANICAL GAZETTE [januasy will be found to comply with the views proposed in this article. It would lead us too far, however, to reproduce these calculations here. In all these cases the conception that mass mutation is the chief cause of the production of twin hybrids evidently makes the sup- position of a labile condition of the factor for laeta superfluous. It seems desirable, therefore, to lay stress on the fact that this sup- position does not rest on the phenomena observed in the produc- tion of these twins. It is mainly derived from other observations, and some of them may be briefly repeated here in order to make this point clear. They refer to the brittleness of O. rubrinervis and O. deserens and to the dwarfish stature of O. nanella. In crosses brittleness behaves in three different ways. With O. biennis Chicago and O. Cockerelli it is recessive to the tough structure of the fibers, since it fails in the first generation and reappears in the second in ratios corresponding to Mendel's law. In crosses with O. Lamarckiana it is sometimes dominant and sometimes recessive, as has been shown. In O. rubrinervis and O. deserens the toughness is wholly absent. From these and other facts it is clear that at least three conditions of this factor are possible. I call them active, labile, and inactive. Whether the labile condition is due to linkage or to some other cause is as yet an open question, which, however, has no influence upon the main contention. The combination '' active X inactive" is assumed to be responsible for Mendelian crosses, but the com- bination " labile X inactive" may cause a splitting in the first gen- eration and produces, as a rule, constant hybrids. The two types of first generation hybrids appear in variable numerical propor- tions according to different circumstances. If one of the groups is so small as not to be represented in every loo specimens, the splitting may seem to fail, and such extremes are of common occur- rence. This would explain the dominance of an evidently recessive character. The case is exactly the same for the dwarfish stature. The factor for tallness must be in the inactive condition in the dwarfs, but in the active condition in O. rubrinervis, since the crosses between these two types follow Mendel's law. In O. Lamarckiana, 1919] DeVRIES— OENOTHERA RUBRINERVIS 25 however, it is labile, since tall and low specimens appear in the first generation of its cross with O. nanella. In many ternary crosses of hybrids of this mutant the dwarfish stature dominates over the tall condition, but the dominance is not always absolute and sometimes 3-5 per cent of tall specimens appear among the dwarfs, as I have shown in Gruppenweise Arthildung. This fact evidently supports our conception. The conclusion from this discussion is that since brittleness and dwarfish stature are in some cases recessive to and in other cases dominant over their antagonists, these latter must be sometimes in the active and in other instances in the labile condition. Summary 1. Oenothera rubrinervis is a half mutant, produced by the copu- lation of a mutated gamete with a normal velutina gamete of O. Lamarckiana. 2. In consequence, it produces about one-fourth empty grains, a mass mutation of about one-fourth pure or double mutants, and one-half specimens of O. rubrinervis, which will repeat the splitting. 3. The pure or double mutant is called O. mut. deserens. It is very similar to O. rubrinervis, but the leaves of its young rosettes and the bracts of its flower spike are broader and more even. 4. O. mut. deserens is constant from seed. It has no hereditary empty grains. 5. The formula for the self-fertilization of O. rubrinervis is therefore O. {deserens -\-velutina) = des.Xdes. -jrvelu. Xvelu. -\-des. X velu. The first combination gives the mass mutation, the second the empty grains, the third the normal plants of O. rubrinervis. 6. In crossing with other species the two kinds of gametes will produce twin hybrids, as, for example, laeta and velutina. This assertion has been controlled by making the corresponding crosses of 0. mut. deserens and O. mut. velutina. The first produce the laeta and the second the hybrid velutina. The result of a cross of 0. rubrinervis is equal to the sum of these two crosses. 7. Outside of the mass mutability into O. deserens, O. rubrinervis is not known to mutate to any noticeable degree. This shows that 26 BOTANICAL GAZETTE [January the internal constitution, which causes the mass mutation, is not in itself a cause for further mutability. 8. The constitution of the gametes of 0. rubrinervis can directly be proven by a cross with O. deserens, since O. rubrinervis = {deserens -\-velutina)xO. deserens produces 0. deserens and (O. deserens X velutina) or rubrinervis. g. Crosses of O. rubrinervis with 0. Lamarckiana give three types of hybrids, besides about one-fourth empty seeds. One type exactly resembles 0. Lamarckiana and is constant in its pro- geny. A second type called lucida has broader and more shiny leaves, and after self-fertilization splits off brittle specimens. The third type is either subrobusta or rubrinervis, and in the first case may produce the brittle form in the second generation. All these phenomena are easily explained by the proposed formula for the constitution of O. rubrinervis as a half mutant. They were con- firmed by means of crosses with O. nanella and some other mutants. 10. O. oblonga is quite analogous to O. rubrinervis, since it must arise through a mutation of the typical sexual cells of 0. Lamarcki- ana, leaving the velutina gametes unchanged. Contrary to O. ru- brinervis, however, the two lethal factors remain in their condition, and moreover the mutated gametes must be assumed to become suppressed in the pollen of the mutant. 11. 0. nanella seems to arise through mutations in the velutina gametes of O. Lamarckiana, since after crosses with other species or mutants it is not split off by the laeta hybrids, but only by those of the type velutina. LuNTEREN, Holland NOTES ON AMERICAN WILLOWS. Ill A CONSPECTUS OF AMERICAN SPECIES AND VARIETIES OF SECTIONS RETICULATAE, HERBACEAE, OV.\LIFOLIAE, AND GLAUCAE Camillo Schneider In this third article, as I said in my first paper,' a key will be given containing the species treated in the first two papers, and also those of the sections Reticulatae and Herbaceae (Retusae), together with a few other species the systematic position of which is not yet fully understood, but which are best placed near one of these groups. I have tried to prepare two separate keys for the determination of the male and female plants as I did in "Conspectus analyticus Salicum Asiae orientalis Himalayaeque " in Sargent, PI. Wils. 3:73. 1916. Typical complete specimens are not always at hand, and without them even such a good key as that given by Coville (in Proc. Wash. Acad. Sci. 3:300. 1901) to the Alaskan willows is insufficient to determine species. To make a key to the sections only proves likewise of little value owing to the great difficulty of exact limitation of the groups, as I shall explain later, Clavis specierum I. SECUNDUM SPECIMINA FEMINEA A. Ovaria (pedicelHs inclusis) etiam juveniha glaberrima. I. FoHa utrinque concoloria, viridia et stomatifera, minima vel parva; amenta serotina, vulgo paucifiora; bracteae pl.m. concolores, flavescentes vel violascentes, vix vel sparse brevi- pilosae; fruticuli minimi prostrati (sed vide 7. S. Peasei). Folia semper crenato-dentata, utrinque tenuiter reticulata, anni praeteriti nunquam persistentia. Ramuli floriferi tenuissimi, breves, fere semper bifoliati; bracteae flavescentes, sparse pilosae 8. 5. herbacea Ramuli floriferi crassiores, longiores, 2-4-foliati; bractae fuscescentes, albido-pilosae 7. S. Peasei 'Box. G.\z. 66:118. 1918. 27] [Botanical Gazette, vol. 67 28 BOTANICAL GAZETTE [January Folia integerrima, adulta sicca per secundum annum vel diutius persistentia. Nervi laterales foliorum utrinque pl.m. elevati, venulae etiam prominulae (confer etiam lo. 5. phlehophyllam, cujus fructus interdum glaberrimi sunt) . .9. S. rotundifolia Nervi laterales foliorum superne tenuissime incisi, sub- tus prominentes, venulae baud visibiles. .11. S. Dodgeana II. Folia discoloria, subtus distincte pallidiora, pl.m. glaucescentia (saepissime pruinosa) ; fruticuli repentes vel parvi, erecti. Bracteae concolores, flavescentes, glabrae vel sparse tantum ciliatae. Fruticuli parvi, erecti, ramis non prostratis et radicantibus ; styli distincti, quam stigmata brevia bifida longiores, baud vel tantum apice fissi; folia circiter 3-5PI0 longiora quam lata. Folia superne cito margine excepto glabra, stomatifera 21. S. chlorolepis Folia superne infimis exceptis pl.m. villosula, esto- matifera 22b. S. hr achy car pa var. glahellicarpa Fruticulus depressus, ramis prostratis radicantibus; stylus brevi§simus, vulgo bifidus stigmatibus bifidis baud longior; folia satis crassa, vix i|plo longiora quam lata 3. S. leiolepis Bracteae pl.m. bicolores, ad apicem vel fere totae fuscae, pl.m. longe sericeo-pilosae; fruticuli parvi, prostrati, ramis radi- cantibus. Ovaria etiam juvenilia distincte pedicellata; pedicellus fructuum glandulam late ellipsoideo-rectangularem vel sub- quadratam circ. 2plo superans; stylus distinctus, apice bifidus stigmatibus brevibus oblongis bifidis subduplo longior; folia superne baud stomatifera, margine vulgo sparse et saepe indistincte denticulata 18. S. arctophila f. lejocarpa Ovaria subsessilia vel breviter pedicellata, pedicello etiam fructuum quam glandula pl.m. breviore vel vix sublongiore. Folia pl.m. glanduloso-crenato-denticulata (saltem ad medium et apicem), rarius subintegerrima, superne sto- 1919] SCHNEIDER—AMERICAN WILLOWS 2Q matifera, adulta marcescentia partim diu persistentia; stipulae saepe distinctae; stylus apice bifidus stigmati- bus brevissimis bifidis 2-2^plo longior. . . .6. S. Uva-ursi Folia integerrima, rarissime basim versus paucidentata; stipulae nunquam distinctae. Amenta cylindrica, 3-4plo longiora quam lata, 3-5 cm. longa (saltem basi) sublaxiflora ; folia majora ultra 3 cm. longa, superne non stomatifera 15. 5. arctica f. glabrata Amenta etiam fructifera vix ad ^plo longiora quam lata, densiflora; folia etiam maxima vix ultra 2 . 5 cm. longa vel superne stomatifera. Stylus satis brevis, stigmatibus mediocribus vix longior; fructus maturi vulgo pl.m. glaucescentes; folia superne (in var. camdcnsi excepta) baud stomatifera, matura subcrassa, subtus conspicue elevato-reticulata 16. S. ovalifolia Stylus elongatus tenuis, stigmatibus angustis longis saepe pl.m. longior; fructus maturi vix glauces- centes; folia superne stomatifera, matura tenuiora, subtus vix conspicue reticulata .... 17. S. stolonijera B. Ovaria (interdum tantum partim vel nonnisi pedicelli) pl.m. dense pilosa; fructus saepe glabriores vel partim glabri. I. Folia (sub anthesi perfecte evoluta) utrinque concoloria, viridia, aequaliter stomatifera, integerrima; fruticuli minimi repentes. Ramuli floriferi breves, tenuissimi, vulgo 2-foliati; folia utrin- que tenuiter reticulata, adulta sicca non persistentia; amenta paucifiora, vel multifiora; ovaria interdum tantum ad apicem pilosa, pl.m. sessilia; stylus distinctus, stigmatibus vulgo longior; glandula i 5. 5". polaris Ramuli floriferi 2-4-foliati; folia adulta marcescentia paren- ch>Tnate evanescente plures annos persistentia; amenta pluriflora, cylindrica; ovaria saepius distinctius pedicellata; glandula etiam dorsalis interdum adest (confer etiam 9. S. rotundijoliam f. pilosiusculam) 10. S. phlebop/iylla 30 BOTANICAL GAZETTE [january II. Folia subtus discoloria, pallidiora vel pl.m. glaucescentia, vulgo pruinosa, vel plantae aliis signis diversae. a. Folia circumcirca satis dense minute glanduloso-serrata, obovata, fere glabra; stipulae distinctae, lanceolatae, serratae; amenta pedunculo excluso 5-6 cm. longa, 10-15 nii^- crassa; stylus distinctus, stigmatibus circ. 2plo longior; planta prostrata 29. S. Chamissonis b. Folia integerrima vel pl.m. crenato-denticulata, vel plantae aliis signis diversae. I. Amenta serotina, pseudoterminalia, anguste cylindrica vel minima pauciflora, pedunculis nudis lis saepe aequi- longis suffulta; bracteae breviter (rarius longius) pilosae, pl.m. concolores, flavescentes vel violascentes; ovaria sessilia vel subsessilia; styli {S. venusta excepta) brevissimi vel nulli, stigmata brevia vel brevissima; glandulae 2 vel interdum plures pseudodiscum lobulatum formantes; folia satis crassa, superne pl.m. inciso- reticulata, rugosa, baud stomatifera {S. venustae?), subtus distincte elevato-reticulata, vulgo pl.m. longe petiolata. Folia coriacea, pleraque vix longiora quam lata vel ultra 4 cm. longa, superne conspicue inciso-reticulata, rugulosa; bracteae intus pl.m. brevipilosae vel utrinque sericeae. Frutex prostratus; folia cito glaberrima vel rarius piKs paucis sericeis obsita, vix ad 5.5:5 cm. magna; petioli elongati, ad 3 cm. longi; bracteae intus tanturn brevipilosae; fructus vix ultra 4.5 mm. longi I. S. reticulata Frutex prostratus vel erectus; folia etiam adulta subtus dense sericea vel majora, oblongiora et margine pl.m. distincte crenulata, vel petioli breves gemmis vix longiores: bracteae utrinque pl.m. sericeo-pilosae; fructus 5-7 mm. longi 2. S. vestita Folia tenuiora (chartacea), minima vel distincte longiora quam lata, vix ultra 3.5:2 cm. magna, superne vix vel indistincte inciso-reticulata; bracteae glabrae vel tantum margine basique pilosae, rarius etiam extus pl.m. 1919] SCHNEIDER— AMERICAN WILLOWS 31 pilosulae; frutices prostrati, interdum minimi suffru- ticulosi (confer etiam 31. S. venustam cujus specimina nondum vidi et quae stylo elongato filiformi fusco distincta dicitur; tantum a Sitka reportata) 4. S. nivalis et var. saximontana 2. Amenta coetanea, rarius pl.m. serotina, lateralia, ovata vel cylindrica, multiflora, rarius parva et pauciflora, sed pedunculi semper foliati; folia pl.m. tenuiter papy- racea, superne nunquam inciso-reticulata. a) Bracteae bicolores, apice vel pro parte maxima fuscae, versus apicem longe sericeae (id est pilis longis sericeis quam bractea vix vel paullo brevioribus instructae), interdum apice tantum ciliatae; fruticuli fruticesque parvi ramis ut videtur semper prostratis (confer etiam 28. S. lingulatam speciem valde incertam). Stylus sub nullus,quam stigmata divaricata bifida duplo brevior; amenta parva circ. 10 mm. longa, ovoidea vel ovato-globosa; bracteae atrae, extus sparse sericeae vel partim glabrae; ovaria breviter pedicellata, albo- sericeo-villoscula; folia ovata vel late lanceolata, obtusa, rigida vix ad 12 mm. longa. . . .30. S. glacialis Stylus semper distinctus, stigmatibus aequilongus vel vulgo longior. Folia sicca anni praeteriti pl.m. persistentia, lingu- lata, lineari- vel anguste oblanceolata (vel lanceolata apice subplicato-acuminata), vix ultra 18:5 mm. magna, superne stomatifera, subtus paullo palli- diora ; amenta parva, etiam f ructif era vix ad 2 . 5 : o . 9 cm. magna pedunculis 0.5-3 cm. longis exclusis; bracteae laxe sericeae, extus saepe glabrescentes; (ovaria et) fructus subsessiles, circ. 4-5 . 5 mm. longi, pl.m. villoso-tomentosi 12. S. cascadensis Folia adulta haud persistentia et plantae aliis signis diversae. Stigmata linearia, elongata, stylo tenui satis longo pl.m. 2-3plo breviora; ovaria fructusque saepe 32 BOTANICAL GAZETTE [january tantum ad apicem sparse pilosi, ceterum ut supra sub S. stolonijera indicata 17. iS*. stolonijera f. suhpilosa Stigmata brevia vel oblonga sed vix linearia vel plantae aliis signis diversae. Folia subtus (in sicco) paullo pallidiora (baud distincte glaucescentia vel albescentia), leviter elevato-nervata sed vix reticulata, laevia, petiolis vix 5-6 mm. longis instructa, vulgo lanceolata vel elliptico-lanceolata, utrinque acuta vel apice obtusa, superne stomatifera et interdum fere inciso-nervata, etiam maxima vix ad 4:1.8 cm. magna, integerrima (rarissime versus basim parce denticulata) ; stipulae nullae vel minimae, caducae; amenta (pedun- culo excluso) 2-4 . 5 (in var. caespitosa interdum ad 6) cm. longa et fructifera ad 1.3 cm. crassa; ovaria subsessilia, dense sericeo-villosa; stylus distinctus, saepe apice breviter bifidus, stig- matibus oblongis bifidis vulgo duplo rarius 3plo longior; glandula ventralis oblonga, ovoideo-conica; fructus pedicello brevi glan- dulam siccam subaequante excluso 4-5 mm. longi; fruticulus ramulis hornotinis flavescenti- bus tenuibus satis brevibus. . .13. S. petrophila Folia subtus distincte discoloria, glaucescentia, pruinosa, superne baud stomatifera vel majora, diversiformia vel plantae aliis signis diversae.^ Glandula ventralis satis brevis et lata, vix duplo altior quam lata, apice late truncata, pedicello fructuum duplo brevior; amenta submatura vel fructifera (3-) 5-10 : i. 2-1.6 cm. magna; ovaria tenuiter villoso- tomentella; styli distincti, apice bifidi, stig- = It is difficult to indicate in such a key the differences between S. petrophila and S. anglorum and its different forms with sufficient clearness. See my remarks under those species in my first paper {I.e.) . 1919] SCHNEIDER— AMERICAN WILLOWS 33 matibus oblongis bifidis paullo vel duplo (rarius fere triple) longiores; fructus pedi- cello excluso 6.5-8 mm. longi; folia superne estomatifera, margine vulgo partim sparse et saepe obsolete denticulata, forma variabilia, majora latiora ad 3-4 (-5) : 2. 5 (-3) cm. magna; frutex procumbens ramulis saepe satis elongatis, 2-3 mm. crassis (si glandula est brevis et lata sed pedicello brevior, confer i8a. S. hudsonensem) 18. S. arctophila Glandula oblonga, vulgo 2^-4plo longior quam lata et pedicelli etiam fructuum quam glandula pl.m. breviores vel rarius sublongi- ores vel plantae aliis signis diversae. Amenta fructifera ellipsoideo-globosa, circ. 1-1.5:1.5 cm. magna; folia subcoriacea, late ovalia vel obovato-rotunda, vix ultra 1.8:1.5 cm. magna, superne estomatifera, in sicco tenuiter reticulata, subtus valde elevato-reticulata, utrinque (saltem initio) ut ramuli novelli villosula 1 6b. S. ovalifolia var. puhescens Amenta fructifera cylindrica, vulgo longi- ora, tenuiora; folia tenuiora vel majora, subtus nunquam conspicue elevato- reticulata. Folia superne baud stomatifera,^ vulgo obovata vel obovato-oblonga, apice rotundata ad subacuta vel plicato- acuta, basi sensim vel subito attenuata, obtusa vel interdum rotundata, majora satis evoluta 3-6:2-4 cm. magna; petioli 9-20 mm. longi; stipulae in ramulis vegetis ovato-lanceolatae vel lanceolatae, 3 With the exception of 5. arctica var. suhcordata, which, however, is easily dis- tinguished from any of the forms of S. anglorum by its much larger leaves and longer stouter catkins. 34 BOTANICAL GAZETTE [january integrae vel subdenticulatae, 2-12 mm. longae; amenta sub anthesi 2.5-4 cm. longa, bracteis atris dense longe sericeis conspicua, fructifera 6 : i . 3 ad 9 : i . 8 cm. magna; ovaria sessilia vel subsessilia, sericeo-villosa; styli distinct!, vulgo in- tegri, quam stigmata oblonga bifida 2-2^plo longiores ; fructus pedicello brevi glandula | ad vix breviore (rarissimo sublongiore) excepto (6-) 8-10 mm. longi (confer etiam i6c. S. ovalifoliam var. suharcticam et i8a. S. hudsonensem) 15. 5. arctica et varietates Folia superne stomatifera, vulgo minora, valde variabilia (vide formas sub S. anglorum enumeratas); petioli vix ultra 10 mm. longi; stipulae nullae vel mino- res; amenta etiam fructifera vix ad 5.5: 1 . 5-1 . 8 cm. magna, saepe distincte minora, tenuiora; fructus ad 7-8 mm. longi pedicello subnullo vel glandula 2 ad Iplo breviore excluso (si folia sunt estomatifera sed plantae aliis signis baud diversae, confer etiam i8a. S. hudsonensem) 14. S. anglorum et varietates 0) Bracteae concolores, flavescentes, stramineae vel brunnescentes (rarius subbicolores, apice violaceae vel leviter fuscae), semper breviter sericeo-villosulae (id est pilis quam bractea brevioribus instructae), intus interdum glabratae. (i) Petioli brevissimi, 1-2 vel vix ultra 2.5 mm. longi, gemmas bene evolutas non superantes et stipulae petiolis aequilongae vel duplo longiores; amenta florifera minima vel p'arva, etiam fructifera vix ad 2.5:1-1.2 (vel in fuller tonensi ad 4 : i . 3) cm. magna. Frutex prostratus ramis repentibus; folia breviter 1 919] SCHNEIDER— AMERICAN WILLOWS 35 lanceolata, elliptico-oblonga vel oblonga, utrinque pleraque acuta, 1:0.5 ^.d 3:0.9-1 .2 cm. magna, superne vulgo sparse stomatifera; stipulae dis- tinctae; ovaria sessilia vel subsessilia, villosulo- tomentosa; stylus satis brevis, integer vel apice bifidus stigmatibus oblongis bifidis subaequi- longus; bracteae oblongae; glandula dorsalis anguste conica, interdum pl.m. bifida, iis duplo brevior; fructus subsessiles, 4.5-6 mm. longi 19. S . fullertonensis Frutices parvi erecti, ramis saepe satis brevibus subtortuosis divaricatis vel subelongatis stricti- oribus et plantae aliis signis diversae. Folia superne stomatifera, oblonga, ad 3 : i cm. magna adulta etiam subtus satis glabrata; bracteae intus glabrae, late ovales vel obovales; stylus distinctus quam stigmata vulgo ultra duplo longior; ovaria sessilia, saepe infra medium glabra; ramuli hornotini satis glabres- centes 21b. S. chlorolepis var. antimima Folia superne baud stomatifera, magis (saltem subtus) villosa vel sericeo-villosa; ramuli hor- notini semper pl.m. dense villoso-tomentosi; bracteae vulgo utrinque pilosae vel interdum extus glabriores. Ramuli hornotini densissime albo-sericeo- villosi; styli ovariorum sessilium, brevissimi, vix ad 0.5 mm. longi, stigmatibus brevibus; folia oblongo-lanceolata vel anguste ovato- lanceolata, apice (infimis exceptis) pl.m. acuta, basi rotundata vel obtusa, 1.5:0.5 ad 3.5:0.9-1 cm. magna, novella utrinque sericeo-villosa; species satis imperfecte cog- nita 20. S. niphoclada Ramuli hornotini vel vulgo tantum novelli minus dense griseo-vel subflavescenti-sericeo- villosi; styli distinctiores, stigmatibus aequi- 36 BOTANICAL GAZETTE [January longi vel saepe distincte longiores; folia elliptico-oblonga, oblanceolata, rarius ovato- vel elliptico-lanceolata, interdum obovato- oblonga, apice obtusiuscula vel subito breviter acuta, basi late cuneata vel obtusa, rarius rotundata ad subcordata, sub anthesi saepe minima vel parva, subspathulata vel lineari- lanceolata, ad 2 . 5-3 : (o . 6-) i vel 3.4:0.8 vel ad 3:1.1 (maxima ad 4.5:1.2) cm. magna; amenta sub anthesi vix ultra 10:4 mm., fructifera 1.5:0.8 ad 2.5:1(1.2) cm. magna; fructus subsessiles vel pedicello glandula vulgo duplo breviore suffulti, 5-7 mm. longi 22. S. hr achy car pa (2) Petioli gemmis vel stipulis longiores, amenta etiam florifera longiora vel plantae alio modo diversae. Folia parva vel mediocra, majora apice ramulorum vulgo baud ultra 4:1.8 cm. magna, lanceolata, oblanceolata, elliptico-lanceolata, anguste elliptica ad obovato-lanceolata, rarius elliptico-vel obovato- oblonga, apice acuta vel subacuta, basi cuneata ad rotundata, integerrima (infima minima tantum brevissime glanduloso-denticulata), superne stoma- tifera, novella pl.m. tenuiter griseo-villosula, demum glabrescentia vel subglabrata, subtus discoloria, glaucescentia, ut superne vel densius villosula (sed pubescentia satis variabili); stipulae nullae vel valde reductae ; amenta sub anthesi 8-15:5 ad 25:6 mm., fructifera ad 2-3:1-1.5 cm. magna; fructus 6.5-8 mm. longi, pedicello brevi quam glandula duplo breviore (rarius ea subaequilongo) excluso; ramuli novelli griseo-villosuli vel subtomentosi, annotini saepe subglabri, purpurasentes vel fere nigro-castanei vel ut vetustiores epidermide griseo- flavescente pl.m. secedente obtecti (conf. etiam 23. S. desertorum, speciem tantum incomplete cogni- tam) 24. S. pseudolappommi Folia majora vel latiora vel superne baud stoma- igig] SCHNEIDER— AMERICAN WILLOWS 37 tifera vel amenta fructifera majora et fructus 8-10 mm. longi peclicellis glandulam ad duplo superantibus. Folia superne haud stomatifera."* Fructus pedicello subnullo vel brevi glandu- lam haud superante suffulti, 6-8(-io) mm. longi; amenta sub anthesi 1-2.5:0.7 cm., fructifera ad 3-5 : i . 2-1 . 5 cm. magna pl.m. densa; folia (var. atra excepta) tantum | ad 2^plo longiora quam lata, elliptica, ovalia, obovato-elliptica, obovato-oblonga, vel ovato- elliptica, apice pl.m. obtusa vel breviter acuta, basi obtusa vel late cuneata ad sub- cordata, 3:2 vel 3.5:2.5 vel 4.5:2-2.3 ad 6:2.8 cm. magna, novella pubescentia (mini- mis infimis exceptis) pl.m. villosa induta, superne glabrescentia, rarius adulta utrinque glaberrima 26. S. cordifolia Fructus pedicello glandulam vulgo \ ad 2plo superante instructi, 7-8(-9) mm. longi; amenta sub anthesi 2-4:0.8-1 cm., fructifera 3 . 5-7 : 1 . 5 cm. magna, basi saepe satis laxi- flora; folia vulgo 2^ ad ultra 4plo longiora quam lata, lanceolata, oblanceolata, elliptico- oblonga, obovato-oblonga vel obovato- elliptica, apice obtusa vel vulgo acuta vel fere breviter acuminata, basi obtusa vel subito sensimve cuneata, mediocra 4.5:2 ad 5:1.5 vel 7:2.3 cm. magna, novella pubescentia satis sericea induta, superne paullo vel omnino glabrescentia, subtus saepe glabriora 25. 5. glaucae varietates Folia superne stomatifera, ceterum ut in vol. 66, p. 349 descripta 27. S. anamesa t Owing to the variability of the nos. 24-26 and our insufficient knowledge of the Greenland forms I cannot avoid using this anatomical character in distinguishing S. anamesa from the other two species. But S. anamesa is apparently a species met with only in Greenland, and therefore the arrangement of the key wUl not be incon- venient to most of the students of American willows. 38 BOTANICAL GAZETTE [january 2. SECUNDUM SPECIMINA MASCULA^ A. Filamenta omnino glabra I. Stamen unicum (rarissime stamina 2 adsunt); bracteae pl.m. purpurascentes et apice atrae, pilis longis argenteis sericeae; folia adulta glabra, discoloria, superne pl.m. nitida et stoma- tifera, pl.m. crenato-denticulata, rariter ad 2.5 cm. longa; frutex depressus vel prostratus 6. S. Uva-ursi II. Stamina semper 2. a. Fruticuli minimi, suffruticosi, ramulis tenuissimis, fere semper radicantibus; folia utrinque concoloria v. sub- concoloria, etiam superne stomatifera, minima vel parva, rarius ultra 25 mm. longa latave; amenta serotina, tenuia, pauci- (rarius multi-) flora. Folia utrinque obtusa vel acutiuscula (cuneata), Integra, nervis primariis superne tenuissime incisis subtus prominentibus, ceterum enervia, 5-8: 3-4 mm. magna; amenta 3-5-flora, rhachi bracteisque concoloribus glabris vel parcissime pilosis; glandulae 2 11. S. Dodgeana Folia utrinque rotundata vel basi cordata vel utraque facie distincte (sed saepe tenuiter) reticulata. Ramuli breves floriferi fere semper bifoliati; folia adulta sicca baud per secundum annum persistentia; ramuli vetustiores 1-2 mm. crassi. Bracteae glabrae subglabraeve, concolores, flaves- centes vel violascentes; glandulae 2; folia semper crenato-serrata 8. S. herhacea Bracteae sericeo-villosulae, fuscae, pl.m. bicolores; glandula i, ventralis; folia saepissime integerrima 5. S. polaris Ramuli floriferi 2-5-foliati; folia adulta sicca pl.m. marcescentia, fuscescentia et per secundum annum persistentia, integerrima; glandulae fere semper 2 (in S. cascadensi ut videtur tantum i). 5 Of the following species the male plant is still unknown: fullertonensis , hud- sonieiisis, leiolepls, Ungulala, Peasei, and ve.nusla. 1919] SCHNEIDER— AMERICAN WILLOWS 39 Amenta minima, 3-8-flora; folia vulgo orbicularia vel late ovalia, utrinque rotundata, vix ad 11:10 mm. magna, adulta sicca anno secundo decidua 9. S. rotundifolia Amenta multiflora, ad 23 mm. longa; folia adulta plures annos persistentia vel pl.m. lineari-lanceolata. Folia elliptica, obovato-oblonga vel late spathu- lata, apice rotundata vel breviter acuta, basi cuneata, subtus vix pallidiora, ad 15:9 mm. magna 19. S. phlehophylla Folia lingulata, lineari vel anguste lanceolata (vel oblanceolata) , apice vulgo subplicato-acuminata^ ad 18:5 mm. magna, subtus subpallidiora 12. 5. cascadensis b. Fruticuli vel frutices vel folia discoloria vel plantae alio modo diversae. I. Bracteae pl.m. discolores, apice vel pro parte maxima fuscae, versus apicem longe sericeae (pilis longis sericeis quam bractea vix vel paullo brevioribus instructae), interdum apice tantum longe ciliatae; fruticuli vel frutices parvi, ramis prostratis, pl.m. radicantibus vel subterraneis, ramulis tantum floriferis pl.m. adscendentibus, vix ad 15-20 cm. altis (vel in S. arcticae formis interdum altioribus); si folia sunt tenuiter sed dense at acute glanduloso-serrulata, vide 29. S. Chamissonis a) Folia subtus (in sicco) paullo pallidiora, leviter elevato-nervata, sed vix reticulata ceterum ut in p. 56 descripta; amenta 1-2.5 cm. longa, vix ad i cm. crassa, pluri- vel multiflora; glandula ventralis pl.m. elongato-conica, dorsalis saepe nulla 13. S. petrophila j8) Folia subtus distincte discoloria, glaucescentia vel albescentia, adulta pl.m. reticulata vel longius petiolata vel majora et forma diversa (specimina mascula specierum sequentium sine speciminibus 40 BOTANICAL GAZETTE [january femineis accurate discernere saepe impossibile est).^ (i) Glandula ventralis satis brevis lataque, vix 2plo longior quam lata, apice late truncata, quam bractea obovata vulgo 2^-3plo brevior;^ amenta 2-2 .5:0. 8-1 cm. magna; folia tantum novella subtus sparse sericea, cito glabra, superne estomatifera, ceterum ut in p. 57 descripta; petioli vix ultra 8 mm. longi 18. S. arctophila (2) Glandula ventralis pl.m. anguste ovato- rectangularis vel anguste conica, apice saepe leviter incrassato truncata, quam bractea vix duplo brevior vel folia superne stomatifera, semper integerrima vel plantae aliis signis diversae. Amenta perfecte evoluta vix ultra 1.5:0.8 cm. magna; folia matura satis crasse papy- racea, subtus pl.m. perspicue et anguste reticulata, vulgo elliptica, late elliptica, obovalia vel rotundata, vix ultra 2,5:1. 2-2 cm. magna; stipulae nullae (vel minimae punctiformes) (si folia superne stomatifera confer, etiam 14. S. anglorum formas et30. S. glacialem speciemvalde incomplete cognitam). Folia superne pl.m. stomatifera; ramuli annotini hornotinique vulgo breves; rami saepe stolones subterraneos tenues vix ad I mm. crassos emittentes. .17. S. stolonifera Folia superne baud (var. camdensis excepta) stomatifera; ramuli hornotini annotini que pl.m. elongati, rami ut videtur baud sto- loniferi 16. S. ovalifolia '5 Of course if their exact locality is known, we shall certainly be able to determine -even young male branchlets. ^ The shape of the ventral gland is often rather variable in any species because the gland may be more or less lobate, bifid, or bipartite. iqiq] SCHNEIDER— AMERICAN WILLOWS 41 Amenta perfecte evoluta vulgo ultra 1.5 et ad 4-5 cm. longa et i-i .3 cm. crassa, bracteis longe et dense pilosis satis sericea; folia matura majora vel tenuiora et subtus vix reticulata, juvenilia magis sericea vel sericeo- villosula, vel stipulae (saltem in ramulis vegetis) pl.m. distincte evolutae. Folia superne baud stomatifera (var. suh- cordata excepta quae foliis amentisque maximis a S. anglorum valde differt) ceterum ut in p. 56 descripta; amenta I -5-5: 1 -3 cm. magna; ramuli peduncu- liferi vulgo 2-4 mm. crassi (confer etiam i6c. 5. ovalij'oliam, v. suhardicam) 15. S. arctica Folia superne stomatifera, ceterum ut in p. 56 descripta; ramuli pedunculiferi i ad vix 2 mm. crassi (confer etiam 27. 5. anamesam e Groenlandia) . . 14. S. anglorum 2. Bracteae pl.m. concolores, fiavescentes et fere glabrae vel stramineae flavobrunnescentesve et pl.m. breviter sericeo-villosae (pilis quam bractea brevioribus instructae) vel subdiscolores sed tantum villosae vel pilis sericeis vulgo tenuissimis praeditae; frutices prostrati vel saepe erecti, 0.3-1 m. alti. Amenta minima vel parva, 5-10 mm. longa vel in niphodada ad 22:4 mm. magna et laxifiora; petioli vix ultra 2 mm. longi. Folia' parva, vix ad 2.5 cm. longa et ad 1.4cm. lata, superne stomatifera, stipulae ut videtur nullae; bracteae fiavescentes vel stramineae, subglabrae vel extus dense breviter pilosae; glandulae 2 21. S. chlorolepis Folia saepe ad 4 cm. longa, superne baud stoma- tifera; stipulae vulgo evolutae; bracteae strami- neae, utrinque brevipilosae ; glandulae 2 20. S. niphodada 42 BOTANICAL GAZETTE [January Amenta 1.2-3.50111. longa, ultra 5mm. crassa; petioli vulgo ultra 2 mm. longi; confer formas diver- sas sub 24. 5". glauca, 25. S. cordifolia, et 26. 5. ana- mesa enumeratas. B. Filamenta pl.m. pilosa (interdum ima basi tantum pilis paucis instructa) I. Amenta serotina, pseudoterminalia, anguste cylindrica vel minima pauciflora, pedunculis nudis iis saepe subaequilongis suffulta; bracteae breviter (rarius longius) pilosae, pl.m. concolores, flavescentes vel violaceae; folia satis crassa, superne saepe pl.m. inciso-reticulata, baud stomatifera, subtus distincte elevato-reticulata, pl.m. longe petiolata; glandulae 2 vel inter- dum plures pseudodiscum lobulatum formantes. Bracteae intus pl.m. brevipilosae vel utrinque sericeae; folia coriacea, pleraque vix longiora quam lata vel ultra 4 cm. longa, superne conspicue inciso-reticulata, rugosa. Frutex prostratus; folia cito glaberrima vel rarius pilis paucis sericeis obsita, vix ad 5 : 5 . 5 cm. magna et petiolis elongatis ad 3 cm. longis instructa; bracteae intus tantum brevi-pilosae ; antherae violaceae i. S. reticulata Frutex prostratus vel erectus; folia etiam adulta subtus dense sericea vel major a, magis oblonga, margine distincte crenata, vel petioli breves gemmis vix longiores: bracteae utrinque pl.m. sericeae; antherae flavae 2. S. vestita Bracteae glabrae vel tantum parce ciliatae; folia tenuiora, minima vel distincte longiora quam lata, vix ultra 3.5:2 cm. magna, superne vix vel indistincte inciso-reticulata; frutices prostrati, interdum minimi, suffruticulosi 4. S. nivalis et var. saximontana II. Amenta coetanea, rarius serotina, lateralia, ovata vel cylindrica, semper multifiora, sed interdum parva, pedunculis semper foliatis. Folia brevissime petiolata, petiolis vix ultra 2.5 mm. longis vel quam gemmae evolutae pl.m. brevioribus, ceterum ut in p. 58 indicata; amenta sub anthesi 5-io(-i5) :2-8(-9) mm. magna; antherae minimae, ellipsoideo-globosae 22. S. hr achy car pa 19 1 q] SCHNEIDER— AMERICAN WILLOWS 43 Folia distinctius petiolata; amenta vulgo ultra 15 mm. longa, crassiora vel antherae magis ellipsoidales oblongiores et folia superne pl.m. stomatifera (specimina mascula specierum se- quentium sub anthesi sine foliis perfecte evolutis accurate discernere saepe impossibile videtur; confer etiam 28. 5. lin- gulatam speciem incertam alaskanam). Folia superne pl.m. (interdum tantum sparse secundum nervos) stomatifera, vulgo oblonga, 2|-4plo longiora quam lata, sed vix ultra 4:1.8 cm. magna. Amenta pedunculis vix ad i cm. longis suffulta, 8-15:7 mm. magna, densifiora; folia pedunculorum (saltem sub- tus) dense breviter sericea vel villosula (confer etiam 27. S. anamesam e Groenlandia) . ... 23. 5. pseiidolap ponum Amenta saepe longius pedunculata, i . 5-3 .5:0. 8-0 . 9 cm. magna et basim versus laxiflora vel folia pedunculorum etiam subtus glabra vel subglabra 24. 5. desertorum Folia superne baud stomatifera, majora, latiora vel longiora. Amenta deflorata ad fere 3 . 5 cm. longa, basi vulgo pi. m. laxiflora vel folia pedunculorum satis oblonga, circ. 2^-3^ plo longiora quam lata 25. S. glaucae varietates Amenta deflorata vulgo baud ultra 2 (-2. 5) cm. longa, etiam basi densiflora vel folia pedunculorum latiora brevioraque vix ad 2|plo longiora quam lata 26. S. cordifolia Enumeratio sectionum specierumque I have omitted from the keys and the following enumeration the well known and easily recognizable S. Candida Fliigge which Ball (1909) includes in his section Arcticae, because I attribute to it a different systematic position. Sect. I. Reticulatae^ Fries in Sylloge PI. Nov. Soc. Ratisb. 2:38 (Consp. Disp. Sahc. Suec). 1828, quoad S. reticulata; for * There is the older name Chamaetia given by Dumortier in Bijdr. Natuurk. Wetensch. 1:56 (Verh. Gesl. Wilgen 15) 1835 to a group, including S. rctusa, S. her- bacea, and S. reticulata. Unfortunately neither the International Rules nor the Philadelphia Code contains a precise rule in regard to the application of names of sections or similar groups. I do not accept Dumortier's name because in his paper he proposes two very different arrangements, and he does not in my opinion make a definite statement. 44 BOTANICAL GAZETTE [January further literature see Schneider in Sargent, PI. Wils. 3 : 146. 1916. — This section, which is represented in America by the following 4 species, is a well defined group. The rather exceptional position of S. reticulata among the other willows, which once led A. Kerner to propose the new genus Chamitea for it, becomes less marked by the addition of the American S. nivalis; and the characters of the Reticulatae are further changed by the inclusion of S. leiolepis with glabrous ovaries. S. glacialis referred by Rydberg to this section is a very imperfectly known species, of which the systematic position is still doubtful. I. S. RETICULATA L., Sp. PI. 2:ioi8. 1753. — S. reticulata a glabra Trautvetter in Ledeb., F1. Alt. 291. 1833. — S. reticulata b normalis And. in Ofv. K. Vet.-Akad. Forh. 15:133 (Bidr. Kanned. Nordam. Pilarter). 1858.^ — 5. reticulata a typica i. glabra And. in DC, Prodr. 16^:301. 1868. — This willow has the most extensive range of all the known species. In Europe it is reported from the high Pyrenees through the whole range of the Alps to the mountains of Croatia, and northward to Scotland, Scandinavia, Iceland, Spitzbergen, and Arctic Russia, while in Asia it is found on the high mountains from the Ural to Kamchatka and in the Arctic zone from Taimyr Peninsula to the Bering Strait. According to Lange it does not occur in Greenland. In North America I have seen specimens from southern Labrador, western Newfoundland, the northern shores of the Hudson Strait, and the western shore of Hudson Bay to the Coronation Gulf and Bernard Harbor (ii4°46' W. long.), and west of 135° W. long, from the Yukon Territory (King Point and Herschel Island to Lake Bennett) and from Alaska. Here it stretches, as Coville has said, over the Arctic zone, but including the extreme north (Camden Bay), and southward it occurs at timber line on the mountains from the Juneau region to Kodiak Island, and westward to the Aleutian, Pribilof, and St. Matthew Islands. There is also a specimen from the ''Rocky Mountains" (no. 85 Herb. H.B.T., ex Herb. Torrey in N.; m., f.), the exact locality of which is unknown to me. No. 86 also 9 The same article has been published with slight alterations in the same year in Proc. Amer. Acad. 4: 50 (Salic. Bor.-Am.) and in Walper, Ann. Bot. 5:744. I do not always repeat these quotations. iqiq] SCHNEIDER— AMERICAN WILLOWS 45 of the same collection is a female specimen which seems to represent a very small form of S. reticulata somewhat similar to 5. nivalis, but showing a distinct reticulation of the leaves. It needs further observation and has already been mentioned by Rydberg (1899), who also cites a specimen of Macoun (18849, 0-) from Silver City in the Rockies, a locality I have not yet been able to identify. Otherwise it is replaced in the Rockies by S. nivalis and var. saximontana. The name 5. orbicularis has been given by Andersson (in DC, Prodr. 16^:300. 1868) to the S. reticulata "in Kamtschatka et in America boreah-occidentali ut ad Sitchka et Unalaschka." Ryd- berg (in Bull. N.Y. Bot. Card. 1:260. 1899) accepted this name for almost all the American material of S. reticulata, citing only three specimens "which may be referred" to Linnaeus' species. But I agree with Coville (in Proc. Wash. Acad. Sci. 3:342. 1901) that the distinguishing characters assigned by Andersson as well as those given by Rydberg are insufficient " to see in our American plant a species distinct from the European." If we wish to dis- tinguish the form with more or less orbicular leaves we may use the name S. reticulata subrotunda Seringe (Essai Mon. Saules Suisse 29, 1815) based on S. reticulata Hoffmann (Hist. Salic, i: pi. 2^, fig. J. 1787). Other variations of this species which have been observed in the Old World are not represented among the American material before me and are not mentioned by previous authors. 2. S. VESTiTA Pursh, Fl. Amer. Sept. 2:610. 1814. — 5. reticu- lata a vestita And. in Ofv. K. Vet. -Acad. Forh. 15:133. 1858, excl. specim. e Siberia et Helvetia. — S. vestita a humilior And. in DC, Prodr. 16^:300. 1868, excl. specim. altaica. — The range of this well known species extends (including the form mentioned later) in the east from northern Labrador southward to western Newfoundland, Anticosti, and the Gaspe Peninsula, and I saw it also from the west shore of Hudson Bay (Churchill, Ig. /. M. Macoun, no. 79143, O., m., f . ; Cor., G.) , while in the west it reaches its northern limit in the Rockies of Alberta and British Columbia toward the 5 2d parallel, extending southward to northwestern Montana and the Wallowa Mountains in eastern Oregon. The western forms have been called 5. Fernaldii by Blankinship (in Mont. Agric. Coll. Sci. Stud. 46 BOTANICAL GAZETTE [January 1:46. 1905), which name is adopted byRYDBERG (Fl. Rocky Mts. 198. 191 7), although he says "perhaps not distinct from the eastern S. vestita Pursh." It certainly cannot be distinguished specifically, and if we intend to apply a special denomination to the more erect form with rather "thinner, narrower, rounded or pointed leaves," we have to use the name var. erecta And. (in DC, Prodr. 16^:300. 1868). The aments are usually longer than in the eastern form, but there are specimens before me from Alberta (Ig. Rehder and also Jack) with the same short fruiting catkins. Professor Fernald kindly pointed out that the shape of the capsules of typical vestita is more ovoid-conical, with a rather pointed apex, while it is more ovoid-ellipsoid, with an obtuser apex, in var. erecta. There seem to be rather intermediate forms, but as a whole this character may be taken for the best one to distinguish these eastern and western forms. Another form has been collected by Fernald and St. John in western Newfoundland which seems closely connected with the next species, but its description has not yet been published and it needs further observation. 3. S. LEiOLEPis Fernald in Rhodora 16:178. 1914.— This is a very peculiar species, which was discovered by Fernald and St. John on the Table Mountain, Port a Port Bay, in western New- foundland, July 17, 1914, on "mossy knolls on the limestone table- land, alt. 200-300 m." (no. 10825, fr.; G.). In habit and foliage it closely simulates, as the author said, S. reticulata and the most dwarfed alpine extreme of S. vestita; but it differs "from both in the glabrous scales and capsules; also from 5. reticulata in its short peduncles and thick fruiting aments, and from S. vestita, which is the most abundant willow of Table Mountain, in its glabrous or quickly glabrate foliage and the smaller and more slender, glabrous, greenish terminal buds." As the type specimen shows, the ovaries are sometimes sparsely pubescent, the bracts frequently provided with a few cilia, the styles very short but more or less distinct, and even the old leaves bear some hairs on the lower surface which are often rather difficult to recognize. Unfortunately the male sex is still unknown; consequently I cannot decide whether S. leiolepis is to be regarded as a good species or as a glabrate variety of S. vestita, 19 iq] SCHNEIDER— AMERICAN WILLOWS 47 representing a rather dwarfed alpine form. The glabrousness or pubescence of the ovaries, a character on which usually so much reliance is placed, cannot always be taken for a decisive taxonomic character. In my notes on the species of the section Ovalifoliae [I.e.) I was able to show that many species with hairy ovaries develop a more or less glabrescent or glabrous variety, or vice versa. 4. S. NIVALIS Hooker, Fl. Bor.-Am. 2:152. 1839; Nuttall, N. Am. Sylva i : 77. pi. 19, fig. sinistra inferior. 1843 ; Rydberg in Bull. N.Y. Bot. Card. 1:262. 1899; Ball in Coult. and Nels., New Man. Rocky Mt. Bot. 139. 1909. — S. reticulata c nana And. in Ofv. K. Vet.-Akad. Forh. 15:133. 1858, excl. specim. e Groenl. et Spitzb. — 5. reticulata fi nivalis And. in DC, Prodr. 16^:301. 1868. — The type of this "elegant and very diminutive shrub" (Nuttall) was collected "near the summits of the peaks in the Rocky Mountains" by Drummond between lat. 52-56°. It occurs most frequently in the alpine region of the Rockies of Alberta and British Columbia, and to a certain degree also on the Tobacco Root Range (Pony Mountains), on Observation Mountain and Mt. Chauvet in southern Montana, and on the Electric Peak in northern Yellow- stone Park. It is also mentioned by Piper from Mt. Rainier, Washington. There are a few "specimens from Colorado (E. L. Greene, no. 517, m.: G.; probably from near Golden City) and from southeastern Utah (Rydberg and Garrett, no. 8787, m., f.; N.; La Sal Mts., West Mt. Peale) which I can hardly distinguish from t^-pical S. nivalis, and which, in my opinion, form connecting links between it and 5. saximontana Rydbg. I take this last species, therefore, only for a variety of S. nivalis, from which it chiefly differs by the characters given later. Rydberg himself said (1899) that 5. nivalis "perhaps represents only the most de- pauperate form" of his S. saximonta^ia, and he repeats in his Cat. FL Mont. 112. 1900 that the latter "seems to grade into S. nivalis," while such an accurate observer as Piper (in Contr. U.S. Nat. Herb. 9:216. 1906) states that "vS. saximontana probably is not specifically distinct from 5. nivalis.'^ 4b. S. nivalis var. saximontana, nov. var. — S. reticulata Bebb in Coulter, Man. Rocky Mt. Bot. 339. 1885, non L. ; Ball in Trans. St. Louis Acad. Sci. 9:90. 1899. — S. saximontana Rydberg in Bull. 48 BOTANICAL GAZETTE [januasy N.Y. Bot. Gard. 1:261. 1899; Ball in Coult. and Nels., New Man. Rocky Mt. Bot. 139. 1909. — S. aemulans v. Seemen in Bot. Jahrb. 29: Beibl. 65:28. 1900. — A typo praecipue differt: ramis crassiori- bus (interdum ad 19mm. crassis), foliis majoribus vulgo 1.5-3.5 cm. longis et 0.8-2 cm. latis interdum minus reticulatis forma ut in typo valde variabilibus (lanceolatis saepe apice subacutis), petiolis ad 1 .8 cm. longis; amentis pluri- vel multifloris masculis 7-12 mm. longis pedunculis vulgo longioribus sparse pilosis exclusis, femineis 8-18 mm. longis fructiferis fere ad i cm. crassis. The type came from Gray's Peak in Colorado (Ig. Rydberg, August 1895, m.; N.), and it is most abundant in the Rockies of this state, reaching its southernmost point on the Truchas Peak of the Taos Mountains in northern New Mexico. Northward its range extends through western Wyoming, Yellowstone Park, and southern Montana to the vicinity of Laggan in Alberta and Skagit Valley in British Columbia, while toward the west it is found on Mt. Rainier in Washington, on the Strawberry and the higher Wallowa Mountains in eastern Oregon, furthermore on the East Humboldt Mountains in northeastern Nevada, and also in Boxelder and Utah counties and on the La Sal Mountains in Utah. Sect. 2. Herbaceae Borrer in Hooker, Brit. Fl. 432. 1830 (Sect. Chamaetia Dumortier, pro parte, see note 8 on p. 43 ; sect. Retusae Kerner in Verh. Zool. Bot. Ges. 10:195 [Niederostr, Weid.]. i860; for further literature see Schneider in Sargent, PL Wils. 3:142. 1916). — As I have already explained {I.e. 143), it seems to me impossible to separate the sect. Retusae from the Herbaceae, but there are the following species which have been added to this group or might be regarded as closely related to its members: S. cascadensis, S. Dodgeana, S. glacialis, S. Peasei, S. polaris, S. phlebophylla, S. rolundifolia, and 5. Uva-ursi. Of these species S. rotundifolia seems to show the most intimate affinity with S. herhacea, but it is distinguished by the persistent leaves, a character also found in S. phlebophylla, S. caseadensis, and S. Uva- ursi. From the last three species 6*. Uva-ursi seems to be widely separated by its bicolor leaves and the single stamen of the male flowers, while on the other hand they all have bicolor or fuscous bracts which are concolor, greenish or yellowish (or partly purplish 19 1 9] SCHNEIDER— AMERICAN WILLOWS 49 toward the apex) in S. herbacea, S. Dodgeana, and S. retusa. The systematic position of 5. polaris is even more puzzhng. Some authors are incUned to regard it as nothing but a variety of S. herbacea because the vegetative characters of both are so similar; but if we base our opinion on the flowers we may come to a very different conclusion; and in Sargent, I.e. 319, I have included this species in the sect. Myrsinites Borrer. If we pay much attention to the presence or absence of a dorsal gland in the male flowers, we might also refer 5. Uva-ursi to this section, but this willow occupies a rather unique position among the American species. With my present knowledge I deem it best to leave the question of the correct limitation of this section and of the true systematic position of these species undecided until I have had opportunity to discuss this problem with such an eminent sahcologist as S. J. Enander, who is preparing a monograph of the whole genus. I have already published (Oestr. Bot. Zeitschr. 65:273. 1915) a short note on the systematic arrangement of the genus and discussed briefly the views taken by Andersson and von Seemen. The main purpose of that note was to show that no systematic grouping on natural lines can be attained unless we make use of every taxonomic character. 5. S. POLARIS Wahlenberg, Fl. Lapp. 261. pi. 13. fig. i. 1812; Rydberg in Bull. N.Y. Bot. Gard. 1:264. 1809; Coville in Proc. Wash. Acad. Sci. 3:335- fig- 27. 1901.— ? S. herbacea var. polaris Kurtz in Bot. Jahrb. 19:475- 1894.— In America this species is only known from the Alaskan coast of Bering Strait, where it has been collected at Port Clarence and Cape Vancouver. I have seen only the Port Clarence specimens of Trelease and Saunders (nos. 3387, f., 3385% m.), which have been described by Coville. They seem to agree with specimens of S. J. Enander's Sahc. Scand. Exsicc. from Spitzbergen, especially with no. 12 "modificatio foHis subovalibus." The ovaries of no. 3387 are partly glabrate, and I cannot at present say whether the American S. polaris is the typical form or not. As to its uncertain systematic position see my preceding remarks. Lundstrom (apud Kjellman in Nordenskjold, Vega Exp. Vet. lakt. 2:21 [Fanerog.-Fl. St. Lawrence-Oii.]. 1883) has described a S. polaris f. subarctica "fohis tenuioribus, subtus margineque pilis 50 BOTANICAL GAZETTE [january longis parce adspersis, squamis atris, obtusis; stylo elongate," the type of which was collected by Kjellman, July 31 to August i, 1879, on the northwestern shore of St. Lawrence Island. I cannot interpret this form without having seen the type, and it is not mentioned by Coville or Rydberg. There is, however, a specimen before me collected by R. L. Shainwald, Jr., on Mt. McKinley, 1200 m., August 26, 1903 (fr. ; N.), which is very similar to S. polaris in every respect, but the fruiting aments measure up to 3 cm. in length and 9 mm. in width. The sessile fruits are 5-6 mm. long, pubescent only toward the apex, and the distinct withered style is a little longer than the stigmas. It looks to me like a new variety of 5. polaris or a new species. 6. S. UvA-URSi Pursh, Fl. Amer. Sept. 2:610. 1814; Rydberg in Bull. N.Y. Bot. Gard. 1:278. 1899; Britton and Brown, 111. Fl. 1:601. Jig. 1477. 1913; Robinson and Fernald, Gray's New Man. 325. fig. 656. 1908. — S. Cutleri Tuckerman in Amer. Jour. Sci. 45:36. 1843; And. in Ofv. K. Vet.-Akad. Forh. 15:132. 1858; in DC, Prodr. 16^:292. 1868. — 5*. Myrsinites var. parvifolia Lange, Consp. Fl. Groenl. 1:108. 1880; 2:278. 1887; Fl. Dan. i7:fasc. 51:13. pi. 30JJ. 1883, non And. — S. ivigtutiana Lundstrom apud Berhn in Ofv. K. Vet.-Akad. Forh. 41:89. 1884. — A well known willow, of which, however, as I said before, the systematic position is by no means settled. Andersson said, ''AS. arhuscula recedit foliis minimis parte superiore serrulatis, amentis subterminalibus et capsuhs glaberrimis. Longius a S. retusa distat. Si ut ferunt auctores americani, fiores masculi staminibus tantum singulis praediti sunt, tum afhnitas cum 5. coesia esset, cui etiam rigiditate et glaucescentia foliorum non absimihs, sed folia subserrata et capsulae glabrae." I have very rarely found two stamens in one flower, and I am at present unable to give a precise opinion as to the real relationship of this pecuHar species. Its range stretches from New York (Mt. Marcy and Mt. Whiteface), Vermont (Camel's Hump, Mt. Mansfield), New Hampshire (White Mts.), and Maine (Mt. Katahdin) northward to the Gaspe Peninsula, southern and western Newfoundland and the whole coast of Labrador, apparently reaching the northern limit of its range at the southern shore of Bafhnsland. Westward it extends through Ungava to the eastern 19 19] SCHNEIDER— AMERICAN WILLOWS 51 shores of Hudson Bay. The species has also been reported from southwestern Greenland by Durand (in Jour. Acad. Nat. Sci. Phil. II. 3:197 [PI- Kaneanae Groenl.] 1856), but Lange (Consp. Fl. Groenl. 1:108. 1880) made the following statement: "S. Uva Ursi Durand ex descriptione (PI. Kan.) ad formam nostram 3 [S. arcio- phila lejocarpa] retuli, etsi specimina Kaneana non vidi, quare dijudicare nequeo, an forsan ad veram S. Uva Ursi Pursh .... pertineant, species, quae tamen a nemine aHoquin in Groenlandia lecta esse constat." I am inclined to believe that Kane's specimens represented the true 5". Uva-ursi, because this species is apparently identical with 5. Myrsinites parvifolia Lange and S. ivigtiitiana Lundstr. Lange said that he did not see specimens of the t>^ical S. Myrsinites L. from Greenland, and his var. parvifolia seems to be distributed from about the 70th parallel to the very south in Green- land. I have seen one specimen from the Tunugdharfik Fjord, Kingua, Ig. L. K. Rosenvinge, August 17, 1888 (fr.; G.), which fully agrees with Lange 's and Lundstrom's descriptions, and, in my opinion, cannot be separated from 5. Uva-ursi, the presence of which may be expected in this part of Greenland. The leaves are distinctly glaucous beneath, while they are green and shining on both sides in 5. Myrsinites. Andersson's var. lahradorica (1868) from near Oakak on Labrador is scarcely different from the type. He based it on Hohenacker's no. 92% which I have not yet seen. On the other hand, the forms referred to 5". Uva-ursi by Hooker (Fl. Bor.-Am. 2:152. 1839) seem to belong at least partly to 5. arcto- phila, but I have as yet seen only a few leaves of Morrison's speci- men in Herb. Bebb in C. which do not have stomata in the upper epidermis, as is always the case in Pursh's species. 7. S. Peasei Fernald in Rhodora 19:223. 191 7. — This willow is known only from the type locahty in New Hampshire, southwest gully of King's Ravine on Mt. Washington, where it was first collected by Pease (no. 12091; the type is Fernald and Pease, no. 16847, fr.; G.). It is certainly a very peculiar species, and needs further observation, the male plant being still unknown. Fernald's description is, as usual, ample and fitting. I am almost sure that it has to be regarded as of hybrid origin. I visited King's Ravine on September 18, 1918, and I found the willow growing in 52 BOTANICAL GAZETTE [january about the altitude given by Fernald on wet cliffs in company with S. herhacea. The main part of S. Peasei I saw was growing about 15-30 m. below S. herhacea on the southern slope of the ravine and covered a rather large area. S. Uva-ursi is very common at a somewhat higher level, but I collected plants of it which were grow- ing just above the place where I saw S. herhacea and S. Peasei close together. Some plants of S. Peasei looked much like vigorous S. herhacea, while the main part of it lower down at first sight could easily be taken for S. Uva-ursi. Fernald states that S. Peasei "finds itself at home on the almost inaccessible wet cliffs," and perhaps he did collect it somewhere else (but apparently not far from where I found it), because this place is by no means "almost inaccessible." Anyone who is a Kttle careful not to start a stone avalanche and is not afraid of some steep cHmbing can easily visit this locality. Unfortunately the weather became misty and pre- vented my exploring the southwestern part of the ravine to a greater extent. On the southeastern slopes (toward the Madison Huts) I could not find a trace of either S. herhacea or S. Peasei, both species seeming to inhabit a very limited area. Under the lens the leaves, above as well as beneath, look as though very finely punctate owing to the presence of stomata. They are not "papillose," as the author says. These stomata are also present in S. Uva-ursi, but usually hardly visible except under the microscope. 8. S. HERBACEA L., Sp. PL 2:1918. 1 753; Lange, Consp. Fl. Groenl. 1:107. 1880; 2:278. 1887; Rydberg in Bull. N.Y. Bot. Gard. 1:277. 1899; Robinson and Fernald, Gray's New Man. 325. fig. 65. 1908; Britton and Brown, 111. Fl. ed. 2. i :6oi. fig. 1478. 1 91 3.— It has almost exactly the same range as S. Uva-ursi; it does not occur, however, in New York and Vermont, and I have not seen specimens from Newfoundland. On the other hand, it is met with on the western shores of Hudson Bay and in western as well as eastern Greenland. It is entirely absent from western North America, but in Europe and Asia its range is even more extensive than that of S. reticulata. Pursh (Fl. Am. Sept. 2:617. 1814) cites a specimen of D. Nelson from "the northwest coast" which I cannot identify. 1919] SCHNEIDER— AMERICAN WILLOWS 55 9. S. ROTUNDiFOLiA Trautvetter in Nouv. Mem. Soc. Imp. Nat. Mosc. 2:304. pi. II (De Salic. Frig. Kochii). 1832; Andersson in DC, Prodr. 16^:299. 1868; Rydberg in Bull. N.Y. Bot. Gard. 1:276. 1899; Wolf" in Izv. S. -Petersburg Liesn. Inst. 5:112. pi. j8. Jig. 15-20. pi. 46. fig. y-g. 1900. — S. polaris var. leiocarpa Chamisso in Linnaea 6:542. 1831. — S. retusa var. rotundifolia Treviranus ex Trautv. in Nouv. Mem. I.e., pro synon.; Trautvetter in Middendorff, Reise Sib. i. pt. 2. Bot. Abt. 1:152 (Fl. Boganid. Phaen.) 1847. — ^- rotundifolia a typica Lundstrom in Nov. Act. S. Sci. Ups. ser. 3. 1877. 30. fig. j. — S. leiocarpa Coville in Proc. Wash. Acad. Sci. 3 :338. pi. 41. fig. 2. 1901. — ''This charming Uttle plant .... grows on the islands and both shores of Bering Sea and the Arctic Ocean, and above timber line on the Pacific coast of Alaska eastward to Prince William Sound," and to these localities given by Coville is to be added Collinson Point on Camden Bay, where it was collected by F. Johansen, June 13, 1914 (no. 39 or 9381 1 O., m.). The typical form has glabrous ovaries, but two of the specimens before me represent a form with more or less hairy ovaries which I deem best to keep distinct as f. pilosiuscula, f. nov. (ab typo ut videtur nonnisi ovariis partim vel omnino villosis differt). As type may be taken no. 3382 of Trelease and Saunders, from Hall Island, July 14, 1809 (f. ; M.), to which no. '3383 from Matthews Island, July 15 (f. ; M.), is to be added. The last specimen has a little longer styles with more or less slender stigmas, thus somewhat resembling S. stolonifera, but otherwise not differing from S. rotundifolia. 10. S. PHLEBOPHYLLA Auderssou in DC, Prodr. 16^:290. 1868; Coville in Proc. Wash. Acad. Sci. 31336. fig. 28. 1901. — S. anglorum Chamisso in Linnaea 6: 541. 1 831, pro parte, quoad specimin. citata; Trautvetter in Act. Hort. Petrop. 6:37. 1879. — S. buxifolia Trev. apud Trautv. inNouv. Mem. Soc. Imp. Nat. Mosc. 2:301. pi. 10. 183 2, "> Wolf, one of our best salicologists, who was curator of the Imperial Institute of Forestry at Petrograd, at least until the outbreak of the war, has made some extremely valuable studies on European and Asiatic willows. Unfortunately his papers are written in Russian, but they are accompanied by excellent sketches. The title of his main paper is (translated) "Materials toward the study of the willows native to European Russia," which appeared in two parts in 1900 in vols. 4 and 5 of the periodical quoted. 54 BOTANICAL GAZETTE [january non Willd. apud Schleicher." — S. retusa Hooker and Arnott, Bot. Beech. Voy. 130. 1832, non L. — S. arctica /3 minor Ledeb., Fl, Ross. 3:619. 1849-51. — S. {retusa) phlehophylla And. in Ofv. K. Vet.- Akad. Forh. 15:131. 1858, pro parte maxima. — S. arctica /3 huxi- folia [recte minor] Ledeb. ex And. in DC, I.e. 290, pro synon. — S. palaeoneura Rydbg. in Bull. N.Y. Bot. Gard. 1:267. 1899. — The history of this peculiar willow, which had been thoroughly described by Trautvetter, is well given by Coville. Andersson's "spe- cies" is always quoted from 1858, but at this time he published the name phlehophylla only as a varietal designation, distinguishing 3 forms: major, media, and minor, which he rightly omitted in 1868. I saw a photograph and fragments of all his types pre- served in Herb K. The f. minor may be represented by a speci- men like Tur7ier's no. 1293 in part (f. ; G.), collected in 1879 on Atka Island. Turner (Contr. Nat. Hist. Alaska 75. 1886) men- tions a S. rotundijolia var. retusa from Atka Island, "with its heads of cottony catkins peering just above the surface of the other vegetation." I am not quite sure whether he refers to the form before me or not. Rydberg has mentioned the specimen which I have seen under S. Dodgeana, but the leaves do not show the finely impressed veins on the upper surfaces, and the female flowers are very similar to those of S. phlehophylla, having, however, a dorsal gland. This form needs further observation, and the species has not yet been recorded so far south in Alaska, where it inhabits the northwestern and northern coast from Norton Sound to Point Barrow. According to Coville it was also collected on the Porcupine River by Turner in 1891, and Seemann (Bot. Voy. Herald 40. 1852) reported it from Belly's Island at the mouth of the Mackenzie River. On the Siberian coast of Bering Strait C. Wright collected it on Arakam (or Kayne) Island. I have also seen a specimen from near Glacier in southeastern Alaska collected in June 1914 by Mary Milvain (m., f. ; A.). II. S. Dodgeana Rydbg. in Bull. N.Y. Bot. Gard. 1:277. 1899; Fl. Rocky Mts. 195. 191 7; Ball in Coult. and Nels., New Man. " According to Seringe, Essai Men. Saules Suisse 54. 181 5, the name S. bitxifolia Willd. was used first by Schleicher, Cat. Sal. i. 1809, where it is a nomen nudum, as well as in ed. 3 of Schleicher's Cat. PI. Helv. 25. 1815. This name has to be used, so far as I understand it, for the hybrid 5. glaiicaXS. reticulata; see Brand in Koch's Syn. D. Schw. Fl. ed. 3. 3:2357. 1907. 1919] SCHNEIDER— AMERICAN WILLOWS 55 Rocky Mt. Bot. 131. 1909.— This delicate suffruticose species is only known from the type locality, Electric Peak, in the northeast corner of Yellowstone Park (Ig. Rydberg and Bessey, August 18, 1897, f.; N.) and from Sheep Mountain in the Teton Forest Reserve, Wyoming (F. Tweedy, no. 292, fr.; N.)- Rydberg calls it "the smallest willow in existence," but there are similar minims among the European (5. ser pyllifolia Scop.) and Himalayan species (5. oreophila var. secta And.) ; see Schneider in Sargent, PI. Wils. 3:146. 1916. The systematic position of 5. Dodgeana is not yet quite understood; it seems to represent a rather singular type among its American congeners. As to a doubtful pubescent variety^ see my remarks under the following species. 12. S. cascadensis Cockerell in Muhlenb. 3:9. 1907; Rydberg, Fl. Rocky Mts. 198. 191 7.-5. tenera And. in DC, Prodr. 16^:288. 1868, non A. Braun (1850); Rydberg in Bull. N.Y. Bot. Gard. 1:269. 1901; Piper in Contr. U.S.N. Herb. 11:216 (Fl. State Wash.). 1906; Ball in Coult. and Nels., New Man. Rocky Mt. Bot. 136. 1909; in Piper and Beattie, Fl. Northwest Coast 117. 1915;. Jepson, Fl. CaUf. 344. 1909, pro parte. — S. phlebophylla Watson in U.S. Geol. Sur. Expl. 40th parallel King's Rep. 5: Bot. 326. 1871,. pro parte, non And. — S. arctica var. petraea Bebb in Watson, Bot. Cahf. 2:90. 1879, pro parte, non And.; Ball in Trans. Acad. St. Louis 9:89. 1899, pro parte. — S. Brownii var. petraea Bebb in Bot. Gaz. 16:107. 1891, pro parte." — S. Brownii var. tenera Jones, The Willow Fam. 19. 1908. — This species has always been regarded as most closely related to S. petrophila, and Ball (1899) mentioned it in the synon^Tny only as "a narrow-leaved form," while he (1909) says "perhaps only a variety of the preceding" (5. petrophila). By a close study of the material before me I have the impression, however, that it possibly might have a more intimate affinity to S. phlebophylla. Both Ball and Rydberg distinguish S. casca- densis from S. petrophila only by the smaller size of the leaves and the few-flowered aments, and neither author mentions the fact that the old leaves are more or less persistent, a character not observed by me in S. petrophila. The leaves of S. cascadensis are occasionally up to 18 mm. long, and in male specimens, hke no. 1074 of Merrill and Wilcox from the Teton Mountains, Wyoming, the catkins bear 5^ BOTANICAL GAZETTE [january more than 20 flowers, and some fruiting aments before me measure up to 2.5:0.9 cm. The systematic position of the species needs further study, however, and it is interesting to note the singular view expressed by Andersson himself as to the relationship of his S. tenera: "Memorabilis forma a nostris 5. retusa, reticulata, et arhuscula quasi composita. S. retusae simiHs: foliis lingulatis, vix f poll, longis, supra medium 2 lin. latis, parallelo-nervosis, integerrimis ; S. reticulatae: capsuhs sessilibus, dense lanatis, • pusilUs; et S. arbusculae: amentis ramulos laterales terminalibus." The type of 5. cascadensis was collected by Lyall in i860 on the ^'eastern summits of Cascade Mountains" (f.; K.), and Lyall's specimens were sent out from Kew under the name S. phlebophylla; therefore Bebb (in Box. Gaz. 16:107. 1891) accused Andersson rather unjustly of the use of this name for Lyall's plant. There is no proof that Andersson himself apphed his own name phlebo- phylla to this form which he (1868) made the type of S. tenera. The writing on the labels of Lyall's specimens is not in Andersson's hand. The specimen collected by Watson in the Uinta Mountains, Utah (no. iioi, f.; G.), which the collector referred to S. phlebophylla because it seemed to match perfectly Lyall's plant, is of pecuhar interest; it looks to me more like a pubescent form of S. Dodgeana than anything else, but I do not venture at present to pass final judgment upon it. Sect. III. Ovalieoliae Rydberg in Bull. N.Y. Bot. Gard. 1:274. 1899, pro parte, sed sensu emend.— Sect. Arcticae Rydbg., I.e. 263, pro parte, non And.; Ball in Coult. and Nels., New Man. Rocky Mt. Bot. 135. 1909, pro parte.— Sect. Diplodictyae Schneider in Sargent, PI. Wils. 3:136- 1916.— See my first note (Box. Gaz. 66:117. 1918). To this section belong the following species I have dealt with {I.e.): 13. S. petrophila Rydbg., 14. S. anglorum Cham., 15. S. arctica Pall., 16. S. ovalifolia Trvt., 17. S. siolonifera Cov., and 18. 5. groenlandica Ldstr. I wish to add some notes to the following species. 15. S. ARCTICA Pall.— Since I wrote the first article I have seen specimens {Evans, no. 439, partim; W.) from Kodiak which represent Trauxvexxer's S. diplodictya mentioned I.e. 123. It is nothing but a mere form with "foliis utrinque concoloribus." igig] SCHNEIDER— AMERICAN WILLOWS 57 17. S. STOLONiFERA Covillc. — I mentioned {I.e. 137) a new f. subpilosa of which as type may be taken B. E. Fernow's specimen from Glacier Bay, Point Gustavus, June 10, 1899 (Cor.). There are specimens before me from the same bay, Muir Glacier, collected between June 8 and 12, 1899, by F. V. Coville and T. H. Kearney (no. 621, f.; W.). Coville is of the opinion that they, at least partly, represent a hybrid between 5. arctica and S. stolonifera. One of the sheets (no. 373447, W.) looks indeed Hke such a hybrid; the leaves seem not to possess stomata in the upper epidermis and are partly larger, while the flowering aments measure up to 3 : i cm. The other specimen (of sheet no. 376920), however, can hardly be separated from 5. stolonifera except by the tomentose ovaries. The leaves have numerous stomata in the upper epidermis, are of the usual size and shape, and the aments are not larger than in typical stolonifera. This specimen looks to me more like f. subpilosa than like a hybrid. It may be, however, that we have to take all the specimens with more or less pubescent ovaries and fruits for those of hybrid origin, but this does not seem hkely to me. 18. S. ARCTOPHiLA Cock.— Profcssor T. D. A. Cockerell kindly informed me that he had published in the last edition of Heller's Cat. N. Am. PI. p. 89, 1910, this new name for S. groen- landica Lundstrom but not of Heer (Fl. Foss. Arct. 1:101. pi. 4. figs. 8-10. 1868). I was not aware of this fact, neither had I seen the sheets of this edition of Heller's Catalogue. Besides, Cocker- ell's name is not found in the card index published by the Gray Herbarium. I accept this new name, but there is no definite statement in the international rules as to the priority of names of paleontological plants. Owing to this change I have to propose the following new combination, S. arctophila var. lejocarpa for S. groenlandica var. lejocarpa Lange {I.e. 141). 1 8a. S. hudsonensis, spec. nov. — S. fullertonensisXS. groen- landica Schneider in Bot. Gaz. 66:342. 1918. — Frutex prostratus habitu S. arctopJiilae, ramis pl.m. subterraneis radicantibus ad 5 mm. crassis, ramuHs elongatis repentibus glabris ut in S. arctophila coloratis, gemmis foliisque ut I.e. a me descriptis sed gemmis inter- dum ad 9 mm. longis et foliis angustioribus acutioribus ad 3 . 5 : i . 7 cm. latioribus ad 3 . 5 : 2 cm. magnis superne vulgo haud stomatiferis 58 BOTANICAL GAZETTE [january maturis glabris vel utrinque sparse pilosis; amentis tan turn fruc- tif eris visis 2 : i . 3 ad 4 : i cm. magnis ramulos f oliatos ad 4 cm. longos terminantibus ; fructibus ad 9 mm. longis pedicello brevissimo vel glandula vulgo distincte breviore excluso. Besides the specimens mentioned {I.e. 342), I examined the following: Hudson Bay: 50 miles south of Cape Eskimo, August 5, 1900, £. ^. and ^. £. Prg^/e (no. 43, fr., type, 46, St.; W.); 25 miles south of Cape Eskimo, August 12, 1900, same coll. (nos. 54, 57, fr.; W.). After having seen the material collected by the Prebles I think it best to propose a new name for this interesting form, which after all seems to represent a new species closely related to S. arctophila, from which it chiefly differs in the shorter pedicels and the more elongated gland. Judging by the flowers alone, one might be inclined to take it for a form of S. anglorum, but the leaves are mostly without any trace of stomata in the upper epidermis, and their color and texture are more like in S. arctophila. Some speci- mens cited in my second note may represent hybrids between this species and 5. Jullertonensis . Unfortunately I have not yet seen young female or male flowers, and further investigation is needed to elucidate the real affinity of this form, which seems to be fairly common along the western shores of Hudson Bay from James Bay to Cape FuUerton and also on the islands in the western part of Hudson Strait. Sect. IV. Glaucae E. Fries, Syllog. PI. Nov. 2:36. 1828, pro parte. — Sect. Arcticae Rydbg. in Bull. I.e. pro parte; Ball in Coult. and Nels, I.e., pro parte. For further synonymy see Schneider in Sargent /.c. 147. 1916. — In my second paper (Bot. Gaz. 66:318. 1918) I have already explained the differences between this section and the Arcticae, and there I have also discussed the following species, which are to be referred to this section: 19. S . fuUertonensis Schn., 20. S. niphoelada Rydbg., 21. 5. ehlorolepis Fern., 22. S. hrachy- earpa Nutt., 23. S. pseudolapponum v. Seem., 24. 5. desertorum Rich., 25. 5. glauea L., 26. S. eordifolia Pursh, 27. S. anamesa Schn., and 28. S. lingulata And. I wish to add the following remarks, because I had the opportu- nity to study some very interesting material of Herb. W., and I 1919] SCHNEIDER— AMERICAN WILLOWS 59 desire to express my gratitude to the curator of the U.S. Nat. Herbarium. 20. S. NiPHOCLADA Rydbg.— Having seen a co-type and other material of this species from Herb. W., I wish to correct my state- ments {I.e. 339) as follows: The co-type collected by Miss E. Taylor has ripe fruits which measure up to 7 mm. in length and show a dis- tinct pedicel (about 1.3 mm. long) which is distinctly longer than the bifid gland. The male type described by Coville {Funston, no. 185) has short aments measuring about 10: 4-5 mm. and loosely flowered at the base, but there is a specimen collected by F. C. Schrader on the John River in northern Alaska, July 10, 1901, of which the female part well agrees with the type of S. niphoclada, while the male aments measure up to 22 : 4 mm., being very slender and loosely flowered. In Funston's specimens the male flowers are younger, but I hardly believe that they could grow to the size of those of Schrader's plant. Otherwise the flowers are identical, having glabrous filaments and ventral glands of a similar shape. Schrader, however, collected another male specimen on Anak- tuvuk River (erroneously spelled Ansktoobah River on the label) August 5, 1 90 1, of which the aments are like those of the John River plant, but the filaments are somewhat hairy at base. Otherwise I cannot separate Schrader's plants, and I believe that the pubescence of the filaments which mostly can be taken for a constant character may not be of reliable taxonomic value in this case. We need, of course, a better acquaintance with all the willows of this region to decide the question whether or not the absence or presence of a pubescence on the filaments is a really important character. A fruiting specimen collected by Schrader at the same place and date as the last mentioned male one seems to me inseparable from 5. niphoclada, but the ripe fruits measure up to 8 mm. in length, are almost sessile, and of a more ovoid-conical shape than in the type. The specimen mentioned {I.e. 339) from Fort Churchill, Hudson Bay, collected by E. A. and A. E. Preble, is really no. 41, not no. 26 (as given by me and by Coville), according to a note by E. A. Preble on the sheet (no. 385093) in Herb. W. From the same place the same collectors brought a male specimen (no. 33; W.) which I refer to 6". braehyearpa, which had been collected there 6o BOTANICAL GAZETTE [january also by F. M. Macoun {I.e. 337). Preble's no. 41 may be more closely connected with S. brachycarpa than with S. niphoclada, but the shape of most of the larger leaves is more elongate-lanceolate in no. 41, where the largest leaves measure up to 4.5: i cm., while those of Macoun's female plant are shorter and broader, and its pubescence is more like in typical S. brachycarpa. In no. 41 the young twigs bear a very thick villose pubescence of long soft hairs, which is far more conspicuous than the tomentum of the forms of S. niphoclada from the west. After all, I am inclined to believe that no. 41 may represent a form of S. hrachycarpacovcv^^kX^hle. to f . poliophylla of 5. glauca acntifolia. We must not forget, however, that we hardly can elucidate those forms without a full under- standing of the true S. desertorum (see I.e. 331). I have thought it best to treat those forms at a considerable length, because we know so little of the willows of the Northwest Territories, and I wish to give an impulse to their closer investiga- tion in the field by future collectors. 21. S. CHLOROLEPis Fern. — I have described {I.e. 339) the var. antimima, which looks rather intermediate between the type and S. brachycarpa, and which on Mt. Albert seems to be connected by hybrids with the latter. There is a small male specimen in Herb. N. collected hy A. P. Low, north of Cape Jones on the eastern coast of Hudson Bay, July 16, 1898 (no. 63272 O.). It agrees well with S. brachycarpa, and has hairy filaments, but I found some stomata in the upper leaf epidermis. We need much more copious material (male and female) from this locality to judge the form properly. 25. S. GLAUCA var. acutifolia (Hook.) Schn.'^ — Having re- ceived, as already stated, very interesting material from Herb. W., I wish to add the following remarks.'^ There are before me several " I stated {I.e. 321) that 2 S. villosa had been published before Hooker described the present form under this name, but there is still a much older S. villosa Hoffmann (Obs. Bot. 15. 1787) which is not registered in the Kew Index, nor can I find this name mentioned by Koch, Fries, Wimmer, Andersson, or v. Seemen. It was sent from Sweden by Thunberg. " In the note {I.e. 321) I made an entirely wrong statement in regard to Pursh's Canadian collections. Reljdng on facts given by Harshberger, which he in his turn took from an article in Bot. Gaz. 7:142. 1882, I said that Pursh did not collect in Canada. Professor N. L. Fernald kindly directed my attention to Penhallow's iQig] SCHNEIDER— AMERICAN WILLOWS 6l' specimens collected at Great Slave Lake, Mackenzie, which, in my opinion, are most closely related to var. aciUifolia, but at least some of them seem to represent a very villose form of it, for which I propose the name: S. glauca var. acutifolia f. poliophylla,''' forma nov. — A typo nonnisi differre videtur ramis annotinis densius villoso-lanuginosis etiam vetustioribus tomento lanuginoso pl.m. obtectis, foliis superne pl.m. laxe adpresse sericeo-lanuginosis subtus villo densissimo moUi pl.m. adpresso albo vestitis. The type is Great Slave Lake, Fort Rae, July 28, 1901, E. A. and A. E. Preble (no. 139, fr.; W.; folia inferiora elliptica vel oblongo-elKptica, utrinque acuta, superiora magis ovato-elliptica apice acutiora, maxima ad 5:2.2 cm. magna; amenta fructifera pedunculo foHato ad 3 cm. longo excluso ad 5:1.3 cm. magna; fructus e basi rhom- boideo ovoideo-conici, ad 9 mm. longi pedicello ad i . 5 mm. longo glandulam subduplo superante excluso). The following specimens seem to me rather intermediate between f. polio phylla and typical var. acutifolia: Great Slave Lake: Fort Resolution, July 14, 1901, E. A. and A. E. Preble (no. 141, fr.; W.; forma minus dense quam no. 139 villosa); June 21, 1903, E. A. Preble (no. 194, f. ; W. ; eadem forma ut videtur ac praecedens sed juvenilis) ; Fort Good Hope, on the Mackenzie River, June 23, 1904, E. A. Preble (nos. 330, f. et fr. anni praeteriti, 332, f. ; W. ; forma a cl. Rydberg ad S. niphocladam relata foliis floribus juvenilibus; sed folia distincte petiolata et fructus adulti magni). There are two other male specimens of E. A. Preble from For-t Resolution, June "Review of Canadian Botany" (Trans. Roy. Soc. Can. III. 4:4. 1897), where he expressly states that "with the exception of his immediate predecessors no botanist had accomplished more than Pursh to make the vegetation of Canada known." When James published the manuscript of Pursh's traveling journal he did "not deal with that part of Pursh's work which was continued into and ended in Canada." "He made extensive collections chiefly through the province of Quebec, but all the material thus accumulated was subsequently destroyed by fire." Regarding Lord Selkirk, I have been informed by J. C. Nelson that "Lord Sel- kirk's Exp." probably refers to Thomas Dundas, Fifth Earl of Selkirk, who (according to Johnson's N. Univ. Cycl. 4:175. 1878) "spent several of the later years of his life in promoting emigration to the Red River of the North, British America." I have not yet been able to consult the tracts which he has published on emigration to those parts of Canada. '"t Derived from iro\i6s, with white hair. 62 BOTANICAL GAZETTE [january 21, 1903 (nos. 196, 198; W.), which Rydberg regards as "near" niphodada but with much broader and shorter leaves." The young leaves do not differ from those of the female specimen no. 194 previously cited, but the young male aments are very short, not exceeding i cm. in length and o . 5 cm. in thickness. Otherwise the flowers agree with those of var. acutifolia. If the size of the male aments should prove a reliable character, and if the male plant should belong to the same form as the female no. 194, this form might prove to be a new one more closely related to S. cordifolia. See also my remarks under S. niphodada as to the size of male aments. There is a male specimen collected by Seton and Pringle, July 19, 1907, near Caribou Island, Great Slave Lake (no. 43 [78305 O.]), which has been named S. atra Rydbg. by Ball. At first sight it resembles a great deal Preble's nos. 196 and 198 mentioned above, but the leaves, which seem to be almost fully grown, measure only up to 3.6:1.6 cm. The flowers of the one small catkin (2.3:0.9 cm.) I could examine show no trace of a dorsal gland, and the hairy filaments are connected for ^ to | of their length. So far this form remains rather doubtful. 26. S. CORDIFOLIA Pursh. — The male syntype of S. lahradorica mentioned {I.e. 345) bears the number 21, not 31, of Waghorne. 27. S. ANAMESA Schn.— In his Fl. Europ. 21:157. 1890, Gan- DOGER has described under S. glauca the following 4 new subspecies from Greenland: 5. eskimorum (type: Peter^ew, JuHanehaab) ; S. groenlandica, non Heer, nee Lundstrom (type: Rink, Godthaab); 5. platycarpa (type: /. Vahl, Ikilok), and {I.e. 158) S. Vahlii (type: /. Vahl, Ikilok). Judging by the characters given in the key, all these names refer to a form having "foHa anguste hneari- oblonga, 5-12 mm. lata," which are "supra semper canescentia et villosa vel tomentosa" in the last two forms, while they are "supra glabra aut glabrescentia et viridia" in the first two. Not having seen the types, I am unable to say to which species these forms really belong, but I presume they may be referable to what I have called S. anamesa. In giving binary names to those forms Gandoger did the same as Andersson did in several instances in his monograph. I beheve, however, that these binomials cannot be regarded as iqiq] SCHNEIDER— AMERICAN WILLOWS 63 proper species, and therefore I do not think it advisable to take up one of Gandoger's names for 5. anamesa. V. Species sectionis incertae. The proper systematic posi- tion of the following species is still doubtful to me, and most of them are so little known that it is impossible to even express an idea as to their affinities. 29. S. Chamissonis And. in DC, Prodr. 16^:290. 1S68; Lund- strom apud Kjellman in Nordenskiold, Vega Exp. Arbet. 2:21 (Fanerogfi. St. Lawrence On). 1883; Coville in Proc. Wash. Acad. Sci. 3*325.^^. 2 J. 1 901. — S. myrsinites Chamisso in Linnaea 6:540. 183 1, non L. — The type of this rare but apparently well marked species was collected by Chamisso and Eschscholtz in 1816 at St. Lawrence Bay; elsewhere it is only known from St, Lawrence Island and Port Clarence in Alaska.'' 30. S. GLACiALis And. in Ofv. K. Vet.-Akad. Forh. 15:131. 1858; in DC, Prodr. 16^:300. 1868; Coville in Proc. Wash. Acad. Sci. 3:329.^^. 24. 1901; Rydberg in Bull. N.Y. Bot. Card. 1:262. 1899. — 5. Uva-ursi Seemann, Bot. Voy. Herald 40. 1852, pro parte. — ^As Coville has already stated, "this species is known only from the type specimen collected by Lieutenant W. J. G. Pullen in 1849, on the Arctic seacoast between Point Barrow and the Mackenzie River, and from specimens collected at Point Barrow by John Murdoch in 1882-3." I have been able to examine a photograph and fragments of the type {Pullen, no. 155, f., fr. im.; K), and I agree with Andersson that it is a forma pusilla of S. ovalifolia. The style is very short but not wanting, and the stigmas are rela- tively long; the pedicel is also distinct and in fruit about \ longer than the gland. It resembles the pubescent form of 5*. ovalifolia, which is only somewhat more vigorous, and shows a more distinct (but short) style. Andersson describes the leaves of the type as '5 After the manuscript of this note had gone to the printer, I found on a sheet with S. amplifolia of Herb. Cor. a male specimen of S. Chamissonis collected by B.E. on St. Lawrence Island, July 13, 1899, which had been seen by Coville but is not enumerated by him in 1901. Had I seen it before I finished my manuscript I would not have included this species in the present key because, judging by the male flowers, it seems to belong to sect. Commutatae, with which I hope to deal later. The male flowers possess only one gland, and the fine and close acute serration of the leaves easily distinguishes this species from all the other willows treated in this paper. 64 BOTANICAL GAZETTE [january ''integerrima," while most of them in the specimen before me show a distinct but fine glandular denticulation in the lower half. Those of the type are entire, with the exception of a few minute teeth at the base of some leaves. They lack stomata in the upper epidermis. If I had seen the type before I finished the manuscript of this article, I would have placed S. glacialis next to 5. ovalijolia, but as long as the male flowers are unknown the true affinity remains unknown. 31. S. VENUSTA And. in DC, Prodr. 16^:288. 1868, non Host 1828, from Sitka, of which, according to Coville, the description ''suggests that the plant may prove to be a form of Salix reticulata grown in a shaded situation," can only be judged by an examination of the type material, which I have not yet had the opportunity to see. 32. S. OBCORDATA And. in DC, Prodr. 16^:291. 1868, which, like the preceding, came from Sitka, is even more obscure than that, and may prove to be another S. reticulata form or a hybrid of it with some other species. Andersson himself placed it without number as S. ohcordata between S. ovalifolia and S. furcata, but he says nothing about its relationship to these species, only mentioning its resemblance to S. venusta and 5. reticulata in his more than meager description. I hope that my notes may prove useful to other students of American willows, and I shall be most thankful to anyone who can correct any mistake I have made or furnish me with good material or information of these or other American willows. There are, of course, quite a number of specimens before me which I have not yet been able to elucidate. Among them are many which I suspect of being of hybrid origin, but I do not intend to deal with the hybrids until I have gained a more thorough understanding of all the American forms of this difficult genus. Arnold Arboretum Jamaica Plaesj, Mass. MORPHOLOGY OF THE GENUS ACTINOMYCES. P Charles Drechsler While the genus Actinomyces has received a large measure of attention in its relations to soil biology and to human, animal, and plant pathology, the natural affinities of the congeneric organisms that it has been customary to include in the group have been the subject of diverse opinions. Under a variety of synonyms, among which Cladothnx, Nocardia, Discomyces, and Oospora have been used nearly as frequently as Actinomyces, the group has been placed with the bacteria, with the Hyphomycetes, or assigned to an intermediate position. In the earher publications on the ray fungus, including the papers by Bostroem (i)^ and by Wolff and Israel (24), this organism was referred to the pleomorphic bacteria. The belief was seriously entertained that cocci, bacteria, and spirilla were produced by the plant, and in such regular succession that a number of investigators were led to draw up detailed ontogenetic schemes of considerable complexity. It is frequently not easy to determine the exact nature of the structures that were interpreted as pleomorphic stages. There are plenty of indications that con- taminating bacteria were often present as secondary invaders; but more frequently aerial spores, segments of spiral or sinuous hyphae, and degenerative bodies of metachromatic substance were mistaken for schizomycetous t>'pes of nearly every description. More recently Actinomyces has been frequently associated with the tubercle and diphtheria organisms on the assumption that they may represent a transition between the Hyphomycetes and the true Schizomycetes. A family of Actinomycetes has thus been erected as a natural group from these diverse components, united chiefly by resemblances in their staining reactions, a usual or an occasional filamentous habit, and the development' of clavate elements in the animal body. It has been supposed by adherents of such a taxonomic disposition that either a progressive phylogeny ' Contribution from the Cryptogamic Laboratories of Harvard University, no. 83. » The bibliography will appear at the end of Part II, which will be printed in February. ^S] [Botanical Gazette, vol. 67 66 BOTANICAL GAZETTE [january has occurred in the family, with Actinomyces at the head of a transition series developing increasingly well marked fungoid characteristics, or else that Actinomyces is the probable progeni- tor of the groups Corynebacterium and Mycobacterium by degenera- tive reduction. The view that the ray fungus represents an organism with hyphomycetous affinities was advanced early by Harz (7) and DeBary (2). These authors regarded as conidia the clavate ele- ments of the actinomycotic lesion, which Bostroem's studies later properly degraded to the rank of degenerative structures. Sauva- GEAU and Radais (20), Domec (3), Thaxter (22), Gasperini (4), and others placed a number of congeneric forms among the Hypho- mycetes on account of their production of aerial spores. It may be mentioned in this connection that an examination of a consider- able number of species has convinced the writer that this disposition is the only one which is in harmony with the morphological condi- tions represented in the genus. The material used in these studies, with the exception of authentic cultures of the species described by Waksman and Curtis (23), and of a number of organisms isolated by H. J. Conn from soil collected near Geneva, New York, was largely obtained from soil collected in Cambridge, Massachusetts. By the use of the dilution method more than 1000 plants belonging to the genus were isolated from this source; and of these about 300, representing probably more than 100 species, were selected for morphological examination. Approximately 400 additional individuals were derived from soil collected in Porto Rico, Cuba, Panama, Montana, Wisconsin, and Kansas, as well as from outdoor air, tap water, horse manure, and gross cultures of dung, dead leaves, and other vegetable matter. The potato scab organism was obtained from Mr. M. Shapavalov, who had isolated it from a diseased tuber, and experimentally estabhshed its pathogenicity. The morphology of the vegetative thallus of Actinomyces, apart from its astonishing minuteness, the diameter of the filaments ranging commonly from o . 5 to i . 2 m, presents no features unusual among the fungi. In most species the mycelium is generally sparsely and irregularly septate; and although in other forms trans- igig] DRECHSLER— ACTINOMYCES 67 verse walls may appear with somewhat greater frequency, there are none in which septation approaches any pronounced degree of regularity or closeness. Ramifications are abundant, and the branching is altogether of the ''true" type. Mace (15), who first carefully observed the formation of branches, found it to proceed by the elongation of lateral buds arising some distance back from the growing point of an axial filament, the branch thus produced giving rise similarly to secondary branches by lateral proliferation. Lachner-Sandoval (12) confirmed Mace, designating the process as monopodial and denying the existence of true dichotomy in the genus, which had been affirmed by previous investigators. Later Neukirch (18) reported that the branching in Actinomyces ochro- leuceus was occasionally of the nature of a true dichotomy. From an examination of very young mycelium (fig. 3)^ it is apparent that, at least in stages following the germination of the spore, filaments are not infrequently terminated by two elements too closely similar in size and angular relationships to be distinguished as bud and axial tip. The branching in such cases must be regarded as dichotomous, although all gradations toward the prevailing well defined monopody may be found. It seems reasonable to suppose, however, that the distinction is one of convenience, not implying any fundamental difference in, manner of development. The cytological structure of Actinomyces is equally devoid of bacterial characteristics. The branches forming the periphery of the actively growing pellicle, or the young sporogenous branches attached at intervals to the superficial mycelium, are iilled with dense protoplasm, which, with haematoxylin, takes a deep homo- geneous stain. Further toward the origin of the hyphae the con- tents become more attenuated, and vacuoles appear, increasing in number and size until they occupy the larger portion of the fila- ments. When individual vacuoles become excessively large and extend through a considerable length of filament, the cytoplasm is in a large measure confined to a peripheral layer, a condition which led Neukirch to distinguish a thin, strongly refringent "Aussenplasma" and a less refringent "Innenplasma." 3 The plates will appear in connection with the second part, to be published in the following number. 68 BOTANICAL GAZETTE [january The presence of large vacuoles is commonly associated with local distentions of the filament wall. These may occur with such regularity in the degenerate mycelium of some species as to suggest the appearance of Leptomitus, each swollen segment being largely occupied by a single elliptical vacuole, separated from the vacuole in the adjacent distention by a protoplasmic partition at the con- striction (figs. 47, 48, 106). In other species and quite generally in the nutritive mycelium of all forms, that is, the mycelium im- mersed in the substratum, no marked regularity in the alternation of inflated portions and constrictions is observable; but pro- nounced deviations in the diameter of filaments may occur with more or less variable frequency. A great deal of importance has been attributed by some writers to a variety of abnormalities "and products of degenerative changes occurring in the thallus of Actinomyces. In the earlier literature on the ray fungus, especially in the publications of Israel (8), JOHNE (9), and MacFadyean (16), bodies described as "micro- cocci," "cocci," or "coccus-like granules" were given minute attention, and assigned an important role in the complex ontogeny ascribed to the parasite, then supposed to belong to the pleomorphic bacteria. Wolff and Israel (24), whose photomicrographs of these bodies leave no reasonable doubt about their identity with structures very frequently observed by the writer (figs. 15, 31, 32, 42, 91), confused them with the spores reported by other authors; and as the structures did not possess the heat resistance common to spores of bacteria, these investigators were inclined to question the production of spores by Actinomyces. Since the organism used by Wolff and Israel was constantly sterile, their conclusion concerning it was undoubtedly correct. Bostroem, who experi- mented with a sporiferous form, did not succeed in avoiding the same confusion, and refers indiscriminately to the unicellular products formed from aerial hyphae, and to the spherical endoge- nous granules, as "spores." Round granules, deeply stained in the living filament by very dilute methylene blue, were studied later by Neukirch. He noted in them a variable size, a method of multiplication, and an orientation related to the regions of growth in the thallus. These 1919] DRECHSLER— ACTINOMYCES 69 observations led him to believe that the structures represented nuclei. Schutze (21), who applied Neukirch's method of stain- ing, designated the bodies as metachromatic granules. After an examination of their occurrence in the aerial mycelium of a con- siderable number of species, such an interpretation seems, in the opinion of the writer, to offer the greater degree of plausibility. The metachromatic material is easily distinguished by a power- ful affinity for most of the stains ordinarily employed in laboratories. In material fixed in alcohol, and treated with Delafield's hematoxy- lin, it retains a nearly opaque stain after all other structures have been completely decolorized. Indications of its presence in the tips of growing filaments, or in sporogenous branches, in general are very infrequent. Some distance toward the origin of the hyphae, associated with a more attenuated or vacuolated proto- plasm, the material makes its appearance in the form of rather minute granules widely separated from one another. As the fila- ment is followed still farther back, the granules increase in size and frequency; often their arrangement is one of much regularity, the individual spherical bodies being of nearly equal size, exactly filling the lumen of the filament, and separated by nearly equal spaces (fig. 42). In other cases the granules seem to coalesce and occupy entire segments of hyphae (fig. 32); and in a few species extensive portions of mycelium were frequently found entirely filled with long unbroken masses of metachromatic substance. It is this property of coalescence of smaller granules, to form incompar- ably larger masses, bearing out the similarity in appearance to a homogeneous liquid with a relatively high surface tension, that makes it diflicult to believe that we are dealing here with anything relating to spores or to nuclei. The function of the metachromatic material in the Actinomyces thallus cannot be ascertained with certainty. A number of views have been advanced regarding the role of metachromatic substance in the cell, none of which has gained universal acceptance. The best explanation, in the opinion of the writer, seems to be that it represents an occluded waste product. While its presence in small or moderate quantities in the sterile hyphae bearing the sporoge- nous branchlets is probably more or less normal, its abundant 70 ' BOTANICAL GAZETTE [january occurrence here, as everywhere else, is an indication of advanced degeneration. In the more mature mycelium of Actinomyces VIII (fig. 47) metachromatic granules are usually very conspicuous, often occupying most of the space in the narrowed constrictions between the large vacuoles in the highly inflated mycelial segments. The appearance of such a thallus is not in the least suggestive of the structure of bacteria, and indicates that the resemblance between Actinomyces and the true Schizomycetes in the consistency of proto- plasm, emphasized by some writers as an important phylogenetic connection, has been unduly overestimated. While the sterile filaments in the nutrient and in the aerial mycelium are relatively uniform in structure throughout the genus, the sporogenous apparatus of many species exhibits a large degree of morphological individuality. This diversity has not usually been recognized by writers, and has undoubtedly been responsible for a portion of the controversy that has arisen, particularly with regard to the method of spore formation. Lachner-Sandoval (12) , from a study of Actinomyces albido-flava, distinguished two kinds of propagative bodies: (i) fragmentation spores appearing as spheri- cal to cylindrical segments in old hyphae, formed by a contraction of the protoplasm; and (2) segmentation spores developed by a septation of the tips of aerial filaments. Segmentation was usually found to involve only lateral branches coming from aerial hyphae, but in submerged growths it frequently extended also to the main filament, leading to the development of a dendroidic system of spore chains. Lachner-Sandoval's figures of these formations, how- ever, are much less striking than might be expected from the description in the text, and do not convey the impression of rami- fications approaching treelike proportions. Neukirch identified the segmentation of Lachner-Sandoval with oidium-spore formation among the fungi, and abandoned the use of a specific term. He vigorously disputed the development of aerial spores by a septation of the mycelium. According to his account the spores are formed as the result of successive contrac- tions of protoplasm until approximately isodiametric portions are separated by regularly alternating empty spaces. This process he identified with the fragmentation of Mace, locahzing it in a differ- iqiq] DRECHSLER— ACTINOMYCES 71 ent region from Lachner-Sandoval, and properly relegating the fragmentation of the latter to the category of degenerative changes. Gilbert (5) found some lateral branches to begin the process of forming spores by becoming differentiated into highly refrac- tive and weakly refractive portions. Constrictions later appear, unassociated with visible changes inside the filament, and soon the spores are completely cut off. Gilbert designated the process as segmentation, following Lachner-Sandoval, who, however, had actually observed septa appearing more or less simultaneously with the constrictions, their appearance being followed by the enlarge- ment and rounding up of the segments to form spores. MiEHE (17), in his study of Actinomyces thermophilus, only inci- dentally examined the mode of sporulation. He believed spores were produced singly on very short stalks attached laterally to the main hyphae, or possibly by successive contractions in chains. In either case conidia were produced, not by the segmentation of a completed fiJament, but by the development of a structure which at no time constituted a cylindrical, continuous hypha. This account, in general, bears strong resemblances to the later descrip- tion by ScHtJTZE (21) of Actinomyces monosporus, a form in which the spores are borne singly on delicate stalks in racemose arrange- ment on a thicker axial filament. It might well be questioned, however, whether forms like this, which depart so widely from the main morphological trend of Actinomyces, are properly to be assigned to this genus, even if allowance is made for much Hberality in the definition of hyphomycetous form-genera. The same criticism, however, cannot be extended to the condi- tion described by Schijtze in his strain of Actinomyces thermophilus. In his account of this species its author strongly defended Neu- kirch's position that the mode of sporulation was one of fragmen- tation. However, while Neukirch found long filaments converted into spore chains by successions of protoplasmic contractions, the long portions finally becoming resolved into ultimate spores, ScHtJTZE found that only short terminal portions or short lateral branches yielding about 5 spores were involved. According to Neukirch, the slightly refractive spaces between the masses of protoplasm that later develop into spores are entirely empty, and 72 BOTANICAL GAZETTE [January the intervening portions of filament wall merely collapse as the spores mature. ScHtJTZE believed that these intervals were filled with attenuated protoplasm, and that by their constriction the spores were delimited without the evacuation of portions of hyphal wall. The spore of Neukirch, consequently, is a structure possess- ing its own spore wall, enveloped, except at its ends, by the rem- nants of the old filament wall; that of Schutze, on the other hand, is without a separate spore wall, the filament wall constituting the only membrane present, and forming a spherical shell everywhere inclosing the protoplasm. Neukirch gave much attention to certain structures he desig- nated as oidium-spores. They developed in submerged growths, the transformation of the filament consisting only in more or less close septation, followed by a slight swelling of the segments. Under suitable conditions filaments grew out from them, an occurrence Neukirch regarded as germination. " Aussenplasma " and "Innenplasma," in his opinion, were sharply defined, but a spore wall was absent. The elements did not exceed the filaments in resistance either to heat or desiccation. Neukirch believed their function to be the dissemination of the fungus in liquid media. Lachner-Sandoval seems to have seen the same structures and regarded them as segmentation spores that had developed in the submerged condition. Gilbert, Schijtze, and Krainsky (id) record their failure to find these bodies without, however, denying their existence. According to Schijtze and Gasperini, sporulation may occur in hyphae which are not truly air hyphae. It seems questionable whether any desirable end is served by calHng Neukirch's elements spores at all. To apply the term to structures with so little individuahty, even though a sort of promiscuous viability may be attributed to them, is approaching very close to the point where all bodies not filaments of uniform thickness are to be regarded as spores. Certainly the distended elements in old mycelium of Actinomyces VIII (figs. 47, 48), which represent enlargements of axial filaments developed gradually in the course of time at the junctures with moderately complex systems of sporogenous hyphae, frequently have an equal or greater resem- blance to reproductive bodies; and the behavior, under similar 1919] DRECHSLER— ACTINOMYCES 73 conditions, of forms like the smuts, Mucor, and Penicillium, would make advisable a larger measure of caution in dealing with fungi growing irregularly in a submerged condition. Although various details associated with the sporulation of Actinomyces have thus been dealt with in the literature, the opinion still seems to prevail widely that the process is of an irregular and miscellaneous nature. Litman and Cunningham (14) in recent years have denied the character of spores to the "gonidia" pro- duced by the potato-scab organism; the elements are believed simply to "serve as a segment of the mycelium, which, by increasing the number of segments, may increase the chances for spread and continuous existence." This view seems, in the opinion of the writer, very much at variance with the distinctiveness of the well characterized sporogenous apparatus found in Actinomyces. In pursuing the present studies a method of mounting material was employed which, in view of the exceptional fragility of all species of Actinomyces, and the great difficulty ordinarily en- countered in attempting to stain undisturbed sporulating condi- tions, gave exceptionally good results. The plants were grown on a suitable substratum, usually potato or glucose agar. Growth on potato agar, as a general rule, is more. prompt and productive of mycelium; but as its use, especially with species exerting a strong tyrosinase reaction, stimulates to excessive guttation and dis- ruption of the sporophores by the extruded droplets, a medium not possessing this property is often found to be advantageous. After the cultures had attained a proper degree of maturity, the whole growth was cut from the agar and removed from the tube as care- fully as possible. A slide smeared with albumen fixative was now brought into firm contact with the mycelium and then separated from it, precautions being taken to avoid altogether any sliding of the two surfaces on each other. If the growth is not too young, this procedure will leave the upper portions of the aerial mycelium adhering to the shde without serious disarrangement, and killing and fixation may be at once effected by the use, for example, of strong alcohol. The material was subsequently stained and mounted in balsam. The quality of preparations in which the spore chains have commenced to disintegrate in large numbers is 74 BOTANICAL GAZETTE ^ [January impaired by the presence of large masses of free spores, which retain their staining properties for some time after maturing. Later the spore walls seem to become entirely impervious to stains, and as a result when the secondary mycelium develops beyond a slight clouding effect no difficulty is encountered from this source. The best results are obtained when the print is made soon after the mycelium begins to adhere readily to the smeared slide. The nature of the killing agent employed was found to have no noticeable effect on the preparation. Flemming's mixtures, both weaker and stronger, picro-formol, picro-acetic, Carnoy's fluid, and 95 per cent alcohol were tried with apparently the same results. Owing to the small diameter of the filaments, the pene- tration is probably so nearly instantaneous that plasmolysis is effectively prevented by nearly any toxic agent capable of readily wetting the material. In order to obviate the necessity of washing, strong alcohol was used almost exclusively. Much more depends on the proper choice of a stain. Saffranin, gentian violet, Bismarck brown, and eosin usually fail to bring about a sufficiently deep coloration. Carbol-fuchsin acts powerfully and rapidly, but is poor for purposes of differentiation. Haiden- hain's iron-alum haematoxylin is good for protoplasmic structures. The most satisfactory results were obtained with Delafield's haematoxylin, which if allowed to act for 24 hours, with the proper degree of decolorization, yields deeply stained, clear preparations, showing vacuoles, metachromatic, and nuclear structures, as well as septa, with remarkable distinctness. The spores of all species of Actinomyces are developed by a transformation of more or less specialized hyphal branches dis- tinguishable from the sterile hyphae of the aerial mycelium at an early stage in their development. In general, with the exception of such inflated hyphae as are shown in figs. 47, 48, and 106, the diameter of any portion of sterile mycelium is attained at the time it arises through the elongation of the growing filament tip, subse- quent increase in thickness being very slight. The sporogenous branches, however, are in the beginning conspicuously thinner than the axial hyphae from which they are derived. Later, when their final linear extension has been nearly attained, increase in thickness 1919] DRECHSLER— ACTINOMYCES 75 generally follows. This increase may be slight, as in some species in which the mature sporogenous hyphae are still somewhat thinner than the vegetative hyphae (fig. 46), or more considerable, as in forms in which they conspicuously exceed the latter in thick- ness (figs. 4-6). The very simple type represented by Actinomyces XIII, in which the aerial mycelium is represented by very long filaments, rarely branching and apparently sporogenous almost to their point of origin in the nutritive mycelium, constitutes the only exception, since in this instance there is no indication of thickening in the young fertile hyphae, nor indeed any variation in the diameter of its vegetative filaments. In a majority of the species the maturation of the sporogenous h>phae is associated with a peculiarity in growth by which they become coiled in more or less characteristic spirals. The tendency toward the coiled condition is usually clearly manifested before the branch has grown to half its final length through the open flexuous habit of the young filament (figs. 5^2, 107). As elongation con- tinues, the turns become increasingly definite, but the contraction leading to the fijial condition, which ranges from that illustrated by Actinomyces XIII with its open, barely perceptible turns, to one in which the spirals are so strongly compressed that its adjacent turns are in continuous contact (figs. 44, 51, 57) in a fashion resem- bling that of the spores of the h>'phomycetous genus Helicobn, is usually delayed until the later growth in thickness of the filament. Specific differences may not only be indicated by the obliquity of the spiral, but involve also the number and diameter of its turns, and its construction with reference to the dextrorse or sinistrorse condition. The range in different species extends from the 2 or 3 turns exhibited in forms like Actinomyces II and XVI, to over 20 turns in others; but the range in a particular species is always considerably smaller. The writer once observed a spiral with 24 turns, but this probably approximates the extreme maximum; spirals with 14-16 turns (figs. 23, 57, 94c) are by no means abundant, and probably no species produces many in which there are more than 12 turns. The diameter of the spirals as a whole is more or less in inverse ratio to the number of turns characteristic of the species. This 76 BOTANICAL GAZETTE [january correlation is very evident in a comparison of types likt Actinomyces II and XVI showing spiral sporogenous hyphae with a few wide turns, and types hke IV, V, and XVII illustrating spirals of many narrow turns. Rotation in the formation of spirals is specifically sinistrorse or dextrorse in different species of the genus; and it is interesting to note that here, as in the vegetable world in general, sinistrorse are much more abundant than dextrorse species. Of the 1 7 species with spiral sporogenous branches figured in the present paper, which have been selected as representative of a much larger number, 5 are dextrorse, 11 are sinistrorse, while the condition of the remaining one could not be ascertained with certainty. In general, the pro- portion appears throughout the entire genus. As a morphological feature, the absolute constancy with which a species adheres to one kind of rotation is noteworthy, particularly in view of the extremely minute dimensions of the structures concerned. An examination and comparison of the relation of the sporoge- nous branches to each other, and to the axial filaments, enables one to recognize several tendencies, the distinguishing characteristics of which are correlated with differences in the sequence of prolifera- tion. Two main types may thus be recognized, approaching each other in apparently intermediate forms, but moderately distinct at the extremes: (i) an erect dendroidic type in which the sequence of development of the sporogenous hyphae is successive; and (2) a prostrate, racemose type in which the development is more nearly simultaneous. In the erect type, well exemplified in Actinomyces I, the develop- ment of the fructification starts from a single erect hypha with a spiral termination. Sporogenesis commences at the tip by the insertion of regularly spaced septa, and proceeds downward toward the base of the filament. Usually before much of the hypha has been involved, a single septum will appear well toward its base, and immediately below it the bud anlage of a new sporogenous branch appears. As the latter is attaining its growth in length and thickness, and its spiral disposition, the basipetal septation in the axial filament proceeds to the septum above the insertion of this first branch, the young spores thus delimited undergo maturation 1 91 9] DRECHSLER—ACTINOMYCES 77 processes, the spiral becomes relaxed, and the chain of spores subject to disruption. The branch now passes through the same course of development as the axial filament and in turn gives rise to a sporogenous branch below a septum a little above its own insertion. The number of sporogenous branches developed below a single septum is generally increased to several by proliferations subsequent to the first; and as the initiation and development of successive orders may be indefinitely repeated, complex fructifica- tions are frequently developed, in which a succession of the processes described are simultaneously taking place at many points. In the second type there is no such clearly defined relation between younger and more mature sporogenous h^^phae. Develop- ment of a fructification is initiated by the proliferation of branches at irregular intervals on the distal portion of a prostrate axial filament which often exceeds i mm. in length. The branches may either stop their more extensive development after forming a spiral, or themselves proliferate a secondary branch a short distance above their own insertion; and this in turn may form a spiral and give rise to a tertiary branch (fig. 43). By a repetition of this process each lateral element may become branched several times, the whole apparatus as well as its insertion' on the axial filament being characterized by an absence of septa. Sporulation, instead of beginning in any individual spiral as soon as it is formed, is usually delayed until the branching and growth of spiral hyphae in the same lateral process have come to an end (figs. 42-44, 46), when it will often proceed rapidly and almost simultaneously in all the spirals (fig. 41). The termination of the axial filament itself develops into a spiral, and behaves essentially like a primary lateral branch. Occasionally the axis of one of these racemose arrangements may be comparatively short, resulting in a rather intricate structure in which the spirals of one lateral branch may be entangled with those of another (fig. 44). The tendencies characteristic of the type, however, are maintained : the absence of a septum above the insertion of branches, and the delay in sporulation in the spirals first formed, until the growth of the last order of sporogenous branches is more or less complete. 78 BOTANICAL GAZETTE [january In addition, however, to species in which these two types are clearly distinguished, a still larger number of species present a combination of the two features. Frequently the open racemose arrangement of the lateral branches on the main axial filament is associated with a successive order of development in the further ramifications of the branches (figs. 63, 79). The presence of a septum above the insertion of a branch is characteristic of more species than is its absence (figs. 2, 5c); and in some species both conditions prevail (figs. 53, 79, 81). In other forms a fructifica- tion with successive development may terminate a long prostrate filament. In a few species, particularly Actinomyces X and XVIII, there are formed, in addition to the more regular fructifications, others of a more miscellaneous tendency. The branching axial filaments are relatively thick, densely filled with protoplasm, and bear at very close and irregular intervals a short, thick, unbranched spo- rogenous hypha with Httle or no spiral modification (fig. 103). It seems quite probable that this type of development is associated with the excessively rapid growth that characterizes the two forms in which it was most frequently observed. The degree of completeness to which the aerial mycelium of Actinomyces is converted into spores has generally been overesti- mated. On the contrary, sporulation is quite strictly confined to terminal elements, never as a rule passing beyond the first junction with another element. The proliferation of the branch nearest the end of the axial filament limits spore production in this filament to the portion beyond the insertion of the branch; and in the same manner the proliferation of a secondary from a primary lateral branch results in a sterilization of the portion of hypha below the insertion of the new branch. In one species, Actinomyces V, sporu- lation is even further restricted by the apparent abortion of a number of potential spores at the proximal end of the unbranched lateral branches. The hyphal portion involved first develops as usual, but when the characteristic septation associated with the delimitation of spores in this species appears in the spiral, it is not extended to the base of the branch, although indications of regu- larly spaced membranes may usually be distinguished (figs. 21, 19 1 9] DRECHSLER— ACTINOMYCES 79 24, 25). Later the unsegmented portion is gradually evacuated and converted into a sterile stalk devoid of protoplasm (figs. 29, 30). It is interesting to note that the basal septum, which in an allied and very similar form, Actinomyces VI, delimits the lowest spore from the axial filament, here also is present as a well developed cross- wall. The delimitation of the ultimate cells in the process of sporula- tion occurs usually as the growth in thickness and the contraction of the spiral (where this is present) are approaching a stage of completion. It has usually been believed by investigators that the details connected with spore formation are uniform throughout the genus. This belief, which the writer was at first inchned to share, must be considerably modified in view of the diversity of conditions actually found. In most species the sporogenous h>phae become divided into regular cylindrical cells separated by septa; the latter generally stain deeply with Delafield's haematoxylin, probably as the result of an association with metachromatic or possibly nuclear material. The species which are thus characterized by clearly defined septation may be assigned to three different categories, separated by differences in the disposition of their septa and in the development of their spores. In the first category, represented by Actinomyces I, the cross- walls in the sporogenous h)q)hae remain without any very pro- nounced change, continuing to separate the adjacent cells until these have developed into a chain of mature contiguous spores. The insertion of these septa progresses from the tip toward the base, and does not break the physiological continuity of the hyphae; for food material apparently is readily transported through them to the young spores at the termination, since these subsequently increase in size, and may deposit a wall of measurable thickness. In the second category the septa apparently split into halves, which are then drawn apart by the longitudinal contraction of the individual protoplasts (figs. 5CI, 8a-/, 59). In Actinomyces II the very pronounced growth in thickness of the sporogenous hyphae, following the insertion of septa, indicates that in this species also septation brings about no impediment in the transfer of food material. This is particularly remarkable on account of the 8o BOTANICAL GAZETTE [january extraordinary thickness of the septa characterizing this species. Actinomyces XVII, however, while less striking, probably repre- sents more nearly the usual condition prevailing in the second category. The segment of the filament wall evacuated by the contraction of each two successive spores undergoes no change until fractured by the disruption of the chain of mature spores. In the third category {Actinomyces IV, V, VI, VII, and XII) the cross-walls first undergo a deep constriction, which by involv- ing the ends of the young cylindrical spores gives to the latter an elongated ellipsoidal shape (fig. joa-d). The constricted septum now gradually loses its staining properties, and appears to become slightly drawn out in a longitudinal direction (fig. joe). A prepara- tion stained with Delafield's haematoxylin usually shows many old spore chains in which the individual spores are thus connected by hyaline isthmuses. Occasionally an isthmus may be found with a remnant of the old deep staining septum still unchanged in its center (figs. i6, joe). Beyond these three t3^es of sporulation another must at least be provisionally recognized, in which septa are either absent from the developing sporogenous hyphae, or are not demonstrated by the use of ordinary stains. The protoplast appears to contract at regular intervals, yielding a series of non-contiguous spores, held together for a time by the connecting segments of evacuated fila- ment wall (fig. 73). It is this type of sporulation which Neukirch and his followers, in opposition to Lachner-Sandoval, believed to prevail throughout the genus. Neukirch's conclusion that septa are never involved in the sporogenesis of Actinofnyces certainly cannot be extended to the large majority of species ; and its applica- tion to any forms whatsoever is associated with some reasonable • doubt. The writer is inclined to believe that cross-walls appear in the development of the sporogenous hyphae probably throughout the genus, but in some members are too thin to be recognized as distinct septa. Such an interpretation is suggested by the wide range in the thickness of septa found to occur, from the very mas- sive structures of Actinomyces II, through those of moderate thick- ness in Actinomyces I, XII, and XVI, to the condition prevailing in 1919] DRECHSLER— ACTINOMYCES 81 Actinomyces III and VIII, where cross-walls can only rarely be distinctly perceived. All investigators, with the exception of Schutze, agree in attributing to the peripheral wall of the filament of Actinomyces an extreme thinness. Indeed, Kruse (ii) and others have urged the single contoured character of the membrane as an evidence of the bacterial affinity of the genus. It is only necessary to examine fungus forms like Chlorosplenium or PJwma, to convince one's self that the single contoured wall is generally characteristic of minute cells, whatever their taxonomic connections may be. Yet while the phylogenetic inference may safely be rejected, it still remains true that the peripheral wall of every species of Actinomyces, except possibly those of some old enlarged hyphae, cannot be made out as a distinct structure with double contour. In evacuated portions its location is indicated by only the faintest indication of its outline. Nor is this surprising when we consider that the maximum resolv- ing power of any combination of lenses employing visible light is approximately 0.17 ju. As this magnitude barely equals the widths of the thinnest cross-walls observed, it is not difficult to suppose that, in the t)pe of sporogenous hyphae represented in Actinomyces XIII, the dimensions of the partitions, like the fila- ment wall generally, fall below the limit of visibility. It is pertinent in this connection to emphasize the peculiarity in the nature of the cross- walls, the appearance of which in many species of Actinomyces initiates the development of the individual spores. Their unusual relative thickness, even in species in which they can be distinguished only with difficulty, but where neverthe- less their thickness must exceed o . 17 /x, in filaments with a diameter of only o . 9 /x, is indicative of a composition essentially different from that of the peripheral wall. This indication is strengthened by the strong affinity for dyes characteristic of the septa, the evi- dent ease with which they permit of the passage of food material, and their apparent plasticity of behavior, resulting in a median split in some species, and in others in a gradual constriction followed by a slow transformation into an attenuated isthmus. The disappearance of the deep staining derivatives of the septa from the ends of the young spores is in some species accompanied 82 BOTANICAL GAZETTE [January by the appearance of one or several deep staining granules within the spore. Whether the latter represent nuclei or bodies of meta- chromatic material cannot definitely be determined. It seems not at all improbable, however, that some of the structures that can be differentiated within the more mature spores, particularly those characterized by uniform size and moderate staining proper- ties (figs. I, 2, ;^^, :^s) are nuclei. In Actinomyces IV and XII they may frequently be distinguished comparatively early, before the septa, with which they alternate in regular succession, show any perceptible constriction, indicating that their existence is not related to the subsequent disposition of the partitions (fig. 67). When two of these bodies occur in the same spore they uniformly occupy opposite or diagonal positions (fig. 2, a, di). The question arises why these bodies, if they actually represent nuclei and not structures originating de novo, cannot be distinguished in the young continuous sporogenous hyphae. The only explanation that can be advanced is that the protoplasm in the earlier stages is too dense to make possible any conspicuous contact between cytoplasm and nucleus. Later, with the attenuation and vacuolization of the cytoplasm that occur with the maturation of the spore, apparently as the result of the deposition of a special wall, the nucleus becomes increasingly distinct, and in some species it constitutes the only spore structure clearly visible in the stained preparation (fig. 41). It cannot be denied, however, that granules having more the appearance of the metachromatic granules found in degenerate sterile filaments occur in the spores of some species, either alone or together with a nucleus-like body. They differ from the latter in taking a deeper stain; in having an absolutely smooth contour; in offering considerable variability in size; and when present in numbers assuming no definite orientation with reference to each other. They have been noted in those species in which the septa associated with the delimitation of spores is particularly massive; and in Actinomyces II (fig. 8/) their derivation from the excessive wall material seems reasonably well established. After the septa have separated along a median plane, the deep staining substance at each end may contract, yielding a number of spherical bodies inside of the spore. This process is probably of a more or less 1919] DRECHSLER— ACTINOMYCES 83 pathological nature, since in the usual type of development the wall material is gradually distributed through the inclosed proto- plasm, causing the normal mature spore, except for the presence of a vacuole, to take an almost homogeneous stain. Another indication of the similarity in nature existing between metachromatic material and the deep staining transverse septa of Actinomyces is found in the occurrence of both within peculiar large spherical structures. These structures appear generally to occupy nearly the entire lumen of the filament, and not infrequently are related to local enlargements. Occasionally, however, their diameter is considerably smaller than that of the hyphae (fig. 103). In any case they may contain either one or several peripherally located metachromatic granules, or a uniformly thick, well defined, deep staining, transverse septum, exactly median in position. It is interesting to note that whenever granules occur their surfaces in contact with the periphery of the structure represent portions of convex spherical surfaces conforming accurately to the confining surface; and whenever a septum is found traversing one of these structures considerably smaller in diameter than the filament, it does not extend into the protoplasm, but remains in its finished state as a curious partial partition. The germination of the spores of Actinomyces takes place readily in dilute nutrient solutions, such as i per cent glucose solution, or nearly any vegetable decoction. During the first few hours of incubation at a moderate temperature they increase considerably in volume by swelling. From i to 4 germ tubes are then produced, apparently more or less successfully, the approximate number being, in a measure, characteristic of the species. Specific characteristics are expressed also in the diameter of the h^-phae, and in the fre- quency of branching. Cryptogamic Laboratories Harvard University Cambridge, Mass. PROTOMARATTIA, A NEW GENUS OF MARATTIACEAE, AND ARCHANGIOPTERIS (with plate I AND THREE FIGURES) BUNZO Ha YATA The Marattiaceae constitute a small family which may be regarded as the survivor of a much larger group. At present only 6 genera are known: Archangiopteris,^ Alacroglossum,^ Angiopteris, Marattia, Christensenia (Kaulfussia), and Danaea. Archangio pteris was first discovered by A. Henry in Yunnan and was published in 1899 by Christ and Giesenhagen as a genus connecting Angiop- teris with Danaea. The type of the genus A. Henryi is shown to be one of the most primitive forms of Marattiaceae by its simply pinnate leaves and simple stelar structure.^ According to Gwynne-Vaughan/ the mature stem of Archangiopteris retains a stage which is rapidly passed through by the young plants of Angiopteris and other genera. The same seems to hold true as to the form of the leaves. As far as I have observed Angiopteris in its habitat in Formosa, the first 2 or 3 leaves from a young stock usually are simply pinnate, but die before they reach maturity and become fertile. The pinnae of these first leaves are much larger than of those that follow, and closely resemble those of Archangi- opteris in shape and size. We may infer, therefore, that Archangiop- teris represents the form of a primitive type to which the ancestor of Angiopteris may have belonged. Archangiopteris is most closely related to Macroglossmn,^ recently established by Copeland, both in the simply pinnate fronds and ' Christ, H., and Giesenhagen, K., Pteridographische Notizen. Flora 86: 72-85. 1899. * Copeland, E. B., The ferns of the Malay-Asiatic region. Philipp. Jour. Sci. 4:1-64. 1909. 3 LoTSY, J. P., Vortrage iiber botanische Stammesgeschichte 2:676. 1906. '' Gwynne-Vaughan, D. T., On the anatomy of Archangiopteris Henryi and other Marattiaceae. Ann. Botany 19:268. 1905. 5 Campbell, D. H., The genus Macroglossnm Copeland. Philipp. Jour. Sci. 9:199-223. 1914; The structure and affinities of Macroglossnm Alidac Copeland. Ann. Botany 28:651-669. 1914. Botanical Gazette, vol. 67] [84 1919] HAVATA—PROTOMARATTIA 85 in the elongated linear sori, but differs from it in the absence of ridges between the sori and in the dorsiventral rhizomes. Gwynne- Vaughan (loc. cit.) says that in the specimens examined by him there was no suggestion of dorsiventrality in the rhizome of Arch- angiopteris, and the leaf arrangement and the vascular structure indicated a radial symmetry. So far as I can judge from the figure given by the author/ however, there are some indications of dorsiventrality in the rhizome, as can be seen in the upward bending of the stipes of the leaves and in the fact that the rhizome ascends somewhat obliquely toward the apex; and it also may be inferred that the remaining portions of the rhizome not given in the figure very probably run horizontally. In Archangiopteris Sotnai, recently discovered in Formosa, and two other new species from Tonkin, which will be described later, the rhizomes are prostrate and show very clear signs of dorsiventrality. In the summer of 191 6 I was sent to Tonkin for collecting. I found there two new species of Archangiopteris and a t}pe of a new genus closely related to the latter. All these plants present an appearance very similar to other ferns, such as Coniogramme japonica, C. fraxinea, Diplaziopsis javanica, or Diplazium hanta- mense, and occur in very small numbers amidst a multitude of the previously mentioned ferns. As it is very rarely that one has the opportunity to meet with these plants of Archangiopteris and the allied new genus, it may not be entirely out of place if I should tell how I was led to discover these very rare and interesting ferns. Some 10 years ago I was greatly interested in learning of Henry's discovery of Archangiopteris, representing as it did an ancient type of the Marattiaceae, and I wondered whether there might not exist another species of the genus in Formosa, the flora of which I have since then been studying. In 191 5, when examin- ing collections sent by the late T. Soma from Formosa, I found among them a curious looking fern labeled Gymnogramme japonica. A glance at the specimen showed me that it was another type of Archangiopteris, which was then named and published as A. SomaiJ * Gwynne-Vaughan, D. T., loc. cit. pi. 10. figs, i and 2. ^Hayata, B., Icones Plantarum Formosanarum. 5:256; 6:154. />/. 79. 86 BOTANICAL GAZETTE [january The next year I went to Formosa to the native place of the plant on the bank of a rivulet in a dense forest at an altitude of about 2000 ft. in the northern part of the island, for I wished to see a living specimen of this highly interesting fern. On the first day of our search we were not successful. The difficulty of finding it is partly due to its extreme rarity and partly to its existence only among other ferns closely resembling it in external appearance. On the second day I was at last successful in finding a few speci- mens of Archangiopteris Somai. In the summer of 191 7 I went to Chapa in the mountainous regions on the boundary between Yunnan and Tonkin. There too I wondered if I might not have the opportunity of finding some Archangiopteris , and so I made a careful search, turning back the leaves of all similar ferns which I came across. At last, as I had expected, I found a stock of the desired genus, in the shade of the forest at an altitude of about 4000 ft., between Chapa and Mueng- Xen. It was just a single stock. The fern resembled A. Henryi, but differed from it in the absence of indusium scales in the middle of the sori. It was a new species which I propose to call A. sub- integra, a description of it being given in the present paper. Later on I went to Mt. Tamdao in the central part of Tonkin, and col- lected in a forest at an altitude of about 3000 ft. There I saw on the side of the forestry service path one poor specimen which I thought most certainly a species oi Archangiopteris before examining the fern. It was a sterile specimen, yet I believed it to be a plant of the same genus, until I found near by some fertile specimens. They revealed the fact that they were different from Archangiop- teris in fructification, all other characters being exactly like the latter. The sporangia of the newly discovered plant were quite fused together, reminding one exactly of those of Marattia, but differing from the latter in the long linear synangium. I thought that it might be a type of a new genus intermediate between Archangiopteris and Marattia; but I shall refer to this later. Not very far from there I collected a true A rchangiopteris, another new species, which I propose to call A. tamdaoensis . Archangiopteris, therefore, formerly a monotypic genus, has come to comprise 4 species. As to distribution, the species are extremely local. iqiq] HAYATA—PROTOMARATTIA 87 A . Henryi is only known from Mentzu (Yunnan) ; A . Somai exists in one or two spots in the northern part of Formosa ; A . subintegra occurs in one place in the mountains of Chapa between Yunnan and Tonkin; and A. tamdaoensis is found in one locality on Mt. Tamdao (Tonkin) ; and the new genus is also limited to one spot on Mt. Tamdao. Returning to the systematic position of the new genus, the most remarkable feature which separates the new type from all the other genera of the Marattiaceae is its elongated linear or even vermiform synangium. The other important characters are its horizontal dorsiventral rhizomes and simply pinnate fronds. Through the dorsiventral rhizomes it is related to Kaulfussia and Archangiop- teris; by the simple pinnate fronds it is aUied to Macroglossum, Danaea, and Archangiopteris; in the structure of the synangium its nearest of kin is to be found in Marattia. It is distinguishable from the latter, however, by its elongated linear synangium, simply pinnate fronds, and dorsiventral horizontal rhizome. It differs from Archaiigio pteris in having a synangium; from Macroglossum in the dorsiventral rhizomes and synangium; from Angiopteris in the simple pinnate leaves, dorsiventral rhizomes, and synangium; from Danaea in the synangium with a longitudinal common slit; and from Kaulfussia in the pinnate leaves and linear synangium. After considering all these cases, I am forced to the conclusion that the new plant must be a type representing a new genus. I propose to call it ProtomaraUia, as it bears exactly the same relation to Marattia as Archangiopteris does to Angiopteris. As was stated, Protomarattia closely resembles Marattia in the reproductive organs, while it is closely related to Archangiopteris in its vegetative organs. The similarity of the type of the new genus and Archangiopteris tamdaoensis in the fronds and rhizomes, even in the serration and venation, is really so very great that I entirely failed to distinguish the one from the other until I saw the sori. The protective arrangement of stipules and commissures of our plant is exactly like that in Archangiopteris. The synangium also, presenting a linear form, with comparatively thinner lateral walls and a little looser connection of locules, more or less tends toward the sorus of Archangiopteris, or even toward that of 88 BOTANICAL GAZETTE [januar-k Macroglossum} There can be no doubt, therefore, that the genus is closely related to MaraUia on the one hand, while on the other it is nearly allied to Archangiopteris. It represents presumably a form of an ancient and conserved type connecting MaraUia with Archangiopteris. Protomarattia, nov. gen. — Rhizoma dorsiventrale oblique vel horizontaliter prostratum, reliquiis stipitum dense obtectum, radicibus e latere ventrali oriundis. Folia circum rhizoma spirali- ter disposita ; stipulis e latere adaxiali stipitis commissura connexis ; stipitibus rectis duobus locis basi et superiore geniculato-incras- satis post finitas functiones e geniculo superiore et iterum basilari solutis; frondibus simpliciter pari- vel impari-pinnatis, pinnis patentibus vel interdum retrorsum refiexis; petiolulis pinnarum incrassatis. Synangium lineare submarginale e margine pinnae 4-6 mm. distans subsessile rima mediana longitudinali apertum; indusio e squamis numerosissimis laceratis inprimis constituto, demum evanescenti. Differt a MaraUia synangio lineari nee ovali, loculis multo numerosioribus, fronde multo minore simplicter nee pluries pinnata et rhizomate dorsiventrali repente nee ereeto; differt a Archangiopteris synangio. In Tonkin incola ad hue monotypica. Protomarattia tonkinensis, sp. nov. — Rhizoma incrassatum horizontaliter vel oblique repens dorsiventrale, in specimine exsic- cato nostro 15 cm. longum cum rehquiis foliorum 2.5 cm. latum, reliquiis stipitum et stipulis dense obtectum e latere ventrali radices incrassatas teretes emittens. Folia versus apicem rhizomatis 5-6 approximatim disposita. Stipites 30-40 cm. longi erecti plano- convexi in sectione duobus locis ad basin et ad 0.3 altitudinem a basi geniculato-incrassati, basi dense sursum pauce squamulati, squamuKs lanceolatis vel hnearibus circ. 2 mm. longis margine erosis; partibus incrassatis basilaribus geniculiformibus 2.5 cm, longis, 1 . 5 cm. latis utroque latere stipula et latere adaxiali com- missura instructis, stipulis coriaceis tenuiter 2-lobatis, lobis abaxi- alibus (anterioribus) minoribus late semirotundatis, i cm. latis, 7 mm. longis margine erosis et membranaceis folium proprium (in 8 Campbell, D. H., The structure and affinities of Macroglossiim Alidae Cope- land. Ann. Botany 28:664. 1914- ■ 1919] HAYATA—PROTOMARATTIA 89 alabastro) obtegentibus, lobis adaxialibus (posterioribus) majori- bus semi-obovatis 1.5 cm. longis 7 mm. latis margine erosis et tenuibus cum commissura alabastrum folii juxta venientis obte- gentibus; partibus incrassatis superioribus geniculiformibus, 2 cm. longis, 7-10 mm. in diametro sectionis in vivo nitidis; foliis post fini- tas functiones ex apice partis incrassatae basilaris et e parte incras- sata superiore texturae degenerationi solutis. Frondes simpliciter pari- vel impari-pinnatae in ambitu obovatae vel ovatae stipitem in longitudine fere aequantes, circ. 30 cm. longae, 20-25 cm. latae, pinnis 4-5, rarius 3-7, alternis inferioribus minoribus superioribus majoribus, patentissimis vel interdum retrorsum reflexis, pinnis superioribus lanceolatis plus minus falcatis, 25-28 cm. longis, 5-6 cm. latis apice ad caudam abrupte acuminatis (cauda apicali lineari, 2-3 cm. longa, medio 4 mm. lata, serrata), basi subito triangulari-cuneatis (parte cuneata i . 5 cm. longa latere recta integra), margine planis nee recurvis nee undulatis minute serrulatis (serrula triangulari-acuta ascendenti), supra atroviridibus nitidis subtus pallidissimis subglabris ad costam et venas paucissime squamulatis (squamula minutissima), costa utraque facie elevata 1-2 mm. lata in exsiccato nigricanti; venis parallels a costa angulo 80° egressis simplicibus vel e basi furcatis fere rectis prope marginem subite recurvo-ascendentibus a se 1.5 mm. distantibus ad apicem serrularum attingentibus apice baud incrassatis in exsiccato nigricantibus; petiolulis pinnarum circ. i cm. longis incrassatis squamulatis; rhachis frondis supra medio tenuiter sulcata in vivo angustissime alata 4-7 cm. longa; textura chartacea vel chartaceo- membranacea. Synangia numerosissima subsessilia secus venas vel venulas utroque latere costae prope marginem i -seriatim approximatimque disposita, e margine 3-5 mm. ditantia, linearia vel vermiformia 4-6 mm. longa, i mm. lata rima longitudinali mediana aperta, primum squamulis laceratis indusiorum obtecta demum nuda, loculis numerosissimis utroque latere receptaculi 20-60 dispositis. Habitat. — Monte Tamdao (Tonkin), in silva ad 3000 ped. alt., leg. B. Hayata, July 191 7. Archangiopteris Christ et Giesenhagen, Pteridographische Notizen, Flora 86:72. 1899; Bitter, Natiirliche Pflanzenfam. 90 BOTANICAL GAZETTE [January i'':439; Christensen, Ind. Filic. 62; Campbell, Eusporangiatae, Publication no. 140, Carnegie Institution of Washington 203; LoTSY, Vortrag Bot. Stammesgeschichte 2:675. Key to species Indusium scales present between two rows of sporangia in a sorus. . . .A. Henryi Indusium scales absent between two rows of sporangia in a sorus. Pinnae nearly entire, scales brown A. subintegra Pinnae serrulate. Scales dark brown or blackish A. tamdaoensis Scales brown A. Somai Archangiopteris Henryi Christ et Giesenhagen, Pterido- graphische Notizen, Flora 86:77. 1899; Gwynne-Vaughan, On _ the anatomy of Archangiopteris Henryi, etc., Ann. Botany 19:257-271. 1905. Habitat. — Mengtzu (Yunnan) ex Henry. Archangiopteris subintegra Hayata, sp. nov. (fig. i). — Rhizoma horizontaliter vel Fig. I. — Archaneiopteris , . , ,. . . • ,. , „ . plus minus oblique repens, in specimme subintegra Hayata. ^ ^ r- ' ... nostro exsiccato 9 cm. longum cum reliquiis stipitum 4 cm. latum dorsiventrale reliquiis stipitum et stipulis persistentibus dense obtectum, radicibus incrassatis e latere ventrah rhizomatis oriundis. Stipes 70 cm. longus erectus basi dense sursum pauce squamulatus, squamulis lanceolatis, duobus locis basi et medio incrassatus, parte basilari incrassata i cm. longa totiusque lata stipulata, commissuris ut videntur obsoletis. Frons simpliciter pari- vel impari-pinnata in ambitu obovata, pinnis 5-7, superioribus majoribus lanceolatis, circ. 25 cm. longis, 5 cm. latis apice acuminatis ad caudas lineares abeuntibus (cauda 2-3 cm. longa serrulata), basi acutis margine cauda serrulata excepta fere integris; costa utraque pagine elevata, venis simplicibus vel a basi furcatis, venis et venulis parallelis a se 2 . 5 mm. distantibus patentissimis a costa angulo 85° egressis subrectis sursum plus minus recurvis; pagine supra nitida atro- viridi, subtus pallidissima; petioluhs incrassatis 7 mm. longis squamulatis; textura membranacea. Sori lineares, i cm. longi, I mm. lati utroque latere costae i -seriatim inter costam et marginem iqiq] HAYATA—PROTOMARATTIA 91 dispositi a se 2.5 mm. distantes, primum squamulis filiformibus indusii obtecti demum nudi. Habitat. — Inter Chapa et Mueng-Xen (Tonkin), in silva ad 4000 ped. alt., leg. B. Hayata, July 191 7. Very distinct from the other members of the genus by the much thinner subentire pinnae. Archangiopteris tamdaoensis Hayata, sp. nov. (fig. 2). — Rhizoma horizontaliter vel plus minus oblique repens dorsiventrale in specimine exsiccato nostro, 9 cm. longum cum reliquiis stipitum 3 cm. crassum e latere ventrali radices incrassatas teretes emittens, reliquiis stipitum et stipulis persistentibus dense obtectum. Stipes 40-45 cm. longus basi dense sursum pauce squamulatus, squamulis lanceolatis apice acuminatis vel ad acumen filiforme abeuntibus, 3-5 mm. longis, locis duobus ad basin et ad 0.3 altitudinen^ a basi incrassatus; parte incrassata basilari geniculiformi, 1-2 cm. longa, i cm. lata utroque latere stipula persistenti instructa, stipulis coriaceis latere adaxiali stipitis a r I • 1 1 X- 1 1 • • Fig. 2. — Archangiopteris commissura 2-iida connexis 2-lobatis, lobis , , • u * tanidaoensts Hayata. anterioribus minoribus rotundatis 7 mm. latis totiusque longis margine erosis interiore recurvis, lobis posteri- oribus majoribus semi-oblongis, 2 cm. longis, 7 mm. latis margine erosis membranaceis; parte incrassata superiore stipitis geniculi- formi, 2 cm. longa, i cm. lata; stipites post finitas functiones ex apice partis basilaris incrassatae et iterum e parte incrassata superi- ore texturae degenerationi soluti. Frons simpliciter pinnata in ambitu obovata pari- vel impari-pinnata, pinnis 3-4, superioribus majoribus lanceolatis, 23 cm. longis, 5 . 5 cm. latis apice gradatim acuminatis rarius' subito acuminatis (acumine lineari 2 .5 cm. longo, 2 mm. lato margine subintegro rarius serrulato), basi triangulari- acutis vel cuneatis margine praeter basin acumenque minute serrulatis (serrula triangulari-subacuta) ; costa utraque elevata, venis lateralibus numerosissimis parallelis patentibus a costa angulo 70° egressis subrectis simplicibus vel e basi furcatis, venulis ad apicem serrulatum attengentibus, a se 1.5 mm. distantibus; pagina supra atroviridi subnitida subtus pallidissima, supra glabra 92 BOTANICAL GAZETTE [JANUARY subtus paucissime squamulata vel glabra; petiolulis 5 mm. longis incrassatis. Sorus linearis utroque latere costae i -seriatim secus venas vel venulas dispositus, 7-8 mm. longus, i mm. latus e costa circ. I cm. e margine circ. 4 mm. distans, primum squamulis filiformibus indusii obtectus demum nudus. M L.i Fig. 3. — Archangiopteris Somai Hayata. Habitat. — Mt. Tamdao (Tonkin), in silva ad 3000 ped. alt., leg. B. Hayata, August 191 7. Allied to A. Somai Hay., but distinguishable from it in the shorter sorus located much nearer to the margin than to the costa, and in the less patent veins. Archangiopteris Somai Hayata, Ic. 6:154. pi. 19 (fig. 3). Habitat. — ^Urai (Formosa) . PL Formosa 5:256; EXPLANATION OF PLATE I Protomaratlia tonkinensis Hayata: i, plant; 2, basal portion of stipe with stipules and a commissura, seen from abaxial side; 3, same portion seen from adaxialside; 4, scale on stipe; 5, young synangium; (5, full grown synangium; 7 , portion of same synangium showing chambers; 8, section of same. BOTANICAL GAZETTE, LXVII PLATE I .:^ •: H J- HAYATA on PROTOMARATTIA CURRENT LITERATURE BOOK REVIEWS Fossil plants The third volume of Seward's Fossil plants^ will be welcomed alike by- students of paleobotany and by those whose primary interest is in the mor- phology and phylogeny of the living vascular plants. The volume, comprising chapters xxvii-xxxix of the complete work, is devoted to Gymnosperms, the space being distributed as follows: Cycadales (recent) 34 pages, Pterido- spermae 140, Cycadofilices 39, Cordaitales 86, Paleozoic gymnospermous seeds 66, Cycadophyta (fossil) 226, Bibliography of Vols. Ill and IV 48, Index 17, making a total of 656 pages. There are 252 figures, many of which are original. The account of the living cycads, from the standpoint of a competent paleon- tologist, is particularly interesting and suggestive to one who, like the reviewer, is somewhat familiar with those forms, but is dependent upon investigators like Seward for descriptions of their extinct predecessors. This introductory chapter is a fitting introduction to the more detailed treatment of paleozoic and mesozoic members of the phylum. The practical advantage of such an intro- duction is sufficient excuse for treating the living cycads first instead of consider- ing them in their natural place at the end of their phylum. The author believes the antiquity of that part of the cycadophyte phylum represented by the living cycads cannot be determined, but it is probable that if cycads, apart from Ben- nettitales, existed in the Jurassic and lower Cretaceous beds, they occupied a very subordinate place in comparison w'ith the Bennettitales. While the living cycads resemble the Bennettitales in many vegetative features, we believe that the reproductive structures show a kind of difference which would make it impossible to derive the living cycads from any forms of the Cycadeoidea type; while, on the other hand, the Cycadofilicales, which Seward prefers to call Pteridospermae, have reproductive structures from which the cones of living cycads might easily be derived. If the living cycads have come from Ben- nettitales, they must have come from ancient types in which the mega- sporophylls still retained a distinct leaflike character. Whether they have come from the Bennettitales or directly from the Cycadofilicales, they must have greater antiquity than is indicated by any material yet discovered. We agree with Seward that the afiinities are still in doubt, but we hope that Triassic material which can be sectioned will be found and that it will clear up relation- ships, for, it seems to us, the differentiation must have taken place long ago. ' Seward, A. C, Fossil plants, a textbook for students of botany and geology. Vol. III. Pteridospermae, Cycadofilices, Cordaitales, Cycadophyta. 8vo. pp. xviii-|- 656. figs. 25J. Cambridge University Press. 1917. 93 94 BOTANICAL GAZETTE [january We should have treated the Pteridospermae, Cycadophyta, and Cycadales together as a cycadophyte phylum. The Cycadofilices, including fernlike plants which may belong to the Pteridospermae but in which seeds have not yet been discovered, naturally follow the known Pteridospermae; but it does not seem natural to treat the Cordaitales between the Pteridospermae and the Cycado- phyta. After a careful reading of the Pteridospermae, we still fail to see why they should not be regarded as an order of the gymnosperms rather than as a group of equal rank. However, these are minor and very insignificant objec- tions. The book is full of detailed descriptions and critical discussions which will make it possible for investigators with far less training than Seward to make valuable studies of such material as may fall into their hands. The Pteridospermae are introduced by an excellent description of Lyginop- teris, the name applied to the plant whose various fragments have been described under the names Lyginodendron (stem), Sphenopteris (leaf), Lagenosloma (seed), Crossotheca (microsporophylls) , and Kaloxylon (root). The descriptions of Heterangium and Medtdlosa, while less complete, give a critical account of what is known up to date. The presentation of these 3 forms, with comparatively fragmentary accounts of others, shows where research is needed, and will enable students to fill in missing phases of life histories as material becomes available. In all the paleozoic forms of the cycadophyte phylum, information in regard to to the gametophytes and embryo, although very desirable, is very scant; but if attached seeds could be found and sectioned, the preservation seems good enough to show the desired features. The treatment of the Bennettitales (Cycadophyta), although it occupies 226 pages, seems short in comparison with the big volumes of Wieland. The English and French contributions to our knowledge of this group are presented in considerable detail, and the author has drawn upon Wieland for numerous excellent figures. If well-preserved reproductive structures of the lower mem- bers of this group, especially Williamsonia, could be found and sectioned, the results could not fail to be important, for they would ahnost certainly throw light upon the origin of the living cycads. The Cordaitales, representing the coniferophyte phylum, do not occupy so much space, but comparatively little is known about the group. If our knowl- edge of these forms were as complete as in case of the Bennettitales, a treatment of the Coniferales would be much simpUfied. As it is, the various stems, leaves, and reproductive organs referred to this group are described under their respec- tive categories, and material is thus accumulating for a connected life history. The chapter on paleozoic gymnospermous seeds is particularly conservative and interesting. Many morphologists would have felt little hesitation in assign- ing most of these seeds to one group or to another, but Seward, throughout the work, recognizes the danger of being too positive when dealing with unattached fragments. The characters of the various types of seeds are described and dis- cussed. Although some knowledge of the internal structure is available, it is very evident that little is known in regard to the gametophyte. A knowledge of 1919] CURRENT LITERATURE 95 the internal structure of the seeds, especially the smaller seeds, might help to connect the Cordaitales with the Pteridophytes. The fact that the geographical distribution of plants at different stages in the development of the earth receives only disconnected treatment is excused by the plea that the space needed for Vols. Ill and IV (now in press) was underestimated, the original plan providing for a treatment of geographical distribution at the end of Vol. IV. How^ever, Seward promises an entire volume devoted to this subject. Such a work would be welcomed by all students of morphology and phylogeny, and we hope that the volume will make its appear- ance at an early date. The complete bibliography and index, together with the critical and con- servative presentation of the entire subject, make the work indispensable to those engaged in research upon fossil plants.— C. J. Chamberlain. NOTES FOR STUDENTS Chlorophyll inheritance.— This subject seems to be a stumbhng-block both for plant geneticists and cytologists. In 1913 Emerson and East^ stated that there were on record only two indisputable cases of non-Mendelian inheritance. Both of these were cases of chlorophyll inheritance. Correns* made reciprocal crosses of a variegated Mirabilis {albomacidata) with normal green plants, and discovered that in this case inheritance was strictly maternal^ the pollen evidently contributing nothing. He explained this by assuming that the variegation was due to a disease of the cytoplasm which destroyed many of the chloroplasts, and that nuclei were immune to this disease. Thus the disease could be transmitted to progeny by the female parent only, since the male is supposed by cytologists to contribute only a nucleus stripped free from its cytoplasm. If one grants Correns' assumptions, the mechanism provided will explain this case of maternal inheritance without any violation of Mendel's law, for here there would be no true inheritance, but merely reinfection. Baur,'< working with a Pelargonium which had white-margined leaves, observed an occasional pure green branch and an occasional pure white branch. Flowers on these branches when self-fertilized gave respectively pure green and pure white progeny (the latter, of course, dying in the seedling stage). A cross either way between the two branches resulted in progeny which were a mosaic of green and white. Such behavior can be accounted for by either of two explanations, but each involves a very bold assumption. If there is a MendeHan determiner responsible for the full green development, and a white ' Emerson, R. A., and East, E. M., Inheritance of quantitative characters in maize. Bull. Agric. Exper. Sta. Nebr. no. 2. pp. 120. figs. 21. 1913. 3 Correns, C. E., Zeitschr. Ind. Abstamm. Vererb. 2:331-340. 1909. ■• Baur, Erwin, Zeitschr. Ind. Abstamm. Vererb. 1:330. 1909. 96 BOTANICAL GAZETTE [January plant lacks that determiner, it would not be an unheard-of thing for a cross between the two to show a mosaic (particulate inheritance). But for pure green and pure white branches to form and breed true sexually would involve somatic segregation. Such an explanation is hard to accept, since we have been confident not only that no general reduction division ever takes place in somatic tissue, but also that segregation in individual pairs of chromosomes or parts of chromosomes is impossible elsewhere than at spore formation. We might accept such a possibility for very rare monstrosities, but the case in hand seems to be a matter of fairly regular behavior. Mutation might also account for these results, but this too could hardly be expected to take place with such regularity. The other explanation seemed much more reasonable to Baur, but that too he acknowledged to be unorthodox. He assumed that this was not a matter of chromosomes but of plastids, and of course somatic segregation of green and white plastids is quite reasonable. If this mechanism be the true one, however, one must also grant that plastid initials are contributed by the male parent. This last is quite unorthodox and seems flatly contradictory of CoRRENs' ideas, for if enough cytoplasm is contributed by the male to intro- duce plastid initials, why should it not also contribute the diseased condition of CoRRENs' albomaculata ? Ikeno,s working on variegated races of Capsicum, confirms Baur's quaUtative results, and makes the case stiU stronger by uncovering some very significant quantitative features. "The offspring arising from the hybridiza- tion between a variegated and a green plant in either of two reciprocal ways contain a relatively far larger number of slightly variegated (less white) plants than those arising from the self-fertilization of the same variegated plant." The intensity of variegation may be progressively diminished by repeated crosses with green plants, but not even a single self-colored green has as yet been obtained in that way. Ikeno concludes that the transmission of varie- gation is not through the nucleus, but through the plastids in the cytoplasm; the male contributes cytoplasm and plastids. It seems impossible to reconcile this behavior with Correns' maternal inheritance. To assume that plastid initials originate within the nucleus might :smooth over the immediate difficulty, and would carry us into further comphca- tions. A more hopeful suggestion is that the disease which Correns speaks •of attacks only mature chloroplasts and that plastid initials are immune, as well as the cytoplasm around them. The easiest assumption, of course, would ibe to claim that Correns overlooked a case of apogamy. Otherwise we may "be driven to acknowledge that chlorophyll inheritance in angiosperms is governed by at least two mechanisms, which are not only quite different but directly contradictory. — Merle C. Coulter. 5 Ikeno, S., Studies on the hybrids of Capsicum annuiim. II. On some variegated races. Jour. Genetics 6:201-229. pi. 8. figs, i, 2. 1917. iqiqI current literature 97 Gonidia of lichens. — In 1905 Elfving, of the University of Helsingsfors, published his studies, which he thought disproved the recent view that the chlorophyllous elements associated with lichens are algae. He continued his work and published his results in 1913. In the interval Danilov* began s-tudies which disproved Elfving's conclusions. Danilov's results were published in Russian in 1910 and in English in 1918.' Elfvixg's conclusion was that the lichen hyphae threw out spherical cells, at first colorless, but later colored and very similar to algae. These he supposed became separated from the hyphae and divided rapidly within the lichen thallus, forming, according to his results, the "gonidia" of lichens. Reviewing these results, Danilov found on careful study that unstained preparations often left the impression that the algal cells might really be outgrowths of the lichen hyphae, with which they are intimately associated. By the use of stains, however, he was able to trace the entrance of the hyphae into the algal cells, thus proving that there is no genetic relationship, but that the relationship is rather that of host and parasite. The "pale gonidia" of ELFV^NG were found to be dead algae which had been killed by the parasitic lichen, and Daxilov was able to see distinctly the lichen hyphae within them. Important and quite apart from the refutation of the once generally accepted view of the origin of the chlorophyllous "gonidia" from the non- chlorophyllous lichens, are the conclusions of Danilov regarding the re- lation of the lichen to its algal host. He admits that there may be osmotic filtration of certain materials from the alga to the lichen, and the like passage of others from the lichen to the alga. However this may be, Danilov finds the final result to be the absorption of the algae by the lichen hyphae, which enter the algal cells and form dense networks of slender, thin-walled or naked absorb- ing threads. Although the lichen thaUus with its prepared peptones and cer- tain other organic materials is probably a favorable substratum for the algae, yet the lichen is parasitic on the algae, which are killed in large numbers as a result of the parasitism. On the whole the algae thrive better outside the association with the lichen, while the lichen does poorly or dies outright outside the association. — Bruce Fink. Sex organs of Phytophthora. — ^In 1913 Pethybridge,* studying a disease of the potato produced by a phycomycetous fungus which he named Phytoph- thora erythroseptica, observed that on the formation of the sexual organs of this * D.A.NILOV, A. N., tJber das gegenseitige Verhaltnis zwischen den Gonidien und dam Pilzkomponten der Flechtensymbiose. Bull. Jard. Imp. Bot. St. Petersb. 10:33- 70. ph. 3. figs. g. 1910. " , The relation between the gonidia and the hy-phae in lichens. Jour. Botany 56:169-181. 1918. * PETHYBRrocE, G. H., On the rotting of potato tubers by a new species of Phytophthora having a method of sexual reproduction hitherto undescribed. Sci. Proc. Royal Dublin Soc. 13:529-565. pis. j. 1913. 98 BOTANICAL GAZETTE [january fungus the oogonial hypha pushes its way entirely through the antheridium, and, after emerging on the side opposite to the point of entrance, enlarges to form the oogonium. This unusual process, together with the subsequent events in the formation of the oospore, has now been more fully investigated by Murphy,' whose cytological evidence bears out the observations of Pethy- BRIDGE. The antheridia and oogonia are found to arise on different branches of the mycelium. During the penetration of the antheridium by the oogonial incept no fusion of the cytoplasm of the two organs occurs. After its emergence the oogonial hypha develops into a more or less spherical multinucleate oogo- nium whose stalk passes through the antheridium. When the sexual organs have reached their full size, about two-thirds of the nuclei in the antheridium and in the oogonium degenerate. The remaining nuclei in both organs then divide once mitotically and simultaneously. During the division the nuclei of the oogonium are arranged in a hollow sphere, with the exception of one, which remains in the center. Immediately after the division the protoplasm of the oogonium separates into a vacuolate hyaline ooplasm and a denser peri- plasm. In the oogonium, and probably in the antheridium also, all the nuclei but one degenerate. During this period a prominent receptive papilla pro- trudes from the base of the oogonium into the antheridium. When the recep- tive papilla is withdrawn, the fertilization tube grows into the oogonium at the same point and discharges one nucleus and the greater part of the cytoplasm of the antheridium into the oogonium. With the completion of this process most of the periplasm has disappeared and the oospore is surrounded by a thin mem- brane with the last vestiges of the degenerating nuclei appressed against its outer surface. The fusion of the two nuclei does not take place until the thickened oospore wall has been completed. — H. Hasselbring. Action of neutral salts on acid inversion of cane sugar. — Lebert'" has studied the action of neutral salts on the acid inversion of cane sugar. His resxilts furnish him a basis for a chemical explanation of certain difficulties some- times encountered when attempts are made to invert cane sugar by means of w'eak acids or stronger acids in quantity just sufficient to effect the inversion. Solutions in which it is desired to invert cane sugar are rarely free from neutral salts, especially sodium acetate left in the solution after clearing with lead acetate and removing the excess of lead with sodium carbonate or sulphate. If the hydrolysis is effected by a relatively large quantity of strong acid, as in the Clerget method, the presence of a small amount of salt is of little conse- quence, since the H ions are in great excess. If organic acids are employed, the presence of their sodium or potassium salts will retard the rate of inversion, 9 Murphy, P. A., The morphology and cytology of the sexual organs of Phytoph- thora erythrose plica Pethyb. Ann. Botany 32:115-153. pis. 3. 1918. "Lebert, M., Action des sels neutres sur I'inversion du sucre par les acides. Rev. Gen. Botanique 30:241-244. 1918. igig] CURRE.XT LITERATURE 99 the decrease in the rate depending upon the strength of the acid; the weaker the the acid the greater the inhibiting action of its salt. The action of acetic acid is completely paralyzed by the presence of sodium acetate equivalent to the proportion of the acid. The effect of a salt other than the salt of the acid used for the inversion depends upon the relations established between the acid and the salt. An example would be HCl in the presence of sodium acetate; NaCl and acetic acid are formed. If the acetate is present in sufiicient quan- tity, all of the HCl is replaced by acetic acid, and if the acetate is still in excess, we have inversion by acetic acid in the presence of its sodium salt, in which case the hydrolysis is always inhibited. The author offers a similar explana- tion for a situation reported by Davis and Daish. They found that 2 per cent citric acid was sutTicient to invert a solution of cane sugar by boiling 10 minutes, but it was without effect in the presence of a certain quantity of sodium acetate. The citric acid reacted with the sodium acetate, giving sodium citrate and liberating an equivalent amount of acetic acid, the action of which was paralyzed by its sodium salt still present in the solution. — Charles O. Appleman. Effect of different oxygen pressures on carbohydrate metabolism of sweet potatoes. — The experiments reported by Hasselbring" in this paper were designed primarily to effect a further separation of the various steps in the transformation of starch to sugar in sweet potatoes. For this purpose different or>'gen pressures were employed. When the sweet- potatoes are killed under a gas pressure of 5 atmospheres, starch hydrolysis is greatly depressed or in- hibited. In the living potatoes starch hydrolysis and cane sugar formation proceeded in the absence of oxygen in the same manner as in air or in an atmos- phere of oxygen. Cruickshaxk working with barley seed, and Boysen- Jensex working with germinating barley and peas, found that cane sugar was not formed in the absence of oxygen. These investigators conclude that the presence of oxygen is one of the necessary conditions for cane sugar formation, but since this was not found to be the case with sweet potatoes, the conclusion is not of general applicability. Anaerobic respiration in sweet potatoes consumes, in a given period of time, a greater quantity of material than is consumed by normal respiration. The energ>'^ derived from a given mass of material is less in anaerobic than in normal respiration. These facts, coupled with the observation that cane sugar is formed with equal facility under anaerobic and aerobic conditions, lead the author to believe that his experiments in a general way support the Boysen- Jensen theory that the respiratory processes furnish the energy for the syn- thesis of cane sugar. In the case of the sweet potato this energy could be furnished by anaerobic respiration. " Hasselbring, Heinrich, Effect of different oxygen pressures on the carbohy- drate metabolism of the sweet potato. Jour. Agric. Research 14:273-284. 1918. lOO BOTANICAL GAZETTE [jaxuary Another very interesting fact brought out by the author's work on sweet potatoes is the apparent stabiUty of cane sugar in relation to the respiratory processes in these roots, as cane sugar does not seem to be consumed by either anaerobic or normal respiration. — Charles O. Applemax. Analysis of quantitative variation. — Brotherton and Bartlett" have presented the results of a very significant piece of research. The investigation as it stands belongs to the field of plant physiology, but probably it is most significant in the bearing upon certain problems of genetics. Plants of Phase- oltis midtiflorus grown in hght and darkness were compared as to length and number of epidermal "cells of a given internode. For the physiologist the results may be summarized in the following statement: "The effect of hght is that it retards extension of the cells, and that as an indirect result there are fewer secondary di\dsions, since relatively fewer primary cells enter the range of length within which division takes place." For the geneticist we quote the following: "The mathematical formulation of the results of size inheritance according to the multiple factor hypothesis should be paralleled by a biological analysis, the object of which is the identification of the several factors con- cerned." Thus size differences may be resolved into number or size of con- stituent cells or both. "In the investigation of quantitative variations of a hereditary nature it seems Ukely that the study by the histological method of reactions to the en\aronment and of the obscure reaction known as 'vigor of heterozygosis' will afford a means of correcting for these disturbing factors." It is probably true that heritable size differences express themselves directly in the cells of tissues deeper than the epidermis, and that the change in the epidermis amounts merely to a mechanical response to these forces within. It would probably be advisable, therefore, to carry the analysis to more sig- nificant tissues. — Merle C. Coulter. Root growth in cuttings. — Curtis'^ has published an important contribu- tion to the physiology of root formation in cuttings. A number of forms were used, but Ligustrum ovaliJoUum furnished most of the experimental material. Nutrient solutions of the strengths used in culture work with seedlings were found to be distinctly injurious to woody cuttings. Treatments with potas- sium permanganate resulted in a xexy marked increase in root growth of various woody cuttings. After discussing several possible explanations for this stimula- tion, the author concludes that it is most probable that the potassium per- manganate increases respiratory activity by catalytically hastening oxidation. It is known that when potassium permanganate comes in contact with organic matter manganese dioxide is precipitated and oxygen is liberated. There was " Brotherton, Wilber, and B.artlett, H. H., Cell measurement as an aid in the analysis of quantitative variation. Amer. Jour. Hot. 5:192-206. 1918. '3 CtJRTis, Otis F., Stimulation of root growth in cuttings by treatment with chemical compounds. Cornell Univ. Agric. Exper. Sta. ^Memoir 14:71-138. 1918. 1919] CURRENT LITERATURE loi some indication that other inorganic compounds may stimulate root growth in cuttings. The author's work gives further strong evidence that callus and root growth is independent of the rest period and that only the buds assume the resting condition. Immature twigs were caused to absorb cane sugar which increased root development. Mature twigs, however, were but slightly bene- fited. When the base of cuttings w-ere placed in sugar solution for a short time, the terminal bud of the twig failed to develop in a normal manner and the lower buds formed shoots instead. The author believes that many of the practices commonly followed by greenhouse and nursery men in the propagation of plants by cuttings are explainable on the basis of better aeration. The discussions of the literature are comprehensive and critical. — Charles 0. Appleman. Vegetation of Newfoundland. — In contrasting the divergent floras of different parts of Newfoundland, Fernald'^ bases his explanation of their differences upon the hypothesis that "the presence or absence of varying degrees of available lime or of other bases in the soil is more fundamental in determining plant distribution than are even considerable differences of temper- ature and humidity." The calcareous and at the same time the most fertile portion of the island is along the west shore, where the ordinary observer would be surprised to find the indigenous flora of the warmest and most fertile region of the island composed very largely of species of high northern distribution, such as Juncus triglumis, Saxifraga oppositifolia, S. aizoides, S. caespitosa, Salix vestita, Dryas integri- folia, and Lesquerella arctica. These Fernald explains as being from the cal- careous habitats of the arctic archipelago and the Canadian Rockies, the Ume being hostile to the plants of the siliceous adjacent mainland. The eastern part of the island, the central tundra district, and the southwest corner, in spite of the fact that they are cold, bleak, and barren, are populated mainly by plants of the southern Atlantic coast region, with an addition of some like Calliina vulgaris and Pedicidaris sylvatica from the acid soils of western Europe. Maps of the distribution of a dozen species give graphic demonstration of the remarkable distribution of some of the more important plants and serve to make the evidence in the support of his hypothesis the more convincing. — Geo. D. Fuller. Physiological role of glucosides in plants. — Continuing his investigations on the physiological role of glucosides in plants, Combes'^ has made the inter- esting discovery that a given glucoside is not toxic to a plant which naturaUy •-t Fernald, M. L., The contrast in the floras of eastern and western Newfound- land. Amer. Jour. Bot. 5:237-247. pis. j. 1918. 's Combes, Raoul, Recherches biochemiques experimentales sur le role physio- logique des glucosides ches les vegetaux. Rev. Gen. Botanique 30:226-237, 245-257. 1918. I02 BOTANICAL GAZETTE [January produces it, but is very toxic to plants belonging to a species in which the glucoside is not naturally found. The toxic glucoside, when added to Knop's culture medium in which the plants are grown, produces very marked abnormal changes in the morphology of the roots, resulting also in a very stunted growth of tops. It appears that we have here another group of substances, the indi- viduals of which possess a constitution suf3ficiently characteristic of the species in which they are found that when they are applied to individuals belonging to nonrelated species they produce abnormal responses. The author has not yet found that glucosides will furnish carbohydrate food for plants when they are grown in a carbon dioxide free atmosphere, as has frequently been found to be the case with glucose. Those wishing to germinate seeds and grow seedlings under aseptic condi- tions will be interested in the detailed descriptions of the apparatus and pro- cedures employed in growing his plants. An excellent review of the mass of literature on the subject and a survey of glucosides in plants will be found in the earlier papers of this series. — Charles O. Appleman. Physiology of fungi. — Duggar, Severy, and Schmitz'^ have undertaken a study of the comparative nutrient value of some of the decoctions ordinarily used in the preparation of culture media for fungi. The decoctions which were prepared on the basis of 50 gm. of dry matter to a liter of water were made from bean, sugar beet, prune, potato, turnip, cornmeal, apple, mangold, celery, carrot, and salmon. The standard decoctions were employed alone and in combination with sugar and various mineral nutrients. It was found that in their nutrient value the decoctions are very dissimilar for different fungi. The addition of sugar in most cases increases the yield, but the addition of sugar with nitrate and phosphate gives a very much greater yield than the addition of any of these substances alone. It is pointed out also that the standardiza- tion of the decoctions on Fuller's scale leaves them differing widely in hydrogen ion concentration. This work brings out the fact that little is really known of the nutrient value of plant decoctions, which it appears are generally deficient in nutrients and require the addition of considerable "fertilizer" to produce the greatest growth of fungi. — H. Hasselbring. Maps of rainfall and crop plants. — Among the recent publications of the United States Department of Agriculture there are two at least of decided interest to ecologists and plant geographers. The first is a rainfall map of the United States'" embodying the data from not less than 3600 stations. The precipitation is given in inches and the map is in 8 shades of blue. An interest- ing insert map gives the rainfall from April i to September 30, and exhibits a '* Duggar, B. M., Severy, J. W., and Schmitz, H., Studies in the physiology of fungi. Ann. Mo. Bot. Gard. 4:165-173; 279-288. 1917. '' KiNX'ER, Joseph B., Atlas of American agriculture. Advance sheet i : Precipi- tation. U.S. Dept. Agric. Weather Bur. 191 7. iqiq) current literature 103 close relationship between the maximum summer rainfall and the grasslands of the country. The second publication'^ contains relief and precipitation maps of the world and numerous larger and smaller maps showing the agricultural produc- tion of all lands. Many other data are contained in the text and in various tables. Several recent papers have successfully related crop possibilities to natural vegetation, but these maps provide material for reversing the process and of relating natural vegetation to areas of crop production. — Geo. D. Fuller. Seedling of dicotyledons. — Sinnott'' has made a comparative study of the seedling throughout dicotyledons, in order to distinguish between conserva- tive and variable characters. It is a very timely distinction to emphasize, for the application of the law of recapitulation to variable characters has led to more or less confusion. The number of protoxylem poles is found to be a very variable character. More constant is the relation between the vascular system of the hypocotyl and that of the epicotyl, two main types being recog- nized. The venation of the cotyledon was found to be very constant; and also an odd number of veins was found to characterize the seedling of all dicotyledons, distinguishing it from that of the gymnosperms. The most conservative character is the structure of the cotyledonary trace. — ^J. M. C. Monographs on experimental biology. — The first volume of a series of monographs dealing with experimental biology and general physiology has appeared under the editorship of Jacques Loeb, T. H. Morgan, and W. J. V. OsTERHOUT. The first monograph^" deals with forced movements, tropisms, and animal conduct. Among the monographs in preparation are "The chromo- some theory of heredity" by T. H. Morgan; "Inbreeding and outbreeding; their genetic and sociological significance," by E. M. East and D. F. Jones; "Pure line inheritance," by H. S. Jennings; "The experimental modification of the process of inheritance," by R. Pearl. This series represents an important event in American science, and deserves the cooperation of the scientific men of the country. — J. M. C. A new phytopathological journal. — The first number of the Annals of the Phytopathological Society of Japan has just appeared, including 5 papers. Some of the papers are in English, and those in Japanese include a summary in English, so that all of them are available for foreign botanists. The contrib- utors to this first number and their titles are as follows: M. Shirai, "On the •8 Finch, V. C, and Baker, O. E., Geography of the world's agriculture. loX 13.5 inches, pp. 149. jigs. 207. 1917. '' SiNNOTT, E. W., Conservatism and variability in the seedling of dicotyledons. Amer. Jour. Bot. 5:120-130. figs. 4. 1918. ^"LoEB, Jacques, Forced movements, tropisms, and animal conduct. 8vo. T^Y). 2og. figs. 42. Philadelphia: J. B. Lippincott Co. 1918. $2.50. I04 BOTANICAL GAZETTE [january development of plant pathology in Japan"; S. Ito, "A preliminary report on a late blight resistant strain of potato"; T. Hemmi, "Vorlaiifige Mitteilung iiber eine neue Anthraknose von Evonymus japonica"; S. Miura, "On the grain of barley or wheat, infected by smut fungus through the flower " ; S. HoRi and U. BoKURA, "Soy bean cake as a substitute for peptone in the preparation of nutrient media."— J. M. C. Evolution of maize. — Weatherwax^' has made a detailed study of the origin of maize, concerning which there has been much discussion. A com- parative study of many varieties of maize and related species has led him to the theory that vestigial organs indicate that Zea, Euchlaena, and Tripsacum are of the same structural type, their present peculiarities being due to the suppression of parts present in a primitive ancestor with perfect flowers and one type of inflorescence. The ear of maize is regarded as the homologue of the central spike of the tassel. The prevailing theory that maize is of hybrid origin he regards as untenable, his conclusion being that Zea and the other two genera mentioned "have descended independently from a common ancestral form now extinct." — J. M. C. Ferns of Borneo. — Copeland" has brought together in a convenient list the ferns of Borneo, accompanied by analytical keys. The fern flora is very impressive, including 697 recorded species, representing 88 genera. In another paper^'' the same author shows that the riches of the fern flora are far from exhausted, for he describes 43 new species from Borneo, 12 of which are species of Cyathea, and also a new genus {Oreogrammatis) related to Polypodium. — J. M. C. Bryophytes of Iceland. — Hesselbo^i has published a rather complete account of the Bryophytes of the island of Iceland. His annotated Ust shows 93 species of the Hepaticae, 20 of the Sphagnales, and 325 of the Musci. These he further discusses as to their aggregation in communities and their altitudinal and horizontal distribution. — Geo. D. Fuller. " Weatherwax, Paul, The evolution of maize. Bull. Terr. Bot. Club 45:309- 342. figs. 36. 1918. "CoPELAND, E. B., Keys to the ferns of Borneo. Sarawak Mus. Journ. 2: 287- 424. 1917. 23 , New species and a new genus of Borneo ferns. Philipp. Jour. Sci. 12: 45-65. 1917. ^ Hesselbo, Aug., The Bryophyta of Iceland. The Botany of Iceland. Ed. by RosENviNGE, L. K., and Warming, Eug. i: pt. II. 397-676. 1918. The University of Chicago Science Series EDITORIAL COMMITTEE ELIAKIM HASTINGS MOORE. Chairman JOHN MERLE COULTER ROBERT ANDREWS MILLIKAN The Evolution of Sex in Plants. By JOHN MekI£ Coulter. x4-I40 pages, i2mo, cloth; $i.oo, postage extra (weight 14 oz.). Individuality in Organisms. By CHARLES MANNING Child. x+aia pages, small lamo, cloth; $1.25, postage extra (weight i lb. 6 oz.). The Origin of the Earth. By Thomas C. Chamberiin. xii+272 pages, small i2mo, cloth; $1 .50, postage extra (w6ight i lb. 6 oz.). Finite Collineation Groups. By Hans F. Blichpeldt. xii+194 pages, small i2mo, cloth; $1.50, postage extra (weight i lb.). A Chemical Sign of Life. By Shiro TaSHIRO. x+142 pages, small i2mo, cloth; $1 .00, postage extra (weight 14 oz.). The Biology of Twins. By HoRATio Hackett Newman, x+i86 pages, small i2mo, cloth; $1.25, postage extra (weight i lb.). Food Poisoning. By Edwin Oases Jordan. viii+ii6 pages, small i2mo, cloth; $1.00, postage extra (weight 11 oz.). The Electron: Its Isolation and Measurement, and the Determination of Some of Its Properties. By ROBERT ANDREWS MiLLIKAN. xii+ 268 pages, small i2mio, cloth; $1 .50, postage extra (weight i lb. 3 oz.). The Living Cycads. By CHARLES JOSEPH Chamberlain. 160 pages, small i2mo, cloth; $1.50, postage extra. The Problem of Fertilization. By FRANK R. LlLUE. 272 pages, small i2mo, cloth; $1.75, postage extra. VOLUMES PLANNED OR IN PREPARATION Black Body Radiation. By Charles E. Mendenhall. Mechanics of Delayed Germination in Seeds. By W. CrocKER. The Rigidity of the Earth and of Materials. By A. A. MiCHELSON. Linear Integral Equations in General Analysis. By EliAkim HASTINGS MoORE. THE UNIVERSITY OF CHICAGO PRESS CHICAGO ILLINOIS AN ANNOUNCEMENT ^ Our activities for the past year have been nearly loo per cent devoted to war production of vital importance to the government, compelling us to discontinue regular manufacture in practically all of our lines. ^ The cessation of hostilities has now enabled us to begin the readjust- ment to a peace production basis. While this adjustment must be gradual, we are pressing it with all possible speed in order to satisfy the requirements of our normal trade at an early date. ^ We appreciate the patience and understanding with which our patrons have reacted to the situation and trust they may continue to exercise such consideration until our manufacturing facilities are once more on a pre-war footing. Bausch ^ Ipmb Optical ©. 41 2 ST. PAUL STREET HOCH ESTER, N.X NEW YORK CHICAGO SAN FRANCISCO WASHINGTON Leading American Makers of Microscopes, Projection' Apparatus (Baloptlcons J .Photographic Lenses, Ophthalmic Lenses and Instruments, Range-Finders and Gun Sights for Armu and Navy, Searchlight Reflectors, Binoculars, and Other High-Grade Optical Products FINE INKS AND ADHESIVES For those who KNOW Higgins' Drawing Inks eternal Writing Ink Engrossing Ink Taurine Mucilage Photo Mounter Paste Drawing Board Paste Liquid Paste Office Paste Vegetable Olue, Etc. Are the Finest and Best Inks and Adhesives Emancipate yourself from the use of corroiive and ill-smellinf inks and adhesives and adopt the Hlggins Inks and Adhesives. They will be a revelation to you, they are so sweet, clean, well put up, and withal so efficient. At Dealers Qenerally CHAS. M. HIGQINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street Brooklyn, N.Y. The Home Typewriter as well as The Business Typewriter IS the Self-Starting All the family can use it. Write for descriptive matter. REMINGTON TYPEWRITER COMPANY Incorporated 374 BROADWAY NEW YORK CITY Pr olume LXVII Number 2 THE Botanical Gazette Editor:. JOHN M. COULTER FEBRUARY 1919 Blister Canker of Apple Trees; A Physiological and Chemical Study. Contributions from the Hull Botanical Laboratory 246 Dean H. Rose 105 iWith ten figures) Morphology of the Genus Actinomyces. H - - Charles Drechsler 147 (With Plates H-IX) Briefer Articles Byron David Halsted ------- Mel. T. Cook 169 (With portrait) Current Literature Book Reviews - - - - - - - - - - -1 71 Soil conditions and plant growth Minor Notices - - - - - - - - - - -173 Notes for Students - - - - - - - - - -174 The University of Chicago Press CHICAGO, ILLINOIS, U.S.A. THE CAMBRIDGE UNfVERSlTY PRESS, London and Edinburgh THE MARUZEN-KABUSHIKI-KAISHA, Tokyo. Osaka. Kyoto. Fukaoka. Sendai THE MISSION BOOK COMPANY. Shanghai Volume LXVII Number 2 The Botanical Gazette A MONTHLY JOURNAL EMBRACING ALL DEPARTMENTS OF BOTANICAL SCIENCE EDITED BY JOHN M. COULTER With the assistance of other members of the botanical staff of the University of Chicago Issued February 15, 1919 The Botanical Gazette is published monthly by the University of Chicago at the University Press, 5750 Ellis Avenue, Chicago, 111. HThe subscription price is $7.00 per year; the price of single copies is 75 cents. Orders for service of less than a half-year .will be charged at the single-copy rate. ^Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Repubhc of Panama, Hawaiian Islands, Philippine Islands, Guam, Samoan Islands, Shanghai. IfPostage is charged extra as follows: For Canada, 35 cents on annual subscriptions (total I7.35), on single copies, 3 cents (total 78 cents); for all other countries in the Postal Union, 84 cents on annual subscriptions (total $7.84), on single copies II cents (total 86 cents). TJPatrons are requested to niake all remittances payable to The University of Chicago Press in postal or express money orders or bank drafts. The following are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. 4. Yearly subscriptions, including postage, £1 12s. 6d. each; single copies, including postage, 35. 6d. each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage, Yen 15.75 each; single copies, including postage, Yen 1.75 each. For China: The Mission Book Company, 18 Peking Road, Shanghai. Yearly sub- scriptions, $7.00; single copies, 75 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 84 cents, on single copies II cents. Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when losses have been sustained in transit and when the reserve stock will permit. DeUveries are not guaranteed within the war zones of Europe, nor in other countries where transportation is disturbed by con- ditions incident to war. Business correspondence should be addressed to The University of Chicago Press, Chicago, 111. Communications for the editor and manuscripts should be addressed to the Editor of the Botanical Gazette, The University of Chicago, Chicago, 111. Contributors are requested to write scientific and proper names with particular care, to use the metric system of weights and measures, and in citations to follow the form shown in the pages of the Botanical Gazette. Papers in excess of thirty-two printed pages are not accepted unless the author is willing to pay the cost of the additional pages, in which case the number of the pages in the volume is correspondingly increased. ........ Illustrations are furnished without cost to author only when suitable originals 'ate supplied. A copy of the suggestions made in the January number, 1907, will be sent on appHcation. It is advisable to confer with the editors as to illustrations required in any article to be offered. Separates, if desired, must be ordered in advance of pubUcation. A table showing approxS mate cost of separates is printed on an order blank which accompanies the proof; a copy vm be sent on request.. Entered as secohd-dass matter Aygust ar, i8g6, at the Post-Ofl5ce at Chicago, Illinois, under the Act_of March 3, 1879. Acceptance for mailing at special rate of postage provided -fc» in Section 1130, Act of October 3, I9i7i authorized on July IS, 1918. I \OLUiME LXVII NUMBER a THE Botanical Gazette FEBRUARY igig BLISTER CANKER OF APPLE TREES; A PHYSIOLOGI- CAL AND CHEMICAL STUDY CONTRIBUTIONS FROM THE HULL BOTANICAL LAB0R.4T0RY 246 Dean H. Rose (with ten figures) Introduction It is now generally recognized that among the most important problems of plant pathology are those coimected with the physiol- ogy of diseases whose etiology is already known. It is also recog- nized that this must be the physiology of the host, of the parasite, and of the two in relation to each other, and, further, that such a comprehensive view of all the factors involved furnishes the only rational approach to an understanding of the principles underlying immunity and disease resistance. In the present paper are given the results of a physiological study of the destructive disease known as Illinois or blister canker, the etiology of which, including the identity of the causal organism, Nummular ia discreta (Schw.) Tul., was worked out by Hassel- BRING (22) in 1902. The work reported here is a continuation of an earlier investigation b}- the writer (30) on the oxidase activity of healthy and diseased bark; in addition there is included an account of the catalase activity and microchemical and macro- chemical analyses of both kinds of tissues. Further work is planned on the chemistry of the disease, on the role of other enzymes than oxidases, and on the physiology of the fungus itself in pure culture. 105 lo6 BOTANICAL GAZETTE FEBRUARY The work was done in part at the Missouri State Fruit Experi- ment Station and in part in the Botany Department of the Uni- versity of Chicago. Historical The problem of oxidation by plant and animal tissues or tissue extracts has been studied by many investigators since the time of the pioneer work by Schonbein, the discoverer of ozone. An immense literature has accumulated, for reviews of which the reader is referred to publications by Clark (14), Kastle (24), Battelli and Stern (5), and Atkins (3). In this paper only those articles will be cited which bear directly on the problem in hand. That pathological conditions in plants are often accompanied by increased oxidase activity has been shown repeatedly in recent years. Woods (35) found greater oxidizing power in the chlorotic portions of tobacco leaves affected with mosaic than in the green portions; this has been confirmed by Allard (i) and by Frei- berg (20). SoRAUER (31, 32) and Doby (17), working with leaf- roll of potatoes, found oxidase activity greater in diseased tubers than in healthy ones, although the former makes the point that this greater enzyme activity is to be considered a symptom of the disease rather than the cause. Bunzell (ii), working with the curly-dwarf disease of potatoes, showed by an extensive series of tests that ''affected plants have a greater oxidase activity than healthy ones of the same age, both in the juice of their tubers and in the juice of their fohage." Similar results were obtained by Bunzell (id) in work with curly- top of sugar beets. All 4 of these diseases are of the so-called physiological type, and the question is still unsettled for the last 3 whether the increased oxidase activity is the cause of the disease or merely the result of disturbances due to the real but at present unknown cause. In the case of diseases whose cause is known the oxidase situa- tion seems to be about the same as for those already mentioned. Reed (29) found that the juice of apples affected with bitter rot (Glomerella cingulata) has greater oxidase activity than that of sound apples. In his previous work the writer (30) found that diseased apple bark shows greater oxidase activity than healthy bark, and is at the same time less acid. This seems to indicate iqiq] rose— blister canker 107 that the oxidizing power of a tissue bears some relation to its acidity, a relation which was rendered more probable by the fact that, according to titration and indicator tests, the acidity rises in the Bunzell apparatus during the course of an experiment at the same time that oxidation gradually decreases and finally ceases. The suggestion was made, therefore, that ''the gradual slowing down of oxidation in the Bunzell apparatus is brought about in part by the accumulation of oxidation products, probably acetic and oxalic acids in the case of pyrogallol, and not by a using up of the oxidase through chemical combination between oxidase and oxidizable substance.'' The validity of this theory in the light of later investi- gation will be discussed in the experimental part of this paper. Experimental OXIDASE ACTIVITY Extracts of fresh bark. — An account will first be given of that part of the work done at the Missouri State Fruit Experiment Station. Extracts of fresh Ben Davis bark were used, prepared as follows: limbs were brought in from the orchard, the bark quickly ground in a meat grinder, and water and toluol added in the propor- tion of 4.25 cc. of toluol for each 100 cc. of water. The mixture was then allowed to extract at 28-30° C. for i hour, with frequent stirring, and filtered through filter paper. The proportions of water and toluol used, assuming that the fresh bark contained 50 per cent water, were such as to make the extracts very nearly equivalent to those prepared for the earlier work (30) with dried bark. All data were corrected to the basis of dry weight deter- mined by weighing and drying samples of the ground bark in triplicate to constant weight in a bath at 95-99° C. Measurement of the amount of oxidation was made by means of the simplified Bunzell apparatus, using i cc. of the extract pre- pared as just described, and either 4 cc. of a i per cent solution of pyrogallol, 0.04 gm. of benzidine, or 2 drops (0.025 gm.) of guaiacol; water was added to make the final volume 6 cc. The various combinations of bark, oxidase reagent, and water were run in duplicate. io8 BOTANICAL GAZETTE [FEBRUARY After the experiment had been set up in the incubator, i hour was allowed for the apparatus and solutions to come to a constant temperature. The manometers were then closed and the solutions mixed. No shaking machine was used, but the apparatus holder was tipped back and forth several times whenever a reading was taken. Allowance for temperature variations was made by run- ning with each experiment a blank containing only water and cor- recting the others by it. Table I gives the results of two representative experiments, showing the amount of oxidation of the 3 different reagents by TABLE I Oxidation of pyrogallol, benzidine, and guaiacol by extracts of healthy AND diseased bark; manometer readings corrected against apparatus containing only water; temperature 28-31° C. Extract of fresh bark Extract of dried BARK Day of Pyrogallol Benzidine Guaiacol Pyrogallol Healthy Diseased Healthy Diseased Healthy Diseased Healthy Diseased Sample 38a Sample 386 Sample 41a Sample 416 Sample j8a Sample 38^ I 2 3 4 5 6 0.0 0.71 1.38 1.70 1.83 0.0 1.62 2.46 2.50 2.50 0.0 0.17 0. 22 0.41 •0.50 0.0 0.82 I . 21 1.76 2.07 0.0 0.02 O.II 0.21 0.27 0.0 0.43 1. 16 1.38 I-5I 0.0 0. 26 0.53 0.77 0.86 1.07 0.0 0-57 0.94 1.20 1.36 I -5° 7 8 9 10 Ratio . . 2.12 2.20 2.36 2.44 I .00 t 2.66 2.73 2.86 2.96 0 I. 21 0.60 0.74 0.77 0.88 1. 00 t 2.31 2.52 2.67 2.93 0 332 0.32 0.32 0-35 0.42 1 .00 t 1.68 1. 91 1.98 2.18 0519 1-51 1. 00 t 1.94 0 1.28 extracts of both healthy and diseased bark. There are included also data from the earlier paper showing the amount of oxidation of pyrogallol by extract of dried bark. The results indicate that for approximately equal amounts of dry matter the dried bark is considerably less active than the fresh (fig. i). The decrease is probably due to the drying; this is shown more definitely by data to be presented later. It is to be noted that the oxidase activity of diseased bark is definitely greater than that of healthy bark, iQig] ROSE— BLISTER CANKER 109 although the ratio between the two is greater where benzidine or guaiacol was used as oxidase reagent than where pyrogallol was used. The writer prefers to follow Bunzell in using the term oxidase activity or oxidizing power rather than ''oxidase." Where the latter term occurs in this paper, it is used only for the sake of brevity, with no intent to imply any fixed notion as to the nature of the agent which brings about the oxidation. Titration and indicator tests on extracts of fresh bark showed the healthy bark to be more acid than the diseased, exactly as had been shown previously in the work with dried bark. No data J) D-- - ../J-- ^_ -'C^ Fig. I. — Oxidation of pyrogallol, guaiacol, and benzidine by extract of fresh bark, healthy and diseased, and extract of dried bark, healthy and diseased: A, pyrogallol and fresh health)^ bark; B, pyrogallol and fresh diseased bark; C, ben- zidine and fresh healthy bark; D, benzidine and fresh diseased bark; E, guaiacol and fresh healthy bark; F, guaiacol and fresh diseased bark; G, pyrogallol and dried healthy bark; H, pyrogallol and dried diseased bark; // = healthy, Z) = diseased. are given, since the true condition, at least for dried bark, was determined more accurately by means of a potentiometer. Extracts of dried bark. — For the work at the University of Chicago bark was used which had been dried at 35-40° C. for 2-3 hours, ground fine enough to go through a 40-mesh sieve, and stored air dry in zinc-capped Mason jars. A few of the experiments were run with oxidases precipitated from an extract of this bark powder, but in most of them the powder itself was used, o.io gm. in each apparatus. The reagents tested were pyrogallol and pyro- catechin, 4 cc. of a i per cent solution; benzidine 0.05 gm.; no BOTANICAL GAZETTE [February guaiacol 2 drops (0.025 gm.). Tests for any given set of condi- tions were always run in duplicate, sometimes in triplicate, or even quadruplicate. All experiments were shaken for 3 hours at the rate of 106 complete excursions per minute in a constant tempera- ture chamber provided with a fan driven from the outside, and then allowed to stand for 10-90 hours. Temperature variations were rarely greater than 0.5° during the shaking period, but some- times amounted to as much as 1.0° afterward, owing to less perfect control when the machinery was not in motion. Corrections for temperature variations were made as before by comparison with a blank containing only water. Potentiometer measurements were made with a hydrogen electrode like that described by Bo vie (8), streaming hydrogen, 3 resistance boxes as described by Michaelis (25, p. 131), a saturated calomel electrode, a normal element checked against another which had been calibrated by the United States Bureau of Standards, and a Leeds and Northrup dead-beat galvanometer. Hydrogen of high purity from a tank of the compressed gas was run through an electrically heated combustion tube containing platinized asbestos and then through the hydrogen electrode tube. The latter, together with the capillary from the calomel electrode, projected through a rubber stopper into the vessel containing the solution to be tested. Escape of hydrogen was provided for by a third opening in the stopper. An error was undoubtedly intro- duced here, due to displacement of CO2 from the solution, in cases where the hydrogen ion concentration was less than io~s (Michaelis, pp. 142-144), but since the only solutions showing this slight degree of acidity were mixtures of bark, water, and pyro- gallol for determination of hydrogen ion concentration before any oxidation had taken place, and since all others were found to be more acid, the error is probably negligible. It could have been avoided entirely by using a Hasselbalch shaking electrode had it and the time for using it been available. Among the first experiments run was one designed to test fully the oxidase activity of healthy and diseased bark when pyrogallol was used as the oxidizing substance. The results, given in table II, are the average of 5 closely agreeing determinations. These results I9I9] ROSE— BLISTER CANKER III agree well with those obtained without a shaking machine in show- ing considerably greater oxidation by diseased than by healthy bark. The ratio between the two, i .00:2.19, is larger than that found previously (1.00:1.28), the difference probably being due to dift'erences in drying or possibly to the shaking itself. T.\BLE II 0.\ID.\TION OF PYROG.4LLOL BY HE.\LTHY AND DISEASED APPLE BARK; S.4MPLES 3 AND 4; TEMPERATURE 27 0=^1.7° C. Time of READING Manometer readings, ex- pressed IN CM. OF MERCURY, CORRECTED AGAINST BLANK CONTAINING ONLY WATER Time of READING Manometer readings, ex- pressed IN CM. OF MERCURY, corrected against blank containing only water Healthy Diseased Healthy Diseased March 19 2:45 P-M 3:00 3:15 3:30 3:45 4:00 4:15 4:30 0.0 0.0 O.IO 0. 16 0.23 0.31 0.41 0.45 0.0 0.23 0.48 0.65 0.80 0.92 I 05 I. II , March 19 4:45P-M.... 5:00 5:15 5:30 5:55 March 20 8:30 A.M... . 0.48 0-53 0.61 0.64 0-59 1 .10 1-25 ^■35 1-45 1-57 1.49 2.41 I In table III are summarized the results of an experiment to test the oxidizing power of both diseased and healthy bark on pyrocatechin, guaiacol, and benzidine. A comparison of the figures in table III with those in tables I and II shows that diseased bark causes greater oxidation of pyro- gallol, pyrocatechin, benzidine, and guaiacol than does healthy bark, and that both tissues cause greater oxidation of the first two reagents than of the last two. It is further shown by tables I and III that the amount of oxidation increases slowly for several days; in fact table III shows that it is practically doubled for all the combinations, except those containing pyrocatechin, during the 64-hour period following the 3 hours' shaking. This fact of an increase of oxidation on standing was observed to a greater or less degree with most of the bark material used in this work, and is in direct contradiction to Bunzell's explicit and repeated statement that oxidation in his apparatus comes to a definite end after 3 or 4 hours' shaking. The only exceptions the writer has 112 BOTANICAL GAZETTE [FEBRUARY noted were in those cases where the bark powder showed low oxidase activity to begin with, possibly due to injury of the "oxi- dase" during drying. TABLE III Oxidation of pyrocatechin, guaiacol, and benzidine by healthy and DISEASED bark; TEMPERATURE 29.4-29. 7° C. Healthy Diseased Ttmf. of reading Benzidine Guaiacol Pyrocate- chin Benzidine Guaiacol Pyrocate- chin June 8, i:30P.M 4:30 after shaking 3 hours . . June 9, 8:10 A.M. . . 2: 20 P.M. . . " 10, 9:15 A.M. . . " 11, 8: 20 A.M. . . 0.0 0.08 0.25 0.38 0.40 0.65 0.0 0.33 0.35 0.48 0-55 0.65 0.0 I 13 1-45 1.65 1.85 2.12 0.0 0.65 0.80 0.98 1 . 27 1-45 0.0 0.75 I .00 1,07 1.20 1-47 0.0 3-77 4-35 4-55 4.87 512 That the rate and temperature of drying have an efifect on the oxidase activity as well as on the hydrogen ion concentration is clearly shown in table IV. TABLE IV Effect of rate and temperature of drying upon oxidase activity and hydrogen ion concentration of healthy and diseased apple bark Oxidation Initial Temperature and duration OF drying Degree of browning Sample After shaking 3 hours After stand- ing 10 hours After stand- ing 15 hours 4 diseased . . 6 " 1.49 2.25 1.58 0-59 1.07 0.35 0.62 2.33 2.42 1 .60 0.80 I. 12 0-35 0.72 2.78 5. 61* 5-45 S-i6 5-15 5 04 5.00 4.80 40°, 2 hours 40°, 2 40°, 4 40°, 2 40°, 2 35°, 4 Slight Slight 2 " Much 3 healthy. . . 5 " ••■ 5a " ... I " 1.23 Very little Slight Very little Much *This figure is the negative logarithmic exponent of 10 where the whole expression 10=5' is a measure of the hydrogen ion concentration in the solution. The larger it is, therefore, the smaller the hydrogen ion concentration it expresses. In this particular case it can be written 2.454X10 "> (6.00 — S .61 =0.30. Antilog 0.39 = 2.454). In the amplified form this becomes 0.000002454 (normal). Samples i, 2, 5, 5a, and 6 were all run in one experiment. Oxidation data for samples 3 and 4 are taken from table II and from another experiment not recorded in this paper. Samples 5 and 5a were parts of the same lot of ground bark but received I9I9] ROSE— BLISTER CANKER 113 different treatments as shown. The results show that oxidase activity is much reduced by drying at 35-40° for 4 hours (sample i, healthy; sample 2, diseased), or at 50° for 2 hours (sample 5a, healthy). Hydrogen ion concentration.— Hydrogen ion determinations on mixtures of bark and water and of bark, water, and pyrogallol, used in the same proportions as in the oxidase apparatus, showed that pyrogallol has no effect on the reaction. It was found pos- sible to get constant initial readings on all mixtures containing healtTiy bark and pyrogallol in 30-45 minutes; the same period sufficed for mixtures containing diseased bark and pyrogallol after they had been shaken in the oxidase apparatus, but not for similar mixtures freshly made up and not shaken. In these cases the potential increased slowly for an hour or two from about Ph = 5 . 60 to Ph = 5.40, but never reached the figure given by healthy bark. Culpepper, Foster, and Caldwell (16), working with normal and diseased Red Astrachan apples, state that when titrations were made on fruit pulp suspended in water " the diffusion of acids out of the tissues continues for many hours and at slower rates in diseased than in normal fruits," but in the light of the following results the writer is inclined to think this increase of acidity was due to oxidation going on in the solutions, and not to diffusion of acids out from the tissues. TABLE V Correlation between oxidase activity and htorogen ion concentration OF MIXTURES CONTAINING PYROGALLOL, WATER, AND EITHER HEALTHY OR DISEASED BARK; TEMPERATURE 29-30.5° C. Stage of experiment Healthy Diseased Oxidation P„ 1 a. Oxidation Ph Before shaking 0.00 0.82 1 . 10 2.00 2.90 5-iS 0.00 2.28 2-59 4.07 4.96 S.61 After shaking 3 hours After standing 15 hours. . . . " 48 " .... 4.82 4.89 " 64 " .... 4.29 4.29 Increase in hydrogen ion concentration during oxida- tion.— Experiments designed to test more fully the theory that oxidation causes an increase in acidity are summarized in table V. 114 BOTANICAL GAZETTE [FEBRUARY It is clear from table V that oxidation in these mixtures is accompanied by a marked increase in hydrogen ion concentration, and the conclusion certainly seems justified that there is a causal relation between the two. It is also seen that when oxidation comes to an end, both mixtures have the same reaction, Ph = 4 • 29, a condition suggesting that at this point the hydrogen ion is the limiting factor. BuNZELL (12) and Reed (28) have studied the effect of hydrogen ion concentration on oxidation, but apparently neither of them has realized that it might increase during the oxidation process (30). They apparently assume that the hydrogen ion concentration established at the beginning of an experiment remains constant until the end, whereas the results given show that in these cases it increased as long as the oxidation continued. In order to discover, if possible, what relation exists between oxidation and hydrogen ion concentration in the oxidase apparatus, further experiments were tried with mixtures of bark, dry pyro- gallol, and, instead of water, 5 cc. of buffer solutions containing various amounts of N/io sodium acetate and either N/io or N/ioo acetic acid. The initial reactions of these mixtures (before shaking) and of the buffers alone are given in table VI and shown graphically in fig. 2. ^ ^ ^ TABLE VI Reaction, Ph, of buffer solutions and mixtures of buffer solutions, bark, and pyrogallol Solution I 2 3 4 5 6 7 8 Q Buffer alone Buffer and healthy bark and pyro- gallol 6.02 5-59 5.76 5-73 5-52 5-7° 5-41 5-36 5-50 5-17 S-I5 5-31 4.80 4-85 5.00 4-53 4.58 4.61 4.21 4.24 4-39 3-90 398 4.08 3.61 3.61 Buffer and diseased bark and pyro- gallol 3-73 Graphs B and C in fig. 2 show that while diseased bark absorbs H+ ions to about the same extent as the healthy, the latter absorbs more 0H~ ions; that is, its titration acidity is greater, which is exactly the condition found by titration with N/20 sodium hydroxide (30). The Ph values at points where B and C cross A iQig] ROSE— BLISTER CANKER "5 \|CC^1) ACETJC Arm TO 1 CC"^^ SOD, ACRTATF. ^10.24 5.12 2.56 1.28 0. 0.32 Q.I6 0.08 0.04 Fig. 2.— Pjj of mixtures of bark, pyrogallol, and various buffer solutions before and after oxidation had ceased: A, P^ of buffer solutions; B, Pjj of mixtures of buffer solutions, pyrogallol, and diseased bark before oxidation; C, P^ of mixtures of buffer solutions, pyrogallol, and healthy bark before oxidation; D, P^ of mixtures of buffer solutions, pyrogallol, and diseased bark after oxidation; E, Pjj of mixtures of buffer solutions, pyrogallol, and healthy bark after oxidation. *Onry acetic acid used here. ii6 BOTANICAL GAZETTE [FEBRUARY (healthy bark about 5.10, diseased about 5.65) agree well with those determined without the buffer (Ph healthy = 5 .15, diseased = 5.61); the latter are taken, therefore, to represent practically the actual acidity in each case. This is based on the assumption that if the acidity of a buffer solution is the same as that of a mixture of bark, pyrogallol, and water, no change in acidity will take place when the buffer is used instead of water. Effect of buffer solutions. — The oxidations brought about by mixtures of bark, pyrogallol, and the various buffer solutions are given in table VII, together with the initial Ph of these mixtures and their Ph after oxidation had practically ceased. TABLE VII Oxidation by mixtures of bark and pyrogallol with various buffer solutions; temperature 29-3q°c. Buffer Healthy Diseased SOLUTION Oxidation Initial P^ Final Pg Oxidation Initial Pg Final Pg I ? CO 4.68 4.78 5 76 485 485 2 2 1.68 5 5 5 4 4 4 3 3 5 52 36 15 85 58 24 98 61 IS 458 5 5 5 5 4 4 4 3 5 70 50 31 00 61 39 08 73 61 4 1; . . ... 2.15 4-34 4 36 4-75 6 7 8 1-95 1.80 1-55 0-53 2.22 398 3.6s 3.56 3-35 4.29 4.48 4.12 4.60 4-25 9 1.82 4.27 3.68 Check 4.29 The principal fact shown by the results in table VII is that the Ph (4 . 29) reached by mixtures of pyrogallol, water, and either healthy or diseased bark when oxidation comes to an end is not sufficient to inhibit oxidation when the mixture has that Ph value to begin with; in fact, a greater degree of acidity does not inhibit entirely, since a healthy bark mixture with an initial Ph of 3.61 gave an oxidation (a mercury rise) of 0.53 cm., and a diseased bark mixture with an initial Ph of 3 . 78 gave an oxidation of 1.82 cm. The check, bark, pyrogallol, and water gave, in the former case, 2.22 cm. mercury rise, and in the latter 4.27 cm. It might seem from this that the acidity brought about in mixtures of bark, pyrogallol, and water is not the factor which iQig] ROSE— BLISTER CANKER II7 brings oxidation to an end. It seems more reasonable to suppose, however, that the time factor is of importance here; that is, that an acidity of Ph = 4.29 is more effective when reached gradually than when established as a starting point. Looking at the situa- tion from another angle, we may say that inhibition is total if the initial hydrogen ion concentration is high enough, but will be only partial if the concentration is lower; but since partial inhibition means some oxidation, which in itself increases acidity, the process in time necessarily comes to an end. The hydrogen ion concen- tration at that point will depend on what it was in the beginning, but will never be equal to that which causes total inhibition. That this theory fits the facts is shown by table VII. Oxidation took place in all the mixtures, the amount depending on the initial hydrogen ion concentration, except where diseased bark was used with buffer no. 4. Acidity increased in all the mixtures but one, diseased bark with buffer no. 6 (see tables VI and VII). The increase in acidity is shown graphically in fig. 2. It is unexpectedly small for diseased bark except where the 3 most alkaline buffers were used, a condition which suggests the need of further experi- ments. In figs. 3 and 4 are shown graphically the oxidation data given in table VII, representing the final amounts of oxidation for each set of tests (healthy and diseased bark with the different buffer solu- tions) . In addition there are shown graphs for several earlier stages in each experiment. These graphs show that below i X io~^ (Ph = 4) for healthy bark, and 2.5X10"^' (Pa = 4. 39) for diseased bark, oxidation drops rapidly as acidity increases. Above these points the changes are not so marked. The hydrogen ion concentration for total inhibition, estimated by extrapolation to the base line, lies between 3.55 and 3.80X10"-* for healthy and between 3.55 and 4. 27X10"^ for diseased bark. All these figures closely approxi- mate those found by Bunzell (12) for potato oxidase, 2 . i- 2.8Xio~'*, and by Reed (28) for apple oxidase, 5.0-7.0X10"^ The results given in table VII show that hydrogen ion con- centration is not the only factor effective in controlling oxidation in the apparatus, and consequently that the lower hydrogen ion concentration of diseased bark cannot account entirely for its ii8 BOTANICAL GAZETTE [FEBRUARY greater oxidizing power. For example, when both kinds of bark were brought to approximately the same hydrogen ion concentra- tion by buffer no. 6, the final amount of oxidation (mercury rise) for healthy bark was i .95 and for diseased 4 .48, the final Ph 3 .98 and 4 . 60 respectively. The total oxidase activity of the diseased plant is the joint oxidase activity of the host and parasite, while Fig. 3. — Oxidation of mixtures of healthy bark, pyrogallol, and various buffer solutions: A, after 3 hours; B, after 22 hours (19 hours without shaking); C, after 29 hours; D, after 48 hours; A, bark, pyrogallol, and buffer solutions as indicated by numbers; B, initial P^; points of plotting marked by vertical broken lines. the oxidase activity of the healthy plant is that of the host alone. This may account in part for the difference both in rate of activity and in the Ph concentration at the time the action ceases. Nature of equilibrium reached. — Bunzell (13), in experi- ments with potato peel powder, has obtained what he considers evidence that "the activity of the plant powder is not paralyzed by the products formed in the course of the reaction." He found iqiq] ROSEr-BLISTER CANKER 119 -^ 9 7 6 4 2 I 5 3.73 4.39 4.61 5.31 5.70 5.76 j:^c 1 / "~ / y B jr~ /» E / • 0 1 / c / 1 v> / 0 1/ / / 1 11 • • III 1 Ph, 1 Fig. 4. — Oxidation by mixtures of diseased bark, pyrogallol, and various buffer solutions: A, after 23 hours (shaken 2 hours of this time) ; B, after 45. 5 hours; C, after 69.5 hours; A and B as in fig. 3. 120 BOTANICAL GAZETTE [FEBRUARY that by adding a second portion of the powder to the apparatus in which oxidation by the first portion had ceased he could cause a further increase in oxidation, the amount of increase varying with the oxidase reagent used. The writer has found a similar increase in oxidation when more oxidase reagent is added, after oxidation ceases. . The results of an experiment of this kind are summarized in table VIII. Results are given beginning with the TABLE VIII Summary of results from an experiment to test effect of adding fresh supply of oxidase reagent. Experiment Effect of adding 7 and 4 drops i per cent benzidine to apparatus 10 and 11 on ninth day, 8 and 9 as checks Effect of adding 10 drops i per cent ben- zidine to apparatus 10 and 11 on eleventh daj^, 8 and 9 as checks Total effect of i per cent benzidine, 8 as check Effect of adding 0.06 gm. of pyrogallol to 8 on twenty-sixth day 8 as check Effect of adding o . 06 gm. benzidine to 9 on fourteenth day, 8 as check Effect of adding 10 drops absolute alcohol to 9 on twenty-first day, 8 as check Stage of experi- ment Increase in oxidation (cm. of mercury rise) 9th to nth 0.00 nth to 14th 0.17 9th to 26th 0.47 j 26th to 41st I 1-53 9th to 26th I 0.47 14th to 2ist i 0.19 2ISt to 26th o.n 0.02 0.23 0.74 0.68 0.49 1. 16 1-95 o. 19 0.71 1-34 ninth day of the experiment. Up to that time oxidation in all 4 of the tubes was practically the same, the average being 3.12 (cm. of mercury rise): Alcohol was used at the beginning of the experiment to discover whether it has an inhibiting effect on oxi- dation, and later, when solid benzidine was added, to bring the benzidine into solution more rapidly. The results show that, in the small quantities used, the alcohol had no inhibiting effect (table VIII, ninth day, apparatus 10 and 11), and probably did bring the benzidine into solution (twenty-first-forty-first day, apparatus 9). The most important fact shown by these results is that after oxidation had practically ended, the addition of more oxidase iqiq] rose— blister CANKER 12 1 reagent was followed by a marked increase in oxidation. For example, in table VIII it is seen that from the ninth to the twenty- sixth day oxidation in apparatus 8, containing pyrogallol and bark extract, showed an increase of only o .47 cm., while tubes 10 and 11, also containing pyrogallol and bark extract to which benzidine solution was added later, showed an increase of i .95 and i .34 cm. respectively. Equally marked excess over the check was obtained when solid pyrogallol or soHd benzidine was added. One might infer that the oxygen admitted, when the tubes were opened to introduce reagents, increased oxidation, but this effect could hardly account for the difference observed. Bunzell states that exhaus- tion of oxygen is not the hmiting factor, and experiments by the writer have shown that, when a fresh oxygen supply is allowed to enter the apparatus, the subsequent increase in oxidation is small. The fact that after oxidation ends it can be started afresh by the addition of fresh plant material or of fresh oxidase reagent suggests that the equilibrium reached is a false one, like the third case described by Hober (23, p. 671), in which a reaction product of the catalytic reaction brings about equiUbrium by an inactiva- tion of the catalyzer. A test for this condition according to Hober is that reaction begins again when more catalyzer is added, as in the case of the hydrolysis of amygdalin by emulsin. The similarity between the two reactions, however, does not prove that the oxida- tion catalyst is an enzyme, for it may be non-enzymic in nature and still be inactivated by the products of the catalytic reaction. An idea of the nature of the oxidase reaction was obtained by testing some of the data by the formula for unimolecular reaction, k= log. . In these calculations the total amount of oxida- / a—x tion (mercury rise) at the end of the shaking period was assumed for the value of a, and the amount of oxidation at the end of each 15-minute interval for the value of x. The figures which should be used, of course, are the total amount of pyrogallol at the begin- ning of the experiment and the amount oxidized at the end of each 15-minute interval, but such figures would be difficult to obtain. The writer sees no reason why the values used for a and x do not truly represent the course of the reduction. 122 BOTANICAL GAZETTE [FEBRUARY In most cases the values of k given in these tables are fairly constant and may be considered a strong indication that the oxidase reaction is unimolecular. In table XI, column 3, table XII, TABLE IX Healthy bark and pyrogallol t (min.) X (mercury rise in cm.) a—x , I , 0 *=7 log. / a—x IC 0.14 0.19 0. 24 0.34 0.44 0.49 0.54 0. 63 0.72 0.67 0.62 0.52 0.42 0.37 0.32 0. 2? *. 0.00514 0.00361 0.00315 0.00365 0.00415 0.00407 0 . 00409 0.00477 0.00503 ,0.00570 0 00660 ao AS. 60 • 7c no lOi? 120 135 I CO 0.68 0.18 0. 74 1 0.12 les 0.79 0.86 0.07 180 Mean. . . . a.ooAZz * Brackets in this and following tables indicate those values of k which were considered in calculating the mean. TABLE X Diseased bark and pyrogallol t (min.) X (mercury rise in cm.) a—x , I , a k=- log. t a—x IC 0-13 0.30 0.50 0.65 0.72 0.85 0.99 1.04 i-iS 1. 18 1-25 1.38 1-25 1.08 0.88 0.73 0.66 0.53 0.39 0.34 0.23 0.20 0.13 0.00286 20 0.00355 0.00434 0.00461 0.00427 4c 60 7c QO 0.00461 0.00522 0.00507 0.00576 .0-00559 0.00621 roe, 120 ISC ICO 16"; 180 Mean .... 0 . 00478 column I, and table XIV, column i, the values for k show a gradual increase throughout the experiment, and can scarcely be taken to indicate a unimolecular reaction. Table XII, column i, how- ever, is checked by tables IX and XII, column 3, the mean value igig] ROSE— BLISTER CANKER 123 of k being nearly the same in all 3 cases, although it is doubtful whether a mean for table XII, column i, is really significant. TABLE XI Values of k calculated from data obtained in experi- ments WITH apple bark, K2CO3, AND PYROGALLOL k t (min) KaCOi and pyrogallol K.COj and pyrogallol Healthy bark, K,C0,. and pyrogallol Diseased bark, KCaOj, and pyrogallol IS 30 45 60 75 90 lOS 120 135 150 165 180 ■ 0.00747 0.00776 0.00774 0.00785 0 . 00808 0.00767 0.00765 0 . 00786 O.O0811 0.00768 0 . 00805 < 0,00552 0 . 00803 0.00773 0.00786 0.00819 0.00750 0.00709 0 . 00890 0.00941 0.00947 0.00936 . 0,00114 0 . 00380 0 . 0043 I 0.00368 0,00472 0 . 0049 2 0.00514 0.00470 .0.00633 0 . 00663 0.00692 0.00525 0 . 00600 0.00615 0 . 00604 0 . 00606 0.00628 0.00642 0.00657 0 . 00700 0.00678 0.00862 Mean . . 0.00781 0.00834 0.00482 0.00635 TABLE XII Values of k calcul.^ted from d.a,ta obtained in experi- ments WITH APPLE B.\RK, PYROG.AXLOL, AND PYROCATECHIN / (min.) IS- 30. 45- 60. 75- 90. 105. 120. 135- 150. 165. 180. Healthy bark and pyrogallol 0.00246 0.00277 0.00322 0.00383 0.00493 O . 00460 0.00501 0.00584 0.00886 Diseased bark and pyrogallol 0.00458 0.00528 00515 00515 00510 00530 00507 00575 00604 00744 Healthy bark and pyrocat- echin 0.00502 0.00429 0.00357 O . 00346 0.00416 0.00433 0.00426 0.00434 o . 00483 o . 00548 0.00590 Diseased bark and pyrocat- echin O . 00483 O . 00494 0.00510 O . 00488 0.00526 0.00536 0.00553 0.00584 0.00613 o . 00690 0.00866 Mean . o . 00430 0.00527 o . 0045 I 0.00521 Confirmation of the results with apple bark is found in table XIII and table XIV, column 2, based on data obtained by 124 BOTANICAL GAZETTE [FEBRUARY BuNZELL (9, 13) with tuKp tree leaves and with potatoes, although the mean value of k in all 3 cases is much larger than that found for bark. Attention has already been called to the fact that the data in table XIV, column i (also from Bunzell's work), fail to fit the equation for a unimolecular reaction. The fact of a marked rise TABLE XIII Values of k calculated from data published by BUNZELL (9) FOR POTATO JUICE AND PYROGALLOL / (min) 10. 20. 30. 40. 50. 60. 70. 80. 90. Mean. k* 0.0315 0.0266 10 30 0.0240 0.0216 45 60 0.0244 75 0.0277 90 0.0255 0.0283 105 0.0262 / (min) Mean . kf 0.0246 0.0277 0.0199 0.0168 0.0174 0.0233 0.0208 * 23i p. 29. table VII, columns i and 4. t 23, p. 26, table II, columns 5 and 7. TABLE XIV Values of k calculated from data published by Bunzell (13) k t (min.) k i (min.) Spinach leaves and para-cresol Tulip tree leaves and phlorhizin Spinach leaves and para-cresol Tulip tree leaves and phlorhizin 15 30 45 60 75 0.00374 0 . 00654 0 . 00640 0 . 00940 0.01092 0.0124 0.0119 0.0133 0.0137 90 IOC O.OIO18 120 IZK Mean. . O.OIS7 in the value of k toward the end of the experiments with bark may mean that at that point the "oxidase" oxidizes not constant frac- tions but constant weights of pyrogallol in a given time (Philip 27, p. 295). The data at hand, however, are insufficient for a veri- fication of this hypothesis. I9I9] ROSE— BLISTER CANKER 125 A unimolecular reaction is one in which the concentration of only one substance is changed. If oxidation of pyrogallol by plant material in the oxidase apparatus be such a reaction, the substance whose concentration is changed is pyrogallol. The "oxidase'' then appears as the catalyst, its concentration remaining unchanged during the course of the reaction. Even at that it is not neces- sarily proved to be an enzyme, since the linear relationship between time and amount of change is also shown in the oxidation of pyro- gallol by potassium carbonate. Effect of adding protective colloids. — Bayliss (6) and Perrin (26) have suggested that the oxidizing enzyme is an active form of the colloidal hydroxide of manganese, iron, or copper, kept in this active state by an emulsion colloid such as gum or albumin, acting as a protective colloid. Tables XV and XVI show the effects of additions of gelatine and gum arabic. Table XV shows that o . 2 per cent gelatine increases considerably the oxida- tion by healthy bark and only sHghtly that by diseased bark. Three other experiments with pyrogallol and 2 with pyrocatechin with o . 2 per cent gelatine added showed similar results. The use of o . 8 per cent gelatine with pyrogallol also showed a similar effect. Both 0.2 and 0.8 per cent gum arabic had little or no effect on healthy bark and a slight accelerating effect on diseased bark. TABLE XV ' Effect of o . 2 per cent gelatine on oxid.ation of pyrogallol by healthy and diseased b.ark; temperature 22-24° c. Time of re.^uing He.althy Diseased Without gelatine With gelatine Without gelatine With gelatine May 18, 7:0^ P.M 0.0 0.77 0.94 1-37 1.65 0.0 0.81 1.32 2.26 2.74 0.0 2. 22 2.40 2.88 317 0.0 10:03 P-M. after shaking 3 hours. . " 19, 8:10 A.M " 20, 0: T,K A.M 2.24 2-54 3.00 3-27 " 21, 8:15 A.M Since gelatine is amphoteric, one might infer that it or its splitting products act as buffers, thus reducing the rate of increase of the hydrogen ion concentration with progress of the oxidation 126 BOTANICAL GAZETTE [FEBRUARY (fig. 5). Table XVII, however, shows that gelatine has little effect on the hydrogen ion concentration of oxidizing mixtures of either healthy or diseased bark. Precipitated oxidases. — Experiments were run using pre- cipitated "oxidases," prepared as follows: 2 gm. of bark were allowed to extract with 10 cc. of water and 5 drops of toluol for I hour; the extract was then squeezed through moist cheesecloth on to coarse filter paper, the beaker washed with five i cc. portions of water and the filter paper finally with two more; 50 cc. of 95 per £ 0 ^ 3 to >- >>■ 0 1 rZ---'''''^ ^ — ^ _^ c 2 1 ^ \\C^ — Time in 1 - r 1 i 1 I 1 1 2 3 20 44 Fig. 5. — Effect of 0.8 per cent gum arable and 0.8 per cent gelatine on oxidation of pyrogallol by healthy and diseased bark: A, healthy bark; B, healthy bark and gum arable; C, healthy bark and gelatine; D, diseased bark; E, diseased bark and gum arable; F, diseased bark and gelatine. cent alcohol were then added to the filtrate (concentration of alcohol about 70 per cent), the whole allowed to stand for 10 minutes and the flocculent precipitate collected on a hard filter by gentle suction with a filter pump; 150 cc. more alcohol were then added to the filtrate (concentration of alcohol now about 90 per cent) and the whole allowed to stand for i hour, since precipitation was slow, before collecting this second fraction on the filter with the first. The precipitate from diseased bark was much browner than that from healthy bark. Whether this bears any relation to its greater oxidase activity is not known. 1919] ROSE— BLISTER CANKER 127 For tests in the oxidase apparatus the combined precipitates were dissolved in 20 cc. of water, and 2 cc. of this solution contain- ing the precipitate obtained from o . i gm. of bark was put in each apparatus together with the usual amounts of pyrogallol and water. TABLE XVI Effect of o. 2 per cent gum arabic, o . 8 per cent gum ar.^bic, and o . 8 per cent gel.atine on oxidation of pyrogallol by healthy and DISEASED bark; TEMPERATURE 21-23° C. Time of reading Healthy No addition Gelatine o . 8 per cent Gum arabic 0.2 per cent o . 8 per cent Diseased No addition Gelatine 0.8 per cent Gum arabic o . 2 per cent 0.8 per cent At beginning. . . . After shaking 3 hours After 18.5 hours After 42 hours. . Average of 0.00 o. 70 I 03 I-50 2 0.00 0.78 1-34 2. 20 0.00 0.65 1 .00 0.00 o. 71 1 .01 1-54 0.00 1.88 2.36 2.73 2 0.00 2.23 2.69 3.18 0.00 2.19 2.71 0.00 2.05 2.46 2-95 In table XVIII are given results showing the oxidizing power of these solutions, with and without gelatine (fig. 6). The relation observed with bark powder still holds here, that diseased material is more active than healthy. On the other hand, gelatine increases oxidation by the precipitate from extract of TABLE XVII Reaction of mixtures of bark and pyrogallol vmn gelatine (o . 2 per cent) .\XD WITHOUT at VARIOUS STAGES OF OXIDATION PROCESS Time of reading • Healthy Diseased Without gelatine With gelatine Without gelatine With gelatine Initial P ] After 15 hours ° [After 64 hours 5IS 4.82 .4-29 SI5 4.84 4-35 5.61 4.89 4.29 5.60 4.86 452 diseased bark, but is without marked effect on that from healthy bark, the reverse of the condition found when bark powder was used. There were indications in the preliminary work that the alco- hoHc precipitate from bark extract was easily separated into 2 128 BOTANICAL GAZETTE [FEBRUARY fractions, hence it seemed worth while to collect these separately. This was done for both healthy and diseased tissue and gave TABLE XVIII Oxidation of pyrogallol by aqueous solutions of precipitated oxidase FROM healthy AND DISEASED BARK, WITH TUSJD WITHOUT gelatine; -temperature 29.3-30.3° C. Healthy Diseased Time of reading Without gelatine With gelatine Without gelatine With gelatine Tune 28 A' i=c p M 0.0 0-31 0-35 0.42 0.0 0.33 0.46 0.46 0.0 0.68 0.0 " 20. 8:4.=; A.M 0.72 " " 11:45 A.M. after shaking 3 hours . . . " ^0. 8: •^o A.M 1. 01 ] 1.24 I. 08 ; 1.^6 Fig. 6. — Oxidation of pyrogallol by precipitated oxidases from healthy and diseased bark, with and without gelatine, shaken only during period from A to B: A, precipitate from healthy bark without gelatine; B, precipitate from healthy bark with gelatine; C, precipitate from diseased bark without gelatine; D, precipitate from diseased bark with gelatine. precipitates whose air dry weights, determined by the use of tared filters, were as follows: From extract of From extract of healthy bark . | diseased bark Fraction i o.oogo gm. o.o'5^2gm. Fraction 2 0.0080 1 0.0164 Total 0.0179 0.0696 The greater amount of precipitate from diseased bark may or may not be directly connected with its greater oxidase activity. iqiq] ROSE— BLISTER CANKER 129 Further study is necessary to show the facts. A test of these pre- cipitates with pyrocatechin showed that while the 2 fractions from healthy bark are about equal in oxidizing power the first fraction from diseased bark is 11 times as active as the second (fig. 7). Fig. 7. — Oxidation of pyrocatechin by precipitated oxidases from healthy bark, without gelatine: A, fraction i; B, fraction 2; C, fractions i and 2 tested together; D, sum of fractions i and 2 tested separately. Other precipitates were prepared using 25 cc. of alcohol for the first fraction and 100 cc. more for the second. The oxidase activity of these, tested separately and combined, with and without gelatine, is shown in table XIX. TABLE XIX 0XID.\SE ACTIVITY OF FIRST AND SECOND FRACTIONS FROM B-^RK EXTRACT TESTED SEPARATELY AND COMBINED; TEMPERATURE 29.5-30.0° C. Without gelatine With gelatine Bark extr.act Sum of fractions I and 2 tested separately Fractions i and 2 combined Sum of fractions i7,„^,.- „„ , ^„a -, _ J i »_j iT Tactions I and 2 ' separVte^ --^"-^ Healthy, after 23 hours. . . . Diseased, "38 " . . . . 0.65 1.82 0.84 2.00 0.76 0.84 2.64 ; 3.06 1 The mechanism by which gelatine increases the oxidase activity is not clear. It is evidently not through buffer action, as shown by its lack of effect on the hydrogen ion concentration (table XVII, figs. 8, 9, 10). Special tests showed that there was no hydrolysis of the gelatine to amino acids, in either healthy or diseased bark, which would increase its buffer effect. If gelatine is effective through its action as a protective colloid, its effect in this direction must be very complex, as shown by its difference in effect on bark mixtures and precipitated oxidases. I30 BOTANICAL GAZETTE [FEBRUARY 1 2 3 23 Time in hours Fig. 8. — Oxidation of pyrocatechin by first and second fractions of healthy bark, with gelatine (for explanation of lettering see legend for fig. 7). 12 3 16 40 Fig. 9. — Oxidation of pyrocatechin by first and second fractions from diseased bark, without gelatine (for explanation of lettering see legend for fig. 7); points of plotting marked by vertical broken lines. Fig. 10.^ — Oxidation of pyrocatechin by first and second fractions from diseased bark, with gelatine (for explanation of lettering see legend for fig. 7); points of plotting as in fig. g. iqiq] rose— blister CANKER 131 The difference in the effect of gelatine and of gum arabic on oxidation by healthy bark may depend on differences in the col- loidal solutions they form. An artificial oxidase prepared by Dony-Henault (18) from manganese formate, sodium bicarbonate, and gum arabic could be destroyed by heat; but one prepared by Trillat (33) from albumin and manganese could not be so destroyed. Bayliss (6, p. 585) thinks the difference here "clearly depends on the nature of the emulsion colloid in association with the metal." On the other hand, what little increase in oxidation gum arabic produces may be due to an oxidase naturally present in it (BouRQUELOT 7), although an experiment designed to test this question gave negative results. One per cent gum arabic plus I per cent pyrogallol, and pyrogallol alone, were placed in separate oxidase tubes and shaken twice during each 24 hours. At the end of 3 days the mercury rise was 0.32 cm. in the first case and o . 20 cm. in the second, a difference almost within the limits of error in reading the manometers. The data given in table XIX show that when the precipitate is collected in 2 fractions, these fractions have a greater oxidase activity if combined than if used separately. This condition seems to be about the same as that described by Bach and Chodat (4) for Lactarius vellereus. They found that by the fractional pre- cipitation of an aqueous solution of the oxidase of this fungus, by alcohol, 2 fractions could be obtained possessing markedly different properties. The first of these was almost insoluble in 40 per cent alcohol and had the properties of a weak oxidase; the second was soluble in 40 per cent alcohol but insoluble in pure alcohol and had no oxidizing powers. This fraction, however, was found to impart greater activity to hydrogen peroxide as an oxidizing agent; it was also found to increase markedly the oxidizing powers of the first fraction. The chief difference between this situation and that found in the work with apple bark is that in the latter case the first fraction has more than a weak oxidase activity, while the second, possibly because of incomplete separation of the fractions, is not entirely without it. No tests have been made of the behavior of the second fraction toward hydrogen peroxide. 132 BOTANICAL GAZETTE [February Oxidase activity of the fungus in pure culture.— A fungus powder was prepared according to the method employed by Reed (29) from mats of Nummularia mycelium grown in the potato extract medium described by Duggar (19). A test with 3 Bunzell tubes using o . i gm. of fungus powder, 4 cc. of i per cent pyrogallol, and i cc. of water gave after 4 days an average mercury rise of 2.35 cm. Quantitative tests on the medium in which the fungus had grown showed ''oxidase" present there also. From these results it appears probable that the greater oxidase activity of diseased bark is due to a summation of the oxidase activity of normal bark and of the canker fungus itself. This may also account for the difference in behavior of the oxidases of the two. The general conclusion to be drawn from the preceding data is that diseased bark has greater oxidase activity than healthy bark, probably because of lower acidity and greater degree of dispersion of the oxidizing agent, and because of an actually greater oxidase content. The lower tannin content of diseased bark (see macro- chemical work) may also be a contributing factor, since tannins are known to cause inhibition of oxidase action. This factor is probably eliminated when precipitated oxidases are used. In reference to the Bunzell apparatus it may be said that while it gives valuable comparative measurements of oxidase activity, those using it must realize its limitations. Conditions within it are artificial; with reference to hydrogen ion concentration, and probably other inhibiting factors, they are unstable and continually moving toward an equihbrium which, so far as we know, does not coincide with the equilibrium obtaining in the plant. Catalase Determinations of catalase activity (table XX) were made on 12 samples of bark, of which nos. 9 and 10 form a set from one tree and nos. 13 to 20 a set from another tree. Nos. 3 and 4 each came from different trees and are the ones used for most of the oxidase work reported in this paper. They were about i year old when tested for catalase. The other samples were freshly prepared for this work in December 19 17 and January 19 18. The hmbs from which they came were carefully cleaned to remove lichens. iqiq] ROSE— BLISTER CANKER 133 Pleurococcus, etc., since microchemical work had shown that such growths have a high catalase activity. The bark was then shaved off, ground in a meat chopper, and allowed to dry on filter paper at room temperature. In the case of samples 9, 10, 14, 16, 18, and 20, calcium carbonate was added during the grinding process at the rate of 0.5 gm. to each 10 gm. of unground bark, to prevent destruction of catalase by the acids of the bark (2) or of the hydro- gen peroxide used. The dried bark was finally ground to a powder and only that part used which passed through an 80-mesh sieve. TABLE XX Catalase activity of apple bark Sample NUMBER 3 4 9 10 13 14 15 16 17 18 19 20 Description of sample Healthy, from sound limb, no car- bonate Diseased, no carbonate Healthy, from sound limb, plus car- bonate Diseased, plus carbonate Healthy, from sound limb, no car- bonate Healthy, from sound limb, plus car- bonate Healthy (?) 5 cm. from canker, no carbonate Healthy ( ?) 5 cm. from canker, plus carbonate Diseased, no carbonate Diseased, plus carbonate Dead, no carbonate Dead, plus carbonate Tempera- TLTIE 23-5 21 .0 21 .0 22.0 20.5 22.5 Positive pressure in cm. After s min. o. 10 0-55 0.83 5-49 0.52 1.83 0.54 0.73 0-55 1-47 5.00 7.01 After 10 min. O.IO 0-9S I . 26 8.59 o. 70 3.02 0.65 1. 01 0.81 2.74 7-37 12.17 Tests were made at room temperature by means of the simpli- fied Bunzell apparatus, using 0.03 or o.iogm. of bark powder, I cc. of water, and 4 cc. of 25 per cent hydrogen peroxide. After the experiment was set up the apparatus was allowed to stand for half an hour, when the manometers were closed and the solutions mixed. The apparatus was shaken for 10 seconds at the end of each minute and readings taken after 5 and 10 minutes. All tests were made in duplicate or quadruplicate, a water blank being included for temperature corrections as in the oxidase work. 134 BOTANICAL GAZETTE [february A test for catalase was run also on the fungus powder pre- viously mentioned, using 0.03 gm. in each tube and calculating the results to the basis of o.iogm. The average mercury rise (positive pressure) produced in 3 tubes was i .65 cm. in 5 minutes and 2.57 cm. in 10 minutes, or, calculated to the basis of o . 10 gm., 5.49 cm. in 5 minutes, and 8.55 cm. in 10 minutes. It is worthy of note that a powder prepared from Nummularia mycelium grown in Raulin's solution, which is acid to litmus, showed no catalase activity. Experiments with different amounts of material showed that the positive pressure varies directly with the amount of material used. It was deemed legitimate, therefore, to calculate all results to the basis of o . 10 gm. of bark powder, and the figures for final tabulation were so calculated. The results for samples 14, 16, 18, all from the same tree, show that diseased bark (sample 18) had more than twice the catalase activity of seemingly healthy bark 5 cm. away from the canker (sample 16), but only nine-tenths of that of bark from a sound unaffected limb on the same tree (sample 14). Dead cankered bark from this tree (sample 20) had 4 times the catalase activity of healthy bark, 12 times that of seemingly healthy bark next the canker, and nearly 5 times that of diseased bark. In the case of samples 9 (healthy) and 10 (diseased), the results are reversed, since the diseased had a catalase activity nearly 7 times greater than that of the healthy bark. The reason for the discrepancy between these two sets is not clear. The high catalase activity of sample no. 10 can hardly have been due to the presence of Kchens, etc., or of an admixture of really dead bark, for precautions were taken when the samples were removed to avoid these sources of error. From the present data the only conclusion that can be drawn is that diseased bark from different trees varies considerably in its catalase activity, and that in general the more completely the bark is destroyed by the fungus the greater is its catalase activity. This condition is probably to be explained by the presence in the diseased bark of considerable amounts of mycelium which, as shown, produces a catalase of its own. The seemingly healthy bark near the canker when compared with sound and with diseased bark appears to form an exception igig] ROSE— BLISTER CANKER 135 in the series. Its catalase activity is less than that of either of the others and seems to be less affected by tissue acids when no car- bonate is added. It is possible that near the canker the host's catalase is injured by materials from the fungus, even in advance of actual invasion by the hyphae. The fungus catalase may not appear here at all, but only later in the diseased bark, and in increasing amounts as the amount of mycelium increases. The oxidase activity of samples 13, 15, 17, 19, together with the catalase activity of samples 14, 16, 18, 20, identical with them except for the addition of carbonate, are given in table XXI. TABLE XXI Catalase activiiy of apple bark Description of sample Healthy Healthy ( ?) 5 cm. from canker Diseased Dead Manometer readings expressed IN CM. OF mercury USING O.I CM. OF BARK POWDER Catalase Oxidase 1. 16 1-47 1-95 0.88 It will be seen that there is a gradual increase in oxidase activ- ity from healthy to diseased bark, but a marked decrease in the case of dead bark. The catalase is considerably lower in apparently healthy bark near a canker than in the bark of an unaffected limb, but very much higher in the bark killed by the fungus than in bark from a healthy Umb. Microchemical analysis Tests for oxidase, peroxidase, and catalase were made on fresh bark, all others on bark preserved in 50 per cent alcohol. The results are given in table XXII. In making the tests for oxidase (direct action) and peroxidase (indirect action), the brownish purple color due to oxidation of benzidine was found most marked at first, in both healthy and diseased bark, in a zone 2 or 3 cells wide just inside the cork and in the pith rays. Later it came to about the same intensity over 136 BOTANICAL GAZETTE [FEBRUARY the whole section. Catalase, judging by evolution of gas when H2O2 was added, was evenly distributed in all the tissues. Tests with FeCl3 on sections of bark successively farther and farther TABLE XXII REStTLTS OF MICROCHEMICAL TESTS ON HEALTHY AND DISEASED BARK Substance Reagent Reaction Healthy Diseased Cellulose IKI and 75 per cent H2SO4 Ruthenium red Phloroglucin and HCl Ferric chloride Diphenylamine in 75percentH2S04 Sudan III 50 per cent acetic acid SO per cent H2SO4 Oxalic acid Fliickigers reagent IKI Picric acid and NaaCOj Berlin blue reaction I per cent benzidine in 50 per cent alcohol I per cent benzidine and H2O2 H2O. + + + + in bast + + Pectin + -1- Lignin Tannin Nitrates + in bast + Fats + Especially in parenchyma next to cork + + Crystals sol- uble + + CaS04 formed + + Crystals not changed + + + + + + + + -\- Same as for Calcium (crystals) . . . Calcium oxalate (crys- tals) Direct reducing sugars Starch fCyanogenic gluco- ] side, probably 1 amvgdalin healthy bark + + Crystals sol- uble -+-+ CaS04 formed -i--|- Crystals not changed + + + Oxidase (direct action) Peroxidase (indirect action) Catalase + + + distant from the badly browned region showed steadily increasing amounts of tannin. Pectin seemed to be present in about equal amounts in both healthy and diseased tissues. Macrochemical analysis Six samples were analyzed. The analytical methods used for 4 of them are based on those^employed by Koch for the quantitative study of animal and plant tissues (21, pp. 199-207). The differ- ence in material required minor variations from these methods, but it is not thought necessary to describe them here. The other 1919I ROSE— BLISTER CANKER 137 2 were analyzed according to a method devised by Kraybill (unpublished work) in a study of the chemical composition of tomato plants. Material for 4 of the samples, healthy i and 2 and diseased I and 2, was taken from 8-10 cm. apple limbs cut in January at the Missouri State Fruit Experiment Station, and shipped from there by express. As soon as these samples arrived they were pre- pared as follows: bark designated as "healthy" was removed from sound limbs with a box scraper and cut into pieces half an inch square; about 150 gm. were then weighed quickly on a torsion bal- ance to hundredths of a gram and put into enough redistilled alcohol (95 per cent) to give an alcohol concentration of approximately 85 per cent. The bottles containing the samples were then set into a steam bath until the alcohol came to a boil, then on top for I hour longer, to inactivate the enzymes. Bark designated as "diseased" was taken from 8-10 cm. limbs showing well developed but not old cankers, usually about 45 cm. long. A strip of moist browned bark 2-3 cm. wide around the outside of the canker was removed with the box scraper, cut up, weighed, and preserved as described. This material usually contained small portions of the seemingly healthy bark outside of the canker, but never any part of the black dead material that often covers the central part of the cankered areas. Healthy samples 3 and 4 were taken from a 7 cm. limb cut in April when the bark peeled easily, to avoid removing small shavings of wood along with the bark, as was inadvertently done in the case of healthy samples i and 2 (see discussion of table XXV) . Healthy samples 3 and 4 were not extracted with hot alcohol and ether as in the method described by Harvey (21); instead the alcohol for pre- serving was filtered into a 1000 cc. flask and made up to volume. One-twentieth aliquots were then pipetted off into small beakers, evaporated to a syrup, and used later for dry weight and other estimations. The partly extracted bark was dried as described for the other samples, weighed, ground, allowed to come to air dry condition, and one-twentieth aliquots weighed out as before. This method of handling the material is much shorter than the Koch method and is very satisfactory if one is not interested in the distribution of substances in the various fractions. 1^8 BOTANICAL GAZETTE [February ^o Dry weight. — One- tenth or one- twentieth ahquots, in tared crucibles or beakers, were brought to constant weight in a vacuum desiccator after intermittent drying for various lengths of time at about 1 00° C. Nitrogen. — Estimations were made by the Kjeldahl-Gunning method, modified to include the nitrogen of nitrates. For healthy samples i and 2 and diseased samples i and 2 estimations were made separately on fractions 2 and 3; no nitrogen was found in fraction 2. Estimations for healthy samples 3 and 4 were made on one-twentieth of the alcohol extract combined with one-twentieth of the partly extracted bark. Carbohydrates. — Healthy samples i and 2, diseased samples i and 2 : in the case of fraction 2, direct reducing sugars, and reducing sugars after mild hydrolysis, were estimated by the Bertrand volumetric method and calculated as dextrose by use of the Munson and Walker tables (34) . The more important details of manipula- tion, including precipitation of non-sugars, are given by Cul- pepper, Foster, and Caldwell (16). The polysaccharides in fraction 3 were estimated as dextrose, but after 2 . 5 instead of 5 hours' hydrolysis (16). Healthy samples 3 and 4: one-twentieth of the air dry, partly extracted bark was further extracted on a filter with about 200 cc. of water at 40° C, the filtrate being collected in a beaker containing one-twentieth of the alcohol extract. Estimation of sugars and polysaccharides in the combined extracts were then made as already described. The results of the analysis are given in tables XXIII and XXIV and summarized in table XXV. The most important differences shown in the tables, as between healthy samples i and 2 and diseased samples i and 2, are as fol- lows: diseased tissue contains 3 . 23 per cent more dry matter than healthy, although here much depends on the manner in which the sample is taken; on the basis of dry weight, fraction i is larger in the diseased by 4.56 per cent (nearly doubled), indicating a synthesis of lipoids by the fungus; fraction 3, the alcohol- water- insoluble residue, is larger by i .83 per cent, while fraction 2, con- taining the alcohol-water-soluble substances, is smaller by 6.27 per cent. These results are strikingly similar to those found by I9I91 ROSE— BLISTER CANKER 139 Culpepper, Foster, and Caldwell (16), working with black rot of apples, caused by Sphaeropsis malorum. The increase in total TABLE XXIII Results of analysis of healthy bark Material Percentage wet weight Percentage wet weight Percentage drj' weight Percentage dry weight Total solids Sample i 51-55 2.56 14.84 34 14 0.23 1-58 0-55 1.07 7.40 8-47 Sample 3 46.13 0.217 0 915 0.634 7-524 Sample 2 51-50 2 59 13.26 35-45 0.23 1 .60 0.54 0.22 7-56 7-78 Sample 4 46.24 0.231 0.949 0.662 7.189- Sample i Sample 2 " F. 4-97 28.79 66.21 0.45 3.06 1.07 2.08 14-35 16.43 Sample 3 " F. 5- 04 25-85 69.08 0.46 3 13 " F, Total nitrogen Direct reducing sugars Reducing sugars after mild hydrolysis Reducing sugars after strong hydrolysis Fi and Fj Reducing sugars after strong hydrolysis F3 I 05 1. 71 Reducing sugars after strong hydrolysis, total Total solids 14 74 16.4s Sample 4 Total nitrogen 0.46 I .96 1.72 16.20 Direct reducing sugars Reducing sugars after mild hydrolysis 0.50 2-05 Reducing sugars after strong hydrolysis 1-45 16.55 TABLE XXIV Results of analysis of diseased bark Material 1 Sample i Percentage wet weight Sample 2 Percentage wet weight Sample i Percentage dry weight Sample 2 Percentage dr>' weight Total solids 54-29 4.69 11.52 38-07 0-45 1.42 0.66 0-13 8.80 8-93 55-12 5-77 11.52 37-94 0-45 1.56 0.59 0.38 8.84 9.22 " F. 8.64 21.21 70.16 0.83 2.62 1.22 0.23 16. 21 16.44 " F. 10.47 20.89 68.80 0.81 2.83 " F3 Total nitrogen Direct reducing sugars Reducing sugars after mild hydrolysis. . . . Reducing sugars after strong hydrolysis F, and F2 Reducing sugars after strong hydrolysis F3 1.07 0.70 16.04 16.74 Reducing sugars after strong hydrolysis, total 140 BOTANICAL GAZETTE [febru.\ry nitrogen in diseased bark may be due to fixation by the fungus or to a withdrawal of nitrogen from the surrounding tissue. Fur- ther data are necessary before a conclusion can be reached. Cul- pepper, Foster, and Caldwell found protein-nitrogen content of fraction 2 for diseased apples larger than for normal ones, but the total nitrogen for the whole tissue smaller for the former than for the latter. TABLE XXV Summary * Average percentage wet weight Average percentage dry weight Material Healthy i Healthy 3 and 2 1 and 4 1 Diseased i and 2 Healthy i and 2 Healthy 3 and 4 Diseased i and 2 Total solids CT e-2 a(\ to 54 70 5-23 11-52 37-57 0.45 1.49 0.62 9.08 " " Fi 2-58 14-05 34-00 0.23 1-59 0-55 8.12 -T- -^ 5.00 27-32 67-65 0.46 3.00 1 .06 16.44 0,48 2.00 1-59 16.38 9-56 21.05 69.48 " F. " F, Total nitrogen 0. 24 0.93 0.65 7-35 0.82 Direct reducing sugars .... Direct reducing sugars after mild hydrolysis Direct reducing sugars after strong hydrolysis 2-73 I 15 16.59 Results with healthy samples 3 and 4 furnish Httle of additional interest. They show, however, that as far as total nitrogen and starch are concerned, the small amount of wood in the other 2 healthy samples had no effect on the results. The difference in the case of dry weight and reducing sugars before and after hy- drolysis is probably due to the fact that samples 3 and 4 were taken from a Hmb cut early in the growing season, while samples i and 2 were taken from limbs cut in the dead of winter. Estimation of tannin The method used was that of Lowenthal, as modified by Proctor (34, p. 150). Material for analysis was taken from 8-12 cm. Ben Davis limbs cut in November, December, and January. The bark was cut off as already described, ground in a meat grinder, and transferred to a glass moist chamber at once. About 10 gm. were then weighed out and set to boil in 400 cc. of I9I9] ROSE— BLISTER CANKER 141 water as required by the Lowenthal method; at the same time duplicate samples were taken for moisture determination. What- ever may have been the errors introduced by this method, the agreement between duplicates taken for moisture determination was very close in most cases, as is shown in table XXVI. TABLE XXVI Percentage of dry matter in duplicate samples of various lots of bark analyzed Sample Healthy i . . . " ' 2. . . Diseased 4 . . . 6... Dead 7 " 8 " 9 Duplicate 2 Average 47 65 47 77 45 76 45 86 46 86 46 91 49 53 49 99 49 41 49 69 49 73 49 80 64 74 64 79 75 83 76 10 The results of the analysis of 9 different samples of bark are shown in table XXVII. TABLE XXVII Percentage of tannin in healthy and diseased apple bark Description of sample Tannin (percent- age dr>- weight) Healthy I 5 16 3 64 3.38 4.06 2. 8-10 cm. from canker 3. From same tree as 6 and 9 Average Diseased A 2.49 5. From same limb as 2 6. 3 56 2.9^ Average 2.99 0.25 1 . 14 Dead (from surface of canker) 7. From same limb as i 8 0. ■ I-5I 0.97 \verage The Lowenthal method probably determines merely the easily water-soluble tannins, but fails to reach those tied up with ..^di 142 BOTANICAL GAZETTE [February the suberin. If suberin is for any reason more abundant in the diseased bark, an error would thus be introduced which might invaUdate any comparisons based on the results obtained. Sub- ject to this possible correction the results shown in table XXVII confirm those obtained in the microchemical analysis; that is, they show a progressive decrease in tannin as the bark is more and more affected by the disease. Healthy bark was found to contain on the average 4.06 per cent of tannin, diseased 2.99, and dead o .97. If sample i healthy, which gave a high figure, and sample 7 dead, which gave a low figure, be eliminated, the averages become healthy 3.51, diseased 2.99, dead 1.33. The figures for samples 3, 6, and 9, all from the same tree, are healthy 3 .38, diseased 2 .93, dead i .51. There is undoubtedly a difference between bark from a sound limb and seemingly healthy bark from a limb that is badly cankered. The latter is usually shghtly browned throughout when first cut ofT and rapidly becomes reddish brown on exposure to air. Really healthy bark under such conditions shows only a slight browning. Whatever the results with apple bark may mean, they are not in agreement with the statement made by Kerr (see Cook and Wilson, 15, p. 26, footnote) that because of the greater stability of tannin and the disappearance of other constituents "all decayed wood and bark give higher tannin contents, no matter what causes the decay." If confirmed by further analyses they would indicate a different relation between host and parasite with reference to tannin in the case of blister canker than obtained in any of the cases studied by Kerr. Leaching of tannin may account for the low percentage found in dead apple bark, as suggested by the chemist of the Chestnut Tree Blight Commission (15, p. 6) for old cankers of chestnut blight, but can hardly be responsible for the condition found in diseased bark. Summary I. Measurements with the simpHfied Bunzell apparatus show that apple bark attacked by Nummularia discreta causes about twice as much oxidation of pyrogallol, pyrocatechin, guaiacol, and benzidine as does healthy bark. 1919] ROSE— BLISTER CANKER 143 2. The gradual slowing down of oxidation in the Bunzell apparatus is shown to be due, in part at least, to increasing hydrogen ion concentration, brought about by the oxidation process itself. The equilibrium reached in the oxidase apparatus seems to be a false one, which can be disturbed by the addition of either fresh oxidase reagent or plant material. When tested by the formula for a unimolecular reaction, the oxidase reaction gives values for k, which indicate clearly a linear relationship between time and amount of change and suggest that the oxidase is a catalytic agent, 3. The hydrogen ion concentration of diseased bark (Ph = 5 -61) is definitely less than that of healthy bark (Ph = 5.i5). Work with buffer solutions shows that this difference is not great enough to account for all of the difference in the oxidase activity of the two kinds of material. When mixtures of the two are brought to the same hydrogen ion concentration by means of buffer solutions, diseased bark still shows greater oxidase activity. 4. The temperature and duration of drying have an effect on the acidity and the oxidase activity of both healthy and diseased bark. 5. Eight-tenths per cent gelatine increases the oxidase activity of both kinds of bark. This may be due to the action of gelatine as a protective colloid which prevents precipitation of the "oxidase." It is not due to buffer action. 6. The concentration of hydrogen ion necessary or complete inhibition of oxidase activity of healthy bark lies between 3 .55 and 3 .8oXio~'*;for that of diseased bark between 3 .55 and4.27Xio~'^. 7. Oxidation in the apparatus comes to an end only after several days instead of after a few hours, as stated by Bunzell. 8. When the "oxidase" is precipitated in 2 fractions, the first has greater oxidizing power than the second, and the 2 combined have slightly greater oxidizing power than when tested separately. 9. Catalase determinations gave the following results: healthy 3.02 (cm, positive pressure), seemingly healthy 5cm. from the canker i.oi, diseased 2.74, dead 12.17; results from oxidase determinations for the same stages were 1.16, 1.47, 1.95, 1.85 (cm. negative pressure). These results show some discrepancies, but justify the general statement that the more severely the bark 144 BOTANICAL GAZETTE [February is attacked by the fungus the greater is its catalase activity, and that catalase activity in part is in indirect ratio to oxidase activity. 10. Microchemical tests indicate, for diseased bark, a partial disintegration of cellulose, a disappearance of cyanogenic glucoside, and a lower content of starch, calcium oxalate, and tannins. 11. Macrochemical analyses show that diseased bark has a higher percentage of dry matter, lipoids, alcohol-water-insoluble residue, and total nitrogen, but a lower percentage of alcohol-water- soluble material than healthy bark. The percentage of carbo- hydrates in both tissues seems to be about the same. Differences in tannin content are definite but not large. Sound healthy bark contains more than diseased bark and diseased bark more than dead bark from the surface of the canker. 12. The greater oxidase activity of diseased bark is probably due to the combined activity of the oxidases of fungus and host, lower acidity, and possibly to a greater degree of dispersion of the oxidizing agent. The lower tannin content of diseased bark may also be a contributing factor. The writer wishes to acknowledge his indebtedness to Dr. William Crocker, Dr. F. C. Koch, and Dr. Sophia H. EcKERSON for valuable suggestions and criticism during the course of the investigation. Thanks are due to Dr. Paul Evans of the Missouri State Fruit Experiment Station for bark material used in the experiments. U.S. Department of Agriculture Washington, D.C. LITERATURE CITED 1. Allard, H. a.. Some properties of the virus of the mosaic disease of tobacco. Jour. Agri. Research 6:649-674. 1916. 2. Appleman, Charles O., Some observations on catalase. Bot. Gaz. 50:182-192. 1910. 3. Atkins, W. R. G., Recent researches in plant physiology. London. 1916. 4. Bach, — , and Chodat, O. R., Untersuchungen liber die Rolle der Peroxyde in der Chemie der lebenden Zelle: IV, tJber Peroxydase Ber. Deutsch. Chem. Gesell. 36:600-605. 1903. 5. Battelli, F., and Stern, L., Die Oxydationsfermente. Ergeb. Physiol. 12:96-268. 1912. iqiqI rose— blister canker 145 6. Bayliss, VV. jM.. Principles of general physiology. London. 1915. 7. BouRQUELOT, Emile, Presence de ferment oxydant dans quelques sub- stances medicamenteuses. Compt. Rend. Soc. Biol. 49:25-28. 1897. 8. BoviE, W. T., A direct reading potentiometer for measuring and recording the actual and the total reaction of solutions. Jour. Med. Research 33:301. 1915. 9. BuxzELL, H. H., The measurement of the oxidase content of plant juices. U.S. Dept. Agric. Bur. PI. Ind., Bull. 238. 1912. 10. , A biochemical study of the curly-top of sugar beets. U.S. Dept. Agric. Bur. PI. Ind., Bull. 277. 28. 1913. II- , Oxidases in healthy and in curly dwarf potatoes. Jour. Agric. Research 2:373-403. 1914. 12. , The relationship existing between the oxidase activity of plant juices and their hydrogen ion concentration, with a note on the cause of oxidase activity in plant tissue. Jour. Biol. Chem. 28:315-333. 1916. 13- , The mode of action of the oxidases. Jour. Biol. Chem. 24:91- 102. 1916. 14. Clark, E. D., The plant oxidases. Diss. Columbia Univ. New York City. pp. 1-113. 1910. 15. Cook, M. T., and Wilson, G. W., The influence of the tannin content of the host plant on Efidolhia parasitica and related species. N.J. Exper. Sta. Bull. 291. 6. 1916. 16. Culpepper, C. W., Foster, A. C, and Caldwell, J. S., Some effects of the blackrot fungus, Sphaeropsis malorum, upon the composition of the apple. Jour. Agric. Research 7:17-40. 1916. 17. DoBY. Geza, Biochemische Untersuchungen iiber die Blattrollkrankheit der Kartoflfel. Zeitschr. Pflanzenkrank. 21:204-211, 401-403. 1912. 18. Dony-Henault, Octave, Contribution a I'etude methodique des oxydases. Mem. Bull. Acad. Roy. Belgique 105-163. 1908. 19. Duggar, B. M., Severy, J. W., and Schmitz, H., Studies in the physiology of the fungi. I\'. The growth of certain fungi in plant decoctions. Ann. Mo. Bot. Card. 4:166. 1917. 20. Freiberg, G. W., Studies in the mosaic diseases of plants. Ann. Mo. Bot. Card. 4:175-232. 1917. 21. Harvey, E. M., Some effects of ethylene on the metabolism of plants. Bot. Gaz. 60:199-207. 1915. 22. Hasselbring, Heinrich, Canker of apple trees. 111. Agric. Exper. Sta. Bull. 70. 225-239. 1902. 23. Hober, Rudolf, Physikalische Chemie der Zelle und Gewebe. Leipzig und Berlin. 1914. 24. Kastle, J. H., The oxidases and other oxygen-catalysts concerned in biological oxidations. U.S. Treas. Dept. Hygienic Lab. Bull. 59. 1-164. 1910. 25. Michaelis, J., Die Wasserstoffionen-konzentration. Berlin. 1914. 146 BOTANICAL GAZETTE [February 26. Perrin, J., Mecanisme de I'electricalisation de contact et solutions colloidales. Jour. Chim. Physique. 3:50-110. 1905. 27. Philip, J. C, Physical chemistry, its bearing on biology and medicine. New York and London. 1910. 28. Reed, G. B., The relation of oxidase reactions to changes in hydrogen ion concentration. Jour. Biol. Chem. 27:299-303. 1916. 29. Reed, H. S., The enzyme activities involved in certain fruit diseases. Va. Agric. Exper. Sta. Rept. 1911-1912. pp. 51-78. 30. Rose, D. H., Oxidation in healthy and diseased apple bark. Box. Gaz. 60:55-65. 1915. 31. SoRAUER, Paul, Die angebliche Kartoffelepidemie, gennant die BlattroU- krankheit. Internal. Phytopath. Dienst (Beigabe, Zeitschr. Ptlanzen- krank.). 33-59. 1908. 32. , Die neueren Untersuchungen von Quanjer iiber die Ursache der BlattroUkrankheit der Kartoffel und der Sorauer'sche Standpunkt. Zeitschr. Pflanzenkrank. 23:244-253. 1913. 33. Trillat, O., Sur le role d'oxydases que peuvent jouer las sels manganeux ex presence d'un colloide. Compt. Rend. 138:274-277. 1904. 34. Wiley, H. W., Oiificial and provisional methods of analysis. Association of Official Agricultural Chemists. As compiled by the committee on the revision of methods. U.S. Dept. Agric. Bur. Chem. Bull. 107 (rev). 1908. 35. Woods, A. F., Observations on the mosaic disease of tobacco. U.S. Dept. Agric. Bur. PI. Ind., Bull. 18. 17-22. 1902. I MORPHOLOGY OF THE GENUS ACTINOMYCES. II Charles Drechsler (with plates ii-ix) Taxonomy of species The obscurity that has surrounded the morphology of Actinomyces, besides involving the genus in improbable specula- tions concerning its phylogenetic relationship to the bacteria, has brought about also a most unfortunate condition in the taxonomy of the many described species. Most of the work on the genus has been done by investigators with bacteriological inclinations, and even where this has not been true, the prevalent view of the nature of these minute plants has lead to the adoption of methods scarcely applicable to mycological research. A discussion of characteristics like the occurrence of endospores, flagella, capsules, sheaths, and involution forms, caimot be regarded as constituting a morpho- logical treatment more satisfactory for species of Actinomyces than for species of Mucor or Boletus. The dependence of certain biochemical processes, particularly chromogenesis, on definite conditions of nutrition, and the con- spicuous differences resulting from comparatively slight changes in the substratum, have long been noted by students of Actitiomyces, yet descriptions of new species have continued to appear, based largely and often quite exclusively, on these variable activities. Very frequently writers have not compared their organisms with others reported by previous investigators; and in recent years there has been a tendency to disregard altogether the taxonomic contributions of the preceding decades. Moreover, while identi- ties have frequently not been recognized where they existed, in other cases organisms have been supposed to be identical on extremely slight evidence. One of these cases that has led to an unusual measure of confusion is that of Gasperini's Actinomyces chromogenus. This species was identified by both Gasperini and Rossi-DoRiA (19) with an organism isolated from the air by the 147] [Botanical Gazette, vol. 67 148 BOTANICAL GAZETTE [February latter and designated as Streptothrix nigra. In culture it was characterized by a dark brown or black discoloration of certain kinds of substrata, a reaction easily obtained on potato agar, for example, and ascribed by Lehman and Sano (13) to the production of tyrosinase. Until recently it has been the custom among writers to refer nearly every member of the genus showing a tyrosinase reaction to Actinomyces chromo genus, Lutman and Cunningham (14) going so far as to identify this species with the potato scab organism. This practice, which would unite forms so different in appearance and method of development as, for example, Actinomyces I and III, regardless of pronounced differences in size and in dextrorse or sinistrorse condition, is not defensible on mor- phological grounds. Krainsky resolved the " chromogenus " com- plex into 4 species; while Waksman and Curtis increased the number of derivatives to 8. Of the 17 morphologically distinct saprophytic species figured in this paper, 11 exhibit a tyrosinase reaction; and these represent less than one-fifth of the number of similarly active species which the writer had occasion to examine. The genus awaits the attention of an investigator in a position to make a comprehensive study involving at least the larger pro- portion of species existmg within wide geographical ranges. The summaries given later, of the more important facts about each species selected for morphological treatment, are not to be regarded as descriptions intended for taxonomic purposes. ACTINOMYCES I Cultural characters. — On glucose agar (o . 5 g. peptone, 10. o g. glucose, 20.0 g. agar, 1000 cc. tap water) nutritive mycelium of individuals smooth, opalescent, more or less confluent; sporula- tion moderately slow and commencing as a light creamy zone near the periphery; no diffusible stain. On potato agar (decoction of 200 g. peeled potatoes, 2 .0 g. glucose, 20.0 g. agar, 1000 cc. water) nutritive mycelium light olivaceous; sporulation moderately abundant, the raised areas where the yellowish gray fructifications are to appear being previously distinguishable by a deep brownish green coloration; guttation never copious, often absent; tyrosinase reaction moderate, but distinct. igig] DRECHSLER— ACTINOMYCES • 149 Morphology. — The development of the erect sporogenous hyphae of this species is strictly successive, and may be followed in the branch d in fig. 2, the younger of any 2 hyphae being dis- tinguishable by its attenuated attachment. The partly disrupted chain of spores di here represents the original prolongation of branch d; the chain d2 represents a secondary branch, the spores here being mature but still retaining their spiral disposition without showing indications of disruption; while dj represents a tertiary branch, in which septation has not commenced. A similar sequence is illustrated in the succession of derivatives bi, b2, bj, and 64 from the branch b, as well as in the 5 elements ci-cj associated with the branch c. A more complex system of fertile h>'phae is shown in fig. i, but the larger fructifications are probably 10 or even 15 times more extensive, and bear many thousands of spores. The species is characterized by close sinistrorse spirals, of 2-6 turns, and 3-4 /x in diameter, which during the later stages of maturation are relaxed, although indications of them may persist in the flexuous or sinuous course of the mature chains of spores. The mature spores are ellipsoidal, 1-2-nucleated, with a distinctly visible wall and a central vacuole of varying size. They measure i . 2-1 .4X1. 4-20 /x, and upon germinating produce 2-4 germ tubes, which early pro- liferate numerous branches, and show at intervals some dark staining granules. Isolated 5 times from soil collected in Cambridge, Massachusetts. ACTINOMYCES II Cultural chail\cters. — On glucose agar, growth moderate; nutritive mycelium colorless, early covered with a cretaceous or downy aerial mycelium ; pigment absent. On potato agar, develop- ment of nutritive mycelium moderately rapid; aerial mycelium appearing in scattered areas, first white, later becoming slightly discolored; substratum stained yellow by a soluble pigment; tyrosinase reaction absent. Morphology. — ^The most conspicuous feature of this species is the extraordinary thickness of the septa (0.3-0.35 fi) associated I50 • BOTANICAL GAZETTE [February with spore production, and their insertion at distensions in the sporogenous hyphae. A comparison of branch ci with the younger branch C2 (fig. 5) corresponding to the conditions shown diagram- matically in figs. Sb and 8c respectively, shows that the growth in thickness of the hyphae takes place subsequent to the appearance of the septa. After the filament has attained its growth, the septa split along a median plane (figs. 5a, 5^7, M), and the 2 halves are drawn apart by the contraction of the delimited protoplasts. Further maturation occurs in the distribution of the deeply stain- ing wall substance, in the strengthening of the peripheral wall, and by an enlargement of the latter, the elimination of the median, hourglass-like constriction of the spore, resulting finally in an approximately cylindrical structure measuring o . 7-0 .9X0. 8-1 . i [x. The terminal and the basal spores of each chain retain a some- what asymmetrical shape, owing to the absence of the massive septum at one of their ends. By an apparently abnormal develop- ment, metachromatic granules may be formed in the spores derived from some hyphae, resulting in a condition illustrated in the lowest spore in fig. 8/. The axial filaments are represented by long prostrate hyphae, branching at irregular, long intervals. Septation is confined to the fertile branches. The sterile hyphal portions below the spo- rogenous terminations taper gradually toward their attenuated attachments. Development and sporogenesis near the axial terminations are successive, and involve the formation of close sinistrorse spirals of 1-5 turns, 3 .5-6.0 ix in diameter. Isolated 3 times from soil collected in Cambridge, Massachusetts. ACTINOMYCES III A. lavendulae Waksman and Curtis Cultural characters. — On glucose agar, nutritive mycelium slightly yellow on reverse side, central area completely covered with velvety aerial mycelium, first white but gradually assuming a beautiful lavender shade; no soluble stain. On potato agar, growth very profuse; mycelium abundant, changing from white to laven- der; guttation moderate to profuse; tyrosinase reaction vigorous. I9I9] DRECHSLER— ACTINOMYCES 151 Identity with A . lavendulae established by comparison of cultural and morphological characteristics. MoRPHOLOGY.^ — The mycelium consists of long prostrate axial filaments, branching rarely except at the end. Sporulation is usually initiated at the tip of the filament, and proceeds basipetally by the insertion and transformation of almost invisible septa, to the point of attachment of the first sporogenous branch (figs. 10, 1 1). The sporogenous branches are rarely crowded, although at the base of the sporogenous axial termination an opposite arrangement is not uncommon. Secondary branching occurs frequently; septa are entirely absent, except when associated with the progressive basipetal delimitation of spores. The sporogenous hyphae terminate in dextrorse, moderately compact spirals of 4-12 turns, 2 .0-3 .8 /^ in diameter. The spores are ellipsoidal, with nuclei not readily demonstrable. Metachro- matic material occurs abundantly in many old hyphae (fig. 15). Isolated 3 times from soil collected in Cambridge, Massachusetts. ACTINOMYCES IV Cultural characters. — On glucose agar, nutritive myceHum colorless; aerial mycelium moderately profuse, velvety in appear- ance, changing from white to smoky blue; no guttation. On potato agar, tyrosinase reaction vigorous; aerial mycelium first produced white, not subsequently much discolored, becoming matted to the substratum as a result of excessive guttation, and later completely overgrown by a loose growth of smoky blue secondary mycelium. Morphology. — The sporogenous branches with dextrorse spirals of 2-12 open turns, i .5-2.5 ix in diameter, are attached to the long axial filaments usually at wide intervals, in a loose race- mose arrangement. Secondary branching, although rare, occurs occasionally, and is then associated with simultaneous sporulation (fig. 18). Development of the 1-2-nucleated spores, o . 7-0 .8X0.9- 1.1 IX, proceeds by the insertion of conspicuous septa, followed by their constriction and subsequent conversion to hyaline isthmuses (figs. 7oa-e). Two germ tubes are usually produced, of a more or 152 BOTANICAL GAZETTE [February less uniform diameter, and proliferating branches at relatively wide intervals. Isolated twice from soil collected in Cambridge, Massachusetts. ACTINOMYCES V Cultural characters. — ^On glucose agar, nutritive mycelium on reverse side slightly yellowish; the surface completely covered with a luxuriant velvety or cottony weft of pinkish-yellow aerial mycelium; guttation slight. On potato agar, nutritive mycehum chocolate-colored, firm lichenoid, crimped around margin; aerial mycelium, as on glucose agar, but less profuse; tyrosinase reaction vigorous. Morphology. — The fertile hyphae, which are attached to prostrate axial filaments at long intervals, are terminated by rela- tively close sinistrorse spirals of 4-12 turns 2.0-4.0/^ in diameter, developing spores (0.6-0.8X0.9-1.1 ju), like Actinomyces IV, by the insertion of conspicuous septa, followed by their constriction and conversion. A peculiar characteristic is found in the sterihza- tion of the basal portion of the fertile hyphae, by an apparent abortion of its lower potential spores. Isolated 3 times from soil collected in Cambridge, IVIassachusetts. ACTINOMYCES VI Cultural characters. — ^On glucose agar, nutritive mycelium colorless, completely covered with a felty aerial mycelium, first white, later assuming a deep smoky tinge. On potato agar, nutri- tive mycelium excessively wrinkled, partially covered with a slightly discolored aerial mycelium; tyrosinase reaction vigorous. Morphology.— The species appears closely allied to Actino- myces V, differing from the latter chiefly in the absence of any evidence of sterilization, and in the shorter length of its sinistrorse sporogenous spirals, which consist of only 2-6 turns, 2.0-4.0 ju in diameter. The spores are uninucleated, measure 0.7-0.8X0.9- I . I /i, and are developed by the insertion and transformation of conspicuous septa. Fertile hyphae are attached to the axial fila- iQig] DRECHSLER— ACTINOMYCES 153 merits with considerably greater frequency, and secondary branch- ing, characterized by the successive type of development, is com- mon. Isolated once from soil collected in Cambridge, IMassachusetts. ACTINOMYCES VII Cultural characters. — On glucose agar, nutritive mycelium colorless, early developing an aerial mycelium from the center outward, the latter changing from white to light gray with increas- ing age. On potato agar, nutritive mycelium luxuriant, developing rapidly; aerial mycelium represented by a slight cretaceous develop- ment toward the top of the slant; tyrosinase reaction vigorous. Morphology. — This species departs from the main trend of the 3 preceding forms in the relatively close arrangement of its branches on the axial filament, and in the elaboration of these branches by further ramifications in a typically successive order. Nearly spherical to elKpsoidal spores, 0.6-0.8X0.7-1 .0 )u, are produced from moderately close sinistrorse spirals of 3-8 turns 2 . 0-3 . o /x in diameter, by the development represented in fig. yoa-e, but the septa are relatively thin, and occasionally fall below the limit of clear visibility. Isolated twice from soil collected in Cambridge, Massachusetts. ACTINOMYCES VIII Cultural characters. — ^On glucose agar, nutritive mycelium nearly colorless, secreting a difi'usible yellow pigment; aerial my- celium moderately profuse, velvety, first white, later changing to a light bluish color. On potato agar, growth similar, but soluble pigment absent; no tyrosinase reaction. Morphology. — ^The fertile hyphae of this species may be attached to the axial filaments in a diffuse racemose arrangement (fig. 46), or crowded in a compact capitate system. The swellings in the axial filaments in figs. 43, 44, 45, and 46 at the bases of sporogenous branches indicate the mode of origin of the Leptomitus- like distensions shown in figs. 47 and 48. 154 ' BOTANICAL GAZETTE [February The small, ellipsoidal, uninucleated spores, 0.5-0.6X0.6- o . 8 ju, are formed from close, sinistrorse spirals of 2-10 turns i . 2- 2 . 5 ju in diameter. Indications of septa can be seen only rarely. The mature spore chains upon collapsing cohere in irregular zoogloea-like masses, a peculiarity of behavior dependent probably on a gelatinization of wall material. Upon germination, i or 2 tubes are produced, relatively thick and abundantly branching. Isolated 6 times from soil collected in Cambridge, Massachusetts. ACTINOMYCES IX Cultural characters. — On glucose agar, nutritive mycelium colorless, forming no soluble pigment; aerial mycelium at first white, becoming light smoky blue in the course of a few days. On potato agar, cultural characters similar, but growth more profuse, guttation moderate, discoloration of aerial mycelium more rapid; tyrosinase reaction absent. Morphology. — -The most characteristic feature of this species is the greater thickness of the fertile hyphae below the second turn of the spiral. The latter are sinistrorse, usually with very close turns, varying in number from i to 16, and measuring i .5-2.0 fx in diameter. They give rise to ellipsoidal, uninucleated spores, 0.5-0.7X0.6-1.0 JU, without the appearance of clearly visible septa. It seems highly probable that cross- walls nevertheless occur, since occasionally a median partition may be differentiated in the hyaline attentuated connections between two spores (fig. 51), suggesting a development similar to that indicated in fig. 'joa-e. Isolated once from soil collected in Cambridge, Massachusetts. ACTINOMYCES X Streptothrix alba Rossi-Doria; Actinomyces griseus Krainsky (?) Cultural characters. — On glucose agar, growth poor and not characteristic. On potato agar, growth excessively rapid, nutritive mycelium colorless; aerial mycelium firm, white, changing rapidly to a yellowish gray; secondary growth occurring in the formation of numerous successive rings of sporodochia, or in the develop- ment of cottony white masses of mycelium from below the thick 1 9 1 9] DRECHSLER—ACTINOM YCES 155 crust of old mature spores; tyrosinase reaction absent; substratum stained a faint greenish yellow in old cultures. Morphology.— According to Waksman and Curtis, the aerial filaments of this species possess only a sHght tendency to branch. The writer was led to a somewhat dififerent conclusion, as the axial hyphae are usually found to proliferate fertile branches at mod- erately close intervals. Occasionally, as jn fig. 58, indications of a successive sequence may be observed, but more frequently the de\'elopment of the different elements of a ramifying system occurs without any recognizable interrelation. The short, cylindrical spores, 0.7X0.7-1 .0 ;u, are formed, as in Actinomyces XVI, by a septation of the fertile h\phae, followed by sphtting of the parti- tions along a median plane, but the septa are usually less conspicu- ous, and often not clearly visible, and the fertile hyphae show no indication of a spiral condition. A striking dimorphism characterizes the mycelium of this species, as well as that of a number of other forms observed by the writer. The deeper sterile aerial h>phae below the sporogenous layer tj^Dically are extremely minute, with a diameter frequently not exceeding o .3 /x; their protoplasmic contents show Httle affinity for stains; and the contours of their walls are uniformly smooth. The more superficial hyphae, which usually attain a thickness of i.o/x, and are distinguishable by markedly irregular contours, contain dense deep staining protoplasm; and when septa are present, they are sometimes associated, as in Actinomyces XVII and XVIII, with spherical structures. The thicker filaments bear the sporogenous branches, and, in general, appear to constitute the expanded prolongations of the minute hyphae (fig. 59). Isolated twice from soil collected in Cambridge, Massachusetts; once from tap water; very frequently from outdoor air; several times from gross cultures of dead leaves; 4 times from horse dung undergoing fermentation at 50-60° C. Synonomy. — In his description of Streptothrix alba, Rossi- DoRiA records two characteristics that estabhsh its identity beyond much danger of confusion: a conspicuous preponderance in number over any of its congeners on plates exposed to the air, and a tendency toward the formation of concentric rings more pronounced than that of any other species. Rossi-Doria 156 BOTANICAL GAZETTE [February attributed this preponderance in the air to its omnivorous char- acter, enabhng it to develop on a large variety of substrata, "Questa Streptothrix cresce, si puo dire, dappertutto, tanto su ter- reni di natura vegetale quanto su terreni di natura animale. E per cio nonche la grande sua produzione di spore, che essa si trova cosi diffusa nell'aria ed altrove. Pare che essa possa svilupparsi anche nel terreno." In spite oif this fortunate and quite distinctive char- acterization, the specific term "albus" subsequently came to be used in a manner as miscellaneous as " chromogenus," being applied generally to any type with a Hght mycelium showing no tyrosinase reaction. The same species was treated in the publication of Waksman and Curtis as Actinomyces griseus Krainsky. I have not been able to satisfy myself fully about the identity of Krainsky's organism; nor would it seem possible to reach any definite con- clusion without an examination of authentic material. ACTINOMYCES XI Cultural characters. ^On glucose agar, nutritive mycelium first colorless, becoming sHghtly reddened with increasing age; aerial mycelium first white, rapidly changing to a bluish violet. On potato agar, nutritive mycelium gradually becomes deep red by the slow accumulation of a slightly diffusible pigment ; tyrosinase reaction absent. Morphology. — More or less erect fructifications are developed along the distal portions of long prostrate filaments. Branching is abundant and only occasionally shows indications of a successive sequence. The aerial hyphae in the dendroidic structures (figs. 64, 66) are often conspicuously vacuolate, especially in the inflated distensions from which a number of fertile branches arise. The latter terminate in sinistrorse spirals of 4-6 turns, 2.0-3.0 m in diameter, from which, by the insertion of conspicuous septa and their subsequent transformation to hyaline isthmuses, spores o . 5-0 .7X1 .0-1 . 2 fj. are produced. Isolated once from soil collected in Cambridge, Massachusetts; identical with an organism isolated by Mr. H. J. Conn from soil collected near Geneva, New York. igig] DRECHSLER— ACTINOMYCES 157 ACTINOMYCES XII A. aureus Waksman and Curtis Cultural characters. — On glucose agar, nutritive mycelium yellowish on reverse side; aerial mycelium changing from white to pale yellowish gray; soluble stain absent. On potato agar, nutri- tive mycelium darker on reverse side; aerial mycelium more profuse, forming a somewhat more deeply colored felty layer; tyrosinase reaction moderate. Identity with Actinomyces aureus established by comparison with authentic material of the latter. Morphology. — In this species long prostrate filaments termi- nate in more or less erect fructifications. Secondary branches are proliferated from the lateral elements, generally in successive sequence. A more or less pronounced cuneate thickening of the hyphae below the insertion of a branch is characteristic of the species. The ellipsoidal, uninucleated spores, o . 5-0 .7X0. 8-1 . 2 n, are formed by the insertion of conspicuous septa in open, sinistrorse spirals of 2-7 turns, 3.0-4.PJU in diameter. Isolated twice from soil collected in Cambridge, Massachusetts. ACTINOMYCES XIII Cultural characters. — On glucose agar, nutritive mycelium light orange-brown, the separate individuals fused into a massive pellicle with a depressed, crimped margin. On potato agar, nutri- tive mycelium dark chocolate-brown, wrinkled, lichenoid, secret- ing a diffusible red pigment; t>Tosinase reaction absent. Aerial mycelium on both substrata loose, cottony; developing slowly, first white, later changing to a dull bluish tint. Morphology. — The aerial mycehum consists of extremely long filaments, which rarely show any evidence of branching (figs. 74 75), and toward their terminations follow an undulating or slightly spiral course. Sporulation occurs as the result of protoplasmic contractions without the appearance of visible septa, the chains of cylindrical spores, 0.4X1 .2-1 .6 /x, being held together for some time by the evacuated portions of hyphal wall, that seem to undergo no apparent constriction. Isolated 3 times from soil collected in Cambridge, Massachusetts. 158 BOTANICAL GAZETTE [February ACTINOMYCES XIV Cultural characters. — On glucose agar, nutritive mycelium usually colorless, but frequently becoming deep brown or black; aerial mycelium consisting of a dense velvety weft, first white, later changing to a creamy yellow. On potato agar, growth similar; tyrosinase reaction absent. Morphology." — This species is characterized by the production of extensive prostrate fructifications through the proliferation of numerous lateral branching processes from long axial filaments (figs. 76, 79, 81). A septum is occasionally present immediately above the attachment of a branch, but more frequently is absent. Secondary ramifications, resulting in more or less complex elements, take place without reference to the stage of sporogenesis in the proliferating branch. The elHpsoidal uninucleated spores. 0.5- 0.7X0.8-1 .2 /x, are derived from sinistrorse spiral h>phae of 1-8 turns, 2 .0-4.0 n in diameter, by the insertion and transforma- tion of relatively thin septa, or without the appearance of demon- strable septa. Isolated 4 times from soil collected in Cambridge, Massachusetts. ACTINOMYCES XV Cultural characters. — On glucose or potato agar, nutritive mycelium opalescent; aerial mycelium first white, becoming only slightly discolored with age; tyrosinase reaction moderate. Morphology. — Microscopically this species closely resembles Actinomyces IV, differing from the latter chiefly in the abundant proliferation of branches of the second or of a higher order. The lateral elements thus formed follow the successive t>pe of develop- ment (figs. 82, St,). The uninucleated spores, 0.7X0.9-1 .0 ju, are formed from dextrorse spiral hyphae of 3-12 turns, i . 8-2 . 5 m in diameter, by the constriction of conspicuous septa, and their trans- formation into hyaline isthmuses. Isolated twice from soil collected in Manhattan, Kansas. ACTINOMYCES XVI Cultural characters. — On glucose agar, growth very meager; never producing an aerial mycelium. On potato agar, develop- iqiq] DRECHSLER— ACTINOMYCES 159 ment rapid; nutritive mycelium dark brown or greenish brown; aerial mycelium profuse, changing from white to violet or pinkish gray; guttation profuse; tyrosinase reaction moderate. Morphology. — In this species the characteristic development consists in the prohferation of a number of long branches in an irregular whorl from a long and somewhat thickened axial filament. Secondary branching is common, but usually more or less remote. Vacuoles associated with h>phal distensions are found in the axial filaments and in the main branches, and metachromatic granules occur abundantly in many of the older sterile hyphae (fig. 91). The long cyhndrical spores, 0.6-0.7X1.0-2.0/1, are formed by the septation of sporogenous h>'phae that terminate in open, sinistrorse spirals of 2-3 turns, 4.0-5.5 y, in diameter, followed by the sphtting of the septa along a median plane, and the separation of the two halves by a contraction of the dehmited protoplasts. The progress of sporogenesis is usually basipetal, but not infre- quently the first divisions may result in a number of segments of varying lengths, which by subsequent divisions are reduced to the magnitude of the ultimate spores. Isolated once from soil collected in Cambridge, Massachusetts. ACTINOMYCES XVII A. scabies (Thaxter) Giissow (6) Morphology. — The aerial mycehum of this species, which is one of the largest of dextrorse forms, consists of long prostrate filaments on which lateral branches are inserted at short intervals. Secondary branching is abundant and usually associated with a successive order of development (figs. 93a7-aj). The more or less cylindrical spores, 0.8-0.9X1 3-1 .5 m, are developed from dextrorse spiral hx-phae of 3-14 turns, 2 .0-3 . 5 m in diameter, by the insertion of conspicuous septa and their subsequent sphtting along a median plane. In many h>phae the septa before their division can be seen to occupy a transverse equatorial position in the pecuHar spherical structures to which reference has been made elsewhere, and which here occupy slight but perceptible hyphal distensions (figs. 92, 93///, 93/^2, loicy). Whenever the spherical structures i6o BOTANICAL GAZETTE [February are absent, the fertile hyphae are uniformly isodiametric. It is not certain whether these structures appear in all sporogenous branches at some time preceding the contraction of the deHmited protoplasts, or are more or less accidental in their occurrence. They also are found associated with septa in the sterile axial fila- ments, and here similarly occupy local hyphal distensions. After the individual spores have become separated, the connecting seg- ments of evacuated hyphal wall contract shghtly to form somewhat narrowed isthmuses, which persist until the mature spore chains are disrupted. In germinating, the spore usually produces i or 2 germ tubes. The preparation from which figs. 92-101 were drawn was derived from one of 5 organisms communicated by Mr. M. Shapavalov, who writes that "all were tested in inoculation experiments in 1912-1913, and proved to be pathogenic." Three of the other organisms were found to be identical morphologically with the one figured in plate VIII, while the fifth did not produce an aerial mycelium sufficiently profuse to permit of a satisfactory microscopic examination, although the general appearance of the culture indi- cated that it also is identical with Actinomyces XVI. ACTINOMYCES XVIII Cultural characters. — On glucose agar, growth meager; nutritive mycelium colorless; aerial mycelium slow to develop, first white, later showing slight discoloration; diftusible pigment absent. On potato agar, development very rapid; nutritive mycelium dark; aerial mycelium profuse, felty, bluish gray; guttation moderate; tyrosinase reaction vigorous. Morphology. — This species is characterized by an unusual degree of variabihty in its fructifications. In figs. 102 and 107 is represented a relationship between axial filament and sporogenous branches common to many members of the genus. Fig. 108 shows a slight departure from this type in the thickening of the subter- minal portion of the axial filament bearing the spiral branches. Further departures are expressed in the tufted grouping of the spiral hyphae in fig. 104, and in the distended and extremely vacuolated condition of the axial filament in fig. 106. A strikingly [ 1919] DRECHSLER— ACTINOMYCES 161 aberrant type is seen in fig. 103, the fertile branches being short, inserted at close, irregular intervals, and showing no spiral tendency; while the axial filament is thick and abounding in spherical struc- tures containing either deposits of metachromatic material or a partial equatorial septum. In the dextrorse spiral hyphae of 1-8 open turns, 2.0-3 o ju in diameter, the ellipsoidal spores, 0.8-0.9X1 o-i .6 ju, are produced by the insertion of conspicuous septa, sometimes in association with spherical structures. The presence of the latter (fig. 106), however, is not here indicated by local distensions. Subsequently the cross- walls undergo constriction and conversion to narrow connecting isthmuses. In the aberrant fertile h>^hae (those without any spiral tendency), sporogenesis appears to take place in a more miscellaneous manner. Definite septa can rarely be distinguished, the spores seeming to result from protoplasmic contractions. Isolated once from soil collected in Cambridge, Massachusetts. Summary 1. The vegetative thallus of Actinomyces consists of a mycelium composed of profusely branching hyphae, the terminal growing portions of which are densely filled with protoplasm. Toward the center of the thallus the vacuoles increase in size and may be associated with the presence of metachromatic granules, the latter having nothing in common with bacterial endospores or "micro- cocci,'' for which they were mistaken by early observers. 2. The vegetative mycelium attains an extent incomparably greater than the branching figures recorded for bacteria of the acid- fast group, and the hyphae lack the uniformity in diameter generally characteristic of the Schizomycetes. 3. The aerial mycelium produced on suitable substrata by most species occurs usually in the form of a mat of discrete fructifications; but in other species these fructi^cations are frequently combined to form numerous and peculiar erect Isarioid sporodochia. 4. In any case each individual fructification represents a well characterized sporogenous apparatus, consisting of a sterile axial filament bearing branches in an open racemose or dense capitate arrangement. The primary branches may function directly as i62 BOTANICAL GAZETTE [February sporogenous hyphae, or may proliferate branches of the second and of higher orders, sporogenesis in the latter case being confined to the terminal elements, the hyphal portions below the points of attachment of branches remaining sterile. 5. Two tendencies in the development of fructifications are recognizable: one leading to an erect drendroidal type, in which successively proliferated fertile elements undergo processes of sporogenesis in continuous sequence; and the other leading to a prostrate racemose type, in which sporogenesis is delayed in the older branches until the younger branches have also attained their final extension. The majority of species show these tendencies combined in different ways. 6. The sporogenous hyphae of most species are coiled in peculiar spirals, sometimes resembling the spores of the hyphomycetous genus Helicoon. These spirals exhibit pronounced specific char- acteristics in the number, diameter, and obliquity of their turns, and especially in the direction of rotation (whether dextrorse or sinistrorse) . 7. Sporogenesis, where it can be followed, begins at the tips of the fertile branches and proceeds basipetally. In the larger num- ber of species the process involves the insertion of septa which, in certain cases, are relatively very massive, and in others so thin as to be barely discernible. The disposition of these septa, while the delimited spores undergo maturation processes, varies with the species: (i) they may remain more or less unaltered; (2) they may sufifer a median split, the two resulting halves being then separated as the result of the longitudinal contraction of the young spores, leaving alternate portions of hyphal walls completely evacuated; or (3) they may first become considerably constricted and sub- sequently converted into non-stainable isthmuses connecting the mature spores. The apparent absence of septa in the sporogenous hyphae of other forms is perhaps attributable to optical difficulties. 8. Granules are readily differentiated in the spores of many species which possess the staining properties and uniformity of size characteristic of nuclei; they generally occur singly, but in the larger spores of a few forms two are often found occup}'ing diagonally opposite positions. iQig] DRECHSLER— ACTINOMYCES 163 9. As in the vegetative thallus, metachromatic granules occur in the aerial mycelium, being very rarely found in the spores or sporogenous hj-phae, but becoming very abundant in degenerate sterile hyphae. 10. The older axial filaments of some species show marked distensions which, in extreme cases, result in figures simulating Leptomitus. These arise as local distensions at the points of attach- ment of the more extensive lateral sporogenous processes. Cuneate modifications of the sterile axial filaments below the origins of branches also occur. 1 1 . Curious spherical structures appear regularly in some forms, both in the sterile axial hyphae, where they may contain either a median septum or a number of peripheral metachromatic granules, and in the sporogenous h}T)hae, where they are associated with the regularly spaced septa. 12. The spores germinate readily in suitable solutions, producing 1-4 germ tubes, the approximate number being more or less char- acteristic of the species. 13. Owing to the absence of any well defined bacterial character- istics, the writer is of the opinion that the view that Actinomyces represents a transition between the Hyphomycetes and the Schizo- mycetes, as well as the phylogenetic corollary based upon it, may safely be abandoned. If mere size is to be regarded as important, it would appear to be equally profitable to look for bacterial affinities in some ascomycetous and sphaeropsideaceous forms, the hyphae of which are similarly very minute. It is doubtful whether far-reaching taxonomic generalizations can be based on the "acid-fast'' staining reaction, especially as this reaction has not played a very important role in mycological research. There seems to be no adequate reason why the genus should not be classed in an unqualified manner with the Hyphomycetes, as a mucedineous group with tendencies toward an erect Isarioid habit. The writer wishes to acknowledge his indebtedness to Professor R. Thaxter, under whose direction this work was done; to Pro- fessor W. G. Farlow for the use of books; and to Professor B. Fink for samples of soil collected on the islands of Porto Rico and l64 BOTANICAL GAZETTE [February Cuba. Thanks are due also to Mr. S. A. Waksman, to Mr. H. J. Conn, and to Mr. M. Shapavalov, for kindness in supplying cul- tures of organisms isolated by them. Cryptogamic Laboratories Harvard University Cambridge, Mass. LITERATURE CITED 1. BosTROEM, Untersuchungen iiber die Actinomykose des Menschen. Beitrage zur Pathologische Anatomie und die AUgemeine Pathologie. 9:1-240. 1890. 2. DeBary, a., Vergleichende Morphologie der Pilze. 377-379. 1884. 3. DoMEC, F., De la morphologie de I' Actinomyces. Arch, de Medecine experimentale. 4:104-113. 1892. 4. Gasperini, G., Ricerche morfologiche e biologiche sul Genere Actinomyces- Harz. Annali deir Istituto d' Igiene Spermentale 2:167-229. 1891. 5. Gilbert, tjber Actinomyces und andere Aktinomyceten. Zeitsch. f. Hyg. u. Infektskr. 47:383-406. 1904. 6. GtJssGW, H. F., The systematic position of the organism of the common potato scab. Science, N.S., 39:431-432. 1914. 7. Harz, C. O., Actinomyces bovis, ein neuer Schimmel in den Gewebe des Rindes. Jahres. d. Thierarzneischule zu Munchen. 1877-78. 8. Israel, J., Neue Beobachtungen auf dem Gebiete der Mykosen des Menschen. Virchow's Archiv. 74:15-53. 1878. 9. JoHNE, Bericht des Veterinar-Wesen in Konigr. Sachsen. 155. 1879. 10. Krainsky, a., Die Aktinomyceten und ihre Bedeutung in der Natur. Centralbl. Bakt. II. 41:649-688. 1914. 11. Kruse, W., Systematik der Streptotricheen. Fliigge: Mikroorganismen. 3d ed. 48-66. 1898. 12. Lachner-Sandoval, v., tJber Strahlenpilze. I. 1898. 13. Lehman, K. B., and Sano, tJber das Vorkommen von Oxydations- fermenten bei Bakterien und hoheren Pflanzen. Arch. Hygiene 67:99- 113. 1908. 14. Lutman, B. F., and Cunningham, G. C, Potato scab. Vermont Agric. Exper. Sta. Bull. 184. 1-64. 1914. 15. Mace, E., Sur les caracters de culture du Cladothrix dichotoma. Compt. Rend. Acad. Sci. 106:1622. 1888. 16. MacFadyean, J., The morphology of Actinomyces. British Medical Journal. 1 339-1 344. June 15, 1889. 17. MiEHE, H., Die Selbsterhitzung des Heues. 1907. 18. Neukirch, H., tJber Strahlenpilze. II. 1902. 19. Rossi-DoRiA, E. D., Su di alcune specie di Streptothrix trovate nell' aria. Annali dell' Inst, d' Igiene. 1892. 1 919] DRECHSLER— ACTINOMYCES 165 20. Sauvageau, C, and Radais, M., Sur le genre Cladothrix. Ann. de 1' Inst. Pasteur 6:242-273. 1892. 21. ScHUTZE, H., Beitrage zur Kenntnis der thermophilen Aktinomyzeten und ihre Sporenbildung. Arch. Hygiene 67:35-56. 1908. 22. Thaxter, R., The potato scab. Conn. Sta. Rep. 15:153-160. 1892. 23. Waksman, S. a., and Curtis, R. E., The Actinomyces of the soil. Soil Science 1:99-134. 1916. 24. Wolff, M., and Isr-Ael, J., tjber Reinkulturen des Actinomyces und seine Ubertragbarkeit auf Thiere. Virchow's Archiv. 126:11-59. 1891. EXPLANATION OF PLATES II-IX All except figs. 7, 8, 70, and loi, which are semidiagrammatical repre- sentations with a magnification of approximately 8000, were drawn with the aid of a camera lucida with a magnification of 2750. PLATE II Actinomyces I Fig. I. — Moderately well developed fructification. Fig. 2. — Somewhat smaller fructification showing successive order of development: a, chain of spores, largely disrupted, developed from termination of axial hypha; b, c, d, secondary branches that have given rise respectively to series of elements bi-b4, ci-cj, and di-dj. Fig. 3. — Spore germinating with 4 germ tubes. Actinomyces II Figs. 4-6. — Portions of aerial mycelium, showing conspicuous septa in fertile branches, and relation of latter to axial filaments: a, b, c, branches proliferated successively from same filament; bi-bj, ci-cj, elements prolif- erated successively from branches b and c respectively. Fig. 7. — Portion of branch c (fig. 5) showing attachment of successively formed spiral elements. Fig. 8a-/. — Successive stages in development of fertile hypha. Actinomyces III Figs. 9-14. — Portions of aerial mycelium. Fig. 15. — Portion of degenerate hypha containing abundance of meta- chromatic material. PLATE III Actinomyces IV Fig. 16. — Short chain of spores showing nuclei, and 2 deep staining remnants of constricted septa in hyaline isthmuses between spores. Figs. 17-20. — Portions of aerial mycelium. Fig. 28. — Spore germinating with 2 germ tubes. l66 BOTANICAL GAZETTE [February Actinomyces V Fig. 21. — Aerial hypha with spiral termination and 2 fertile branches, more mature elements showing failure of spore to develop in proximal portion. Fig. 22. — Aerial hypha with 2 spiral elements. Fig. 23. — Young spiral branch of 15 turns attached to axial hypha con- taining metachromatic granules. Figs. 24-26. — Spiral branches soon after insertion of septa, showing cross- walls absent from portion above basal septum. Fig. 27. — Young spiral branch. Figs. 29, 30. — Spiral branches with spores mature, and non-septate portion completely evacuated. Figs. 31, 32. — Degenerate filaments containing much metachromatic material. Actinomyces VI Fig. S3- — Portion of aerial mycelium showing 2 spiral elements with nuclei in mature spores of one; septum in axial filament associated with basal septum in branch. Fig. 34. — Similar to fig. ^^j but without visible nuclei. Figs. 35-37. — Other portions of aerial mycelium. Actinomyces VII Figs. 38-40. — Portions of aerial mycelium with sporogenous branches in various stages of development. PL ATE. IV Actinomyces VIII Fig. 41. — Sporogenous apparatus with mature spores cohering in zoogloea- like masses. Fig. 42. — Prostrate hypha containing numerous metachromatic granules and bearing a branch with many crowded spiral ramifications. Fig. 43. — More open type of sporogenous apparatus with lateral elements attached to axial hypha at intervals. Fig. 44. — Young sporogenous apparatus with spiral branches more or less crowded. Fig. 45. — Somewhat older system of spiral hyphae, some of which have become converted into zoogloea-like masses of spores. Fig. 46. — ^Lateral element bearing 8 spiral branches. Figs. 47, 48. — Portions of degenerate mycelium showing Leptomitiis-Wkt enlargements occupied by vacuoles, and metachromatic granules in con- strictions. Figs. 49, 50. — Spores germinating with i germ tube. 19 ig] DRECHSLER— ACTINOMYCES 167 Actinomyces IX Fig. 51. — Portion of aerial mycelium showing spiral termination con- verted into chain of uninucleated spores, and presence of remnants of septa in hyaline isthmuses. Fig. 52. — Similar to fig. 51, but without indications of septa between spores. Fig. 53. — Portion of aerial mycelium showing septa in axial filament above insertions of some sporogenous branches. Fig. 54. — Sporogenous branches with portion below second turn of spiral conspicuously thickened. Fig. 55. — Sporogenous branch of 11 turns. Fig. 56. — Spore germinating with i germ tube. Fig. 57. — Sporogenous branch of 15 close spiral turns. Actinomyces X Figs. 58-60. — Portions of aerial mycelium. Figs. 61, 62. — Spores germinating with i and 2 germ tubes respectively. PLATE V Actinomyces XI Figs. 63, 64. — More or less erect fructifications terminating long prostrate filaments, showing origin of groups of sporogenous branches from local hyphal distensions occupied by conspicuous vacuole. Figs. 65, 66. — Intermediate portions of aerial mycelium. Fig. 113. — Spore germinating with i germ tube. Actinomyces XII Figs. 67, 68. — Erect fructifications terminating long prostrate aerial fila- ments, exhibiting a pronounced tendency toward successive type of develop- ment, and showing cuneate hyphal enlargements below insertions of branches. Fig. 69. — Intermediate portion of aerial mycelium. Fig. 700-e. — Progressive stages in development of sporogenous hypha, occurring in this and numerous other species. Fig. 71. — Spore germinating with 3 germ tubes, PLATE VI Actinomyces XIII Fig. 72. — Two portions, a terminal, h subterminal, of one long, unbranched, continuous sporogenous hypha showing very slight spiral tendency. Fig. 73. — Chain of spores with deep staining polar granules. Figs. 74, 75. — Portions of aerial mycelium showing branching. Actinomyces XIV Figs. 76-81. — Portions of aerial mycelium showing arrangement of spo- rogenous branches on hyphae, and method of sporulation. l68 BOTANICAL GAZETTE [February PLATE VII Actinomyces XV Fig. 82. — Portion of aerial mycelium: elements ai-aj and bi-bj suc- cessively proliferated from branches a and b respectively. Fig. 83. — Sporogenous branch a with 2 secondary branches; younger (aj) associated with a septum above insertion (successive type) ; older (02) not set off by septum. Fig. 84. — Axial filament with 3 branches, bearing successively poliferated elements ai-aj, as well as branch ax, latter not associated with septum in primary branch above point of attachment. Fig. 85. — Fructification developed entirely in successive sequence, with 2 chains of uninucleated spores. Figs. 86-88. — Spores germinating with 2 germ tubes. Actinomyces XVI Fig. 89. — ^Large fructification consisting of axial filament a-ai , with whorl of 5 primary branches a2-a6, each bearing i or more secondary branches. Fig. 90. — Spiral termination of sporogenous branch. Fig. 91. — Old filament containing many metachromatic granules. PLATE VIII Actinomyces XVII Fig. 92. — Portion of aerial mycelium showing spherical structures asso- ciated with septa in local distensions of sporogenous branch. Figs. 93, 94. — Portions of aerial mycelium, some lateral elements bearing secondary branches (indicated by numerals above i) developed successively; 94c, unusually long fertile branch. Fig. 95. — Portion of aerial mycelium similar to one shown in fig. 92. Figs. 96-100. — Spores germinating with i or 2 germ tubes. Fig. loia-e. — Successive stages in development of sporogenous branch, ex and cy representing either alternative or probably successive stages. PLATE IX Actinomyces XVIII Fig. 102. — Sporogenous branch of usual type soon after appearance of septa. Fig. 103. — Portion of fructification bearing aberrant fertile branches without spiral terminations. Fig. 104. — Aerial filament with several spiral branches borne terminally. Fig. 105. — Chain of mature spores developed from branch of spiral type. Fig. 106. — Degenerate axial filament containing large vacuoles and spheri- cal structures and bearing a fertile spiral branch. Figs. 107, 108. — Portions of aerial mycelium showing fertile hyphae of spiral type. Figs. 109-112. — Spores germinating with i or 2 germ tubes. BOTANICAL GAZETTE, LXVII PLATE II 'c/)s/er DRECHSLER on ACTINOMYCES BOTANICAL GAZETTE, LXVII PLATE III DRECHSLER on ACTINOMYCES BOTANICAL GAZETTE, LXVII PLATE IV DRECHSLER on ACTINOMYCES BOTANICAL GAZETTE, LXVII PLATE V DRECHSLER on ACTINOMYCES BOTANICAL GAZETTE, LXVII PLATE VI DRECHSLER on ACTINOMYCES BOTANICAL GAZETTE, LXVII PLATE VII ^.dl J3^ DRECHSLER on ACTINOMYCES BOTANICAL GAZETTE, LXVII PLATE VIII DRECHSLER on ACTINOMYCES BOTANICAL GAZETTE, LXVII PLATE IX DRECHSLER on ACTINO^IYCES BRIEFER ARTICLES BYRON DAVID HALSTED (with portrait) With the passing of Byron David Halsted we have lost another of our pioneer botanists. Although not one of the earliest pioneers, he was a pioneer in many respects. He was one of that group of botanists who laid the foundation of the science in America at a time when the sub- ject was recognized by very few American col- leges and universities. He was one of that still smaller group who took up the study of applied botany and worked faith- fully for its advancement. Our younge;- plant pathologists know how difficult it is to find a disease of an economic crop that is not at least mentioned in his reports. He was among the first to report the introduction of several foreign patho- genic organisms. Born at Venice, Cayuga County, New York, June 7, 1852, he was left an orphan at an early age and was cared for by relatives. He graduated from the Michigan Agricultural College with the degree of B.S. in 1871, and received the M.S. degree from the same college in 1874. In 1878 he received the Sc.D. degree from Harvard, being the first man to take the doctorate in cryptogamic botany from that university. He was managing editor of the American Agriculturist from 1880 to 1885; Professor of Botany in the Iowa Agri- cultural College 1885-1889; and Professor of Botany in Rutgers College 169] [Botanical Gazette, vol. 67 I70 BOTANICAL GAZETTE [February and Botanist of the New Jersey Agricultural Experiment Station from 1889 until his death, August 26, 1918. Had he lived until February, 1919, he would have rounded out a full 30 years in the service of the state of New Jersey. During the greater part of these 30 years he was active in both College and Station, but in the latter part of his career poor health necessitated his retirement from the classroom. Although a very busy man, he found time to serve his science by acting as Associate Editor of the Bulletin of the Torrey Botanical Club from 1890 to 1893 and as a contributor to the Systematic Flora of North America. In 1877 the Massachusetts Horticultural Society honored him with its silver medal. He was a member of a number of scientific societies, serving as president of the Society for the Promotion of Agri- culture f-rom 1877 to 1879 and of the Botanical Society of America in 1900-1901. Dr. Halsted was a true lover of nature, and nature made him a most warm hearted and lovable man. He loved to commune with nature and was a most enthusiastic collector. In addition to his own studies, he furnished a great quantity of material for study by other mycologists and from which many new species were described; in fact, the mycological collections not only of America but of the entire world contain material collected by him. The writer looked upon him as a botanist of the old school, and yet he was an up-to-date botanist in every way. After devoting the greater part of his career to mycology, poor health and a failing eyesight forced him to abandon his favorite line of work. He could not leave the field of botany, however, but merely transferred his efforts to a line of plant breeding which did not require the use of the microscope, and worked with the renewed energy and the enthusiasm of a boy. Dr. Halsted was more than a botanist; he was a broad, scholarly man and a pubHc spirited citizen. He was always interested in athletics and in his youth was an amateur baseball pitcher. He never lost his interest in the sport, but was a regular attendant at intercollegiate games, always placing himself so that he could observe the work of the pitcher. His love for literature and his keen interest in the state and community were made manifest by a poem which he wrote on the occasion of a civic parade when the nation was called to arms in 191 7. He was the author of many papers, and while most of us will think of him as a scientist, it should be remembered that many of his papers had to do with other subjects.— Mel. T. Cook, New Jersey Agri- cultural Experiment Station, New Brunswick, N.J. CURRENT LITERATURE BOOK REVIEWS Soil conditions and plant growth The rapid progress that is being made in the scientific field encompassed by plant physiology and soil science is evidenced in the appearance of a third and enlarged edition of Russell's' Soil conditions and plant growth in the series of Monographs on Biochemistry edited by Plimmer and Hopkins. The first edi- tion (191 2) contained 166 pages, including 13 pages of citations; the second edition (191 5) contained 188 pages, including 14 pages of citations; and the third edition (191 7) contains 240 pages, including 18 pages of citations. The growth of the subject is marked by the addition of a chapter on "The relationship between the micro-organic population of the soil and the growth of plants" to the second edition, and a chapter on "The colloidal properties of soil" to the third edition. In his preface to this new edition the author states that he has not attempted to refer to every paper published on the subject since the first edition, but that his "guiding principle has been to include only those that brought in some new idea or profoundly modified an old one." The choice of papers is naturally a question of judgment, yet the reviewer feels that certain important omissions have been made, and he will therefore direct the reader's attention to these in the hope that they may serve to supplement this admirable treatment of practical plant physiologj'. Certain papers pubhshed since 191 7, and therefore not referable to as "omissions," will be included in order to bring the subject up to date. Chapter II on "The requirements of plants " presents a modern physiological basis for the rest of the book. This chapter has been enlarged to the extent of 5 pages over the first edition. In the discussion of limiting factors the treat- ment of oxygen (only 9 lines) and temperature seems quite inadequate. Atten- tion is therefore directed to the following papers: On temperature relations of plants. — Leitch, I., Some experiments on the influence of temperature on the rate of growth of Pisum sativum. Ann. Botany 30:25-46. 1916; Lehenbauer, p. a.. Growth of maize seedlings in relation to temperature. Physiol. Researches 1:247-288. 1914; Lepeschkin, W. W., Zur Erkenntnis der Einwirkung supramaximaler Temperaturen auf die Pflanze. Ber. Deutsch. Bot. Gesell. 30:703-714. 1912; Maximow (Maksimov), N. A., Chemische Schutzmittel der Pflanzen gegen Erfrieren. Ber. Deutsch. Bot. ' Russell, E. J., Soil conditions and plant growth. 3d ed. pp. 243. figs. 14. New York: Longmans, Green and Co. 1917. 171 172 BOTANICAL GAZETTE [February Gesell. 30:52-65, 293-305, 504-816. 191 2; Groves, J. F., Temperature and life duration of seeds. Box. Gaz. 63:169-189. 1917. On oxygen relations of plants. — Shull, Charles A., The oxygen minimum and the germination of Xanthium seeds. Box. Gaz. 52 : 453-477. 1911 ; Cannon, W. A., and Free E. E., The ecological significance of soil aeration. Science N.S. 45:178-180. 1917; LiviNGSXON, B. E., and Free, E. E., The effects of deficient soil oxygen on the roots of higher plants. Article in " Contributions to Plant Physiology." Johns Hopkins University. Reprint from Johns Hopkins University Circular. March 19 17. The new edition deserves high praise for its comprehensive treatment and impartial judgment on the modern developments in soil chemistry. The dis- cussion on the use of dilute acids in soil analysis, based on the author's own work in the Rothamsted laboratories, is the first contribution on this subject that has forsaken empirical experimental methods and adopted a modern physico- chemical procedure. The review of the highly controversial literature on soil acidity is eminently fair. The attention of the reader is called to the following recent papers, each of which contains an extensive bibliography: Chrisxensen, H. R., Experiments in methods for determining the reaction of soils. Soil Science 4: 1 15-178. 1917; Truog, E., Soil acidity. I. Its relation to the growth of plants. Soil Science 5:169-195. 1918; Rice, F. E., and OsuGi, S., The inversion of cane sugar by soils and allied substances and the nature of soil acidity. Soil Science 5:333-358. 1918. The biological aspects of soil science are fully treated, including the author's own interesting ideas on soil protozoa and partial sterilization. The reviewer feels that Russell has been somewhat partial to his own views in not referring to Sherman's studies. Sherman concludes, on the basis of his experiments, that "no evidence has been obtained which indicates that the beneficial effect of partial sterilization is due to the elimination of a biological factor which is harmful to the bacteria." Bolley's interesting views on "Soil sanitation" deserve mention in a chapter that bears such an all-inclusive title as "The relationship between the micro-organic population of the soil and the growth of plants." The following papers should be read in connection with the chapter: Sherman, J. M., Studies on soil protozoa and their relation to the bacterial flora. Jour. Bacteriology 1:35-66, 165-185. 1916; Kopeloff, N., and Coleman, D. A., A review of investigations in soil protozoa and soil sterilization. Soil Science 3:197-269. 1917; Bolley, H. L., Wheat-soil troubles, causes of soil sickness, etc. Bull. 107. N.D.Agric. Exper. Sta. 1913; Bolley, H. L., Con- servation of the purity of soils in cereal cropping. Science N.S. 32:529-541. 1910; Bolley, H. L., Cereal cropping: sanitation, a new basis for crop rotation, manuring, tillage, and seed selection. Science N.S. 38:249-259. 1913; Hop- kins, C. G., The bread supply. Science N.S. 38:479-481. 1913. Keen's mathematical studies on the retention of water by soil are amply discussed, but no mention is made of Shull's even more important contribution to the problem of the wilting coefficient. The reader is therefore referred to the 1919] CURRENT LITERATURE 173 following: Shull, C. A., Measurement of the surface forces in soils. Bot. Gaz. 62:1-31. 1916. The final chapter is devoted to a theoretical discussion of soil analysis. An appendix describes analytical methods used in England. The reviewer believes ■hat the reader should be cognizant of the following discussions of American methods: Bear, F. E., and Salter. R. M., Methods in soil analysis (Technical Bulletin). Bull. 159. West Virginia Agric. Exper. Sta. Morgantown. 1916; Ames, J. W., and Schollenberger, C. J., Liming and lime requirement of soil. Bull. 306. Ohio Agric. Exper. Sta. Wooster. 1916; Truog, E., .\ new test for soil acidity. Bull. 249. Wisconsin Agric. Exper. Sta. Madison. 191 5; Bou- Youcos, Geo. J., and McCool, M. M., The freezing point method as a new means of measuring the concentration of the soil solution directly in the soil. Tech. Bull. 24. Michigan Agric. Exper. Sta. East Lansing. 191 5. — H. L. Walster. MINOR NOTICES History of phytopathology. — Whetzel,^ in his History of phytopathology, aims "only to set forth in outline what appear to be the most outstanding features in the evolution of the science, and to indicate the proper relation thereto of the men who have chiefly shaped its progress." The chief captions are: (i) The Ancient Era, to the end of the 5th century (5 pp.) ; (2) The Dark Era, 6th to i6th centuries (i p.); (3) The Premodern Era, 1600 to about 1850 (19 pp.); (4) The Modern Era, 1853 to about 1906 (65 pp.); (5) The Present Era, 1906 (8 pp.). As is indicated by the page allotment, the first and second topics are treated very briefly, being barely sketched. The third and fifth topics are treated somewhat more fully, while the most page space is given to "The Modern Era." The book is in the main a series of brief biographical sketches, often with portraits, arranged chronologically under the captions indicated. It will be a convenient reference book for those who may need ready access to such biographies. — F. L. Stevens. Winter botany. — To supplement his pocket manual of woody plants, already noted in this journal,^ Trelease^ has compiled and published a com- panion volume for use in naming our common trees and shrubs when without foliage. The range, extending to 326 genera and over 1000 species, includes most introduced as well as native woody plants. The notable features of the volume, aside from its convenient pocket size and abundant illustrations from most accurate drawings, are the numerous keys and the many citations of litera- ture dealing with winter characters of the various genera and species. The * Whetzel, H. H., An outline of the history of phytopathology, pp. 130. Saun- ders Co. 1918. ^ Box. Gaz. 65:194. 1918. ■» Trelease, William, Winter botany. i6mo. pp. 394. figs. 327. Urbana, 111. Published by the author. 1918. $2.50. 174 BOTANICAL GAZETTE [February drawings are of bud, leaf-scar, pith, and other twig characters upon which the keys are based, so that with the use of a hand lens it should be possible to determine readily the genera, and for the most part the species, for native and introduced trees and shrubs. The author is to be congratulated in making such a fund of unusual information available in such a compact and readily available form. — Geo. D. Fuller. American trees. — Another book on trees, by Emerson and Weed,s has been added to the already large number upon the same subject. It is essen- tially a book for the amateur, since its chief virtue lies in the excellent photo- graphs by Emerson, an entire page being devoted to each species. The absence of keys of any sort renders the book comparatively useless for the identification of an unknown species, but the quality and abundance of the illustrations will make it one the tree lover will wish to have upon his table. — Geo. D. Fuller. NOTES FOR STUDENTS Physiological balance in soil and other nutrient solutions. — Hibbard^ has just published a piece of work on physiological balance in soil solution which is to mark a decided advance (both theoretically and practically) , if the future development of the work approximates its present promise. He extracted the soil solution from an infertile very sandy soil and from a fertile sandy loam by the Van Suchtelen oil pressure method as improved and extended in usefubess by Morgan. 7 Hibbard speaks of this as giving a more concentrated solu- tion than any other extraction method. The solution thus extracted from the poor sandy soil had an osmotic pressure of o . 193 atmospheres, and that from the good soil 1 .81 atmospheres. The soil extracts showed an order of production similar to the soils from which they came. The soil extracts were used instead of distilled water to prepare the Shive 3-salt (KH2PO4, CaCNOj)^, MgS04) nutrient solution. The total concentra- tion of nutrient salts added gave an osmotic pressure of 1.75 atmospheres, and in the 36 different solutions made up from each soil extract and from dis- tilled water the proportions of each salt varied from 10 to 80 per cent of the total nutrient salt osmotic concentration. In the nutrient solution made from the extract of the poor soil the optimum osmotic proportions of the KH.PO4, Ca(N03)2, and MgS04 for the growth of Fultz wheat were 7:1:2 respectively, with a total osmotic pressure of (i . 75+ o . 193) 1 . 94 atmospheres ; in that made from the extract of the good soil 2:7:1 respectively, with a total osmotic pressure of (i .75 + 1 -81) 3.56 atmospheres; and in that made with distilled water 5:2:3, with a total osmotic pressure of s Emerson, Arthur I., and Weed, C. M., Our trees, how to know them. New ed. pp. xxi-l-295. pis. 149. Philadelphia: Lippincott Co. $3-50- (• Hibbard, R. P., Physiological balance in the soil solution. Tech. Bull. Mich. Agric. Exper. Sta. no. 40. pp. 44. 191 7. 7 Tech. Bull. Mich. Agric. Coll. Exp. Sta. no. 28. iQio] CURRENT LITERATURE 175 1 . 75 atmospheres. Analyses of the soil extract of the poor soil showed it to be proportionally low in K^O and PA- Also addition of KH2PO4 to the poor soil at the rate of 150-300 pounds to the acre greatly increases the growth of wheat. A similar analysis of the soil extract of the good soil showed it to be as poor in nitrates, nitrites, and ammonium salts as the extract from the poor soil, but with a relatively high calcium magnesium ratio. Addition of Ca(N03)2 to the good soil at the rate of 200-400 pounds to the acre greatly improved the growth of the plants. Hibbard concludes that the soil is deficient in calcium and perhaps nitrates. If, as Hibbard's results seem to indicate, one can tell the fertilizer needs of a soil by such a study of the soil extracts, he has given us a new practical method of great significance. He has also hinged the fertilizer question largely on the balance between nutrient ions. Hibbard worked only on the early stage of the growth of wheat. It would be of interest to know whether the same ratios apply to the middle and late stages. It will also be interesting to see the nature of these ratios when plants with high sulphur demands (alfalfa, cabbage, etc.), those with high potash demands (potato, tobacco, etc.), or those with high general nutrient demands (squash, cucumber, etc.) are used. These stand in contrast with the wheat with its moderate demands. The method also needs the test of a great range of soil types of various degrees of fertility. Considering all of these variables, the application of the method may become rather complex. The work is sure to stimulate a great amount of investigation. The use of a 3-salt nutrient solution requires that the variation in concen- tration of the several nutrients shall always be in pairs. In Shive's solution a change in concentration of sulphur is always accompanied by a correspond- ing change in the concentration of magnesium, potassium by phosphorus, and calcium by nitrogen. There is no reason for thinking that the magnesium demands of a plant are thus tied up with the sulphur demands; in fact, in soils we think of the sulphur as existing mainly as calcium sulphate, and in fertilizer practice it is more generally added in this form. This is not an adverse criticism of the use of 3-salt nutrient solutions; it is rather an acknowl- edgment of the complexity and innate difficulties of the problem. Since so many variations in proportions of nutrients are possible, the investigator must study only a portion of these, if he completes his investigations within a reason- able time. Livingston and Tottingham* recognize this shortcoming of the 3-salt nutrient solution when they mention the 6 possible combinations of salts and proceed to investigate the best proportions in the second combination. The possible combinations are: I II in IV V VI Ca(N03). Ca(N03)a Ca(H.P04). CaCH.PO^)^ CaSO^ CaSO^ KH.PO^ K.SO4 KNO3 K.SO4 KNOj KH.PO4 MgS04 Mg(H.P04). MgS04 Mg(N03)a Mg(H.P04). Mg(N03). « LiviNXSTOx, B. E., and Tottingham, W. R., A new 3-saIt nutrient solution for plant cultures. Amer. Jour. Bot. 5:337-346. 1918. 176 BOTANICAL GAZETTE [FEBRUARY They grew Fulcaster wheat for 18 days in 12 different molecular propor- tions of the salt combination II. The nutrient solutions had a total osmotic pressure of 1.75 atmospheres. Of the 12 volume molecular proportions tried the following proved best: KNO3 0.0216, CaH2(P04)2 0.0026, MgS04 0.0150. This finding must not be taken too seriously, however, for it is based on the extremely slim experimental evidence of a single set of cultures with 6 seedlings in each concentration ; also very different molecular proportions (KNO3 o , 0036, CaH2(P04)2 0.0078, MgS04 0.0300) gave almost as good results as this opti- mum and very much better results than the intermediate combinations. Using combination I (Shive's solution) with a total osmotic pressure of 1.75 atmospheres, Shive and Martin' have determined the optimum volume molecular proportions for the development of buckwheat during its early period of growth (24 days) and during its period of maturity. The following table gives the results based on maximum production of dry weight: Yield of top.s. roots. Volume-molecular partial CONCENTRATION THOSE FROM Knop's solu- Tottin^ham's KH,P04 Ca(NO,)» MgSO, tion taken solution as 1 . 00 taken as i . 00 Buckwheat f Water /Tops 0.0144 0.0052 0.0200 I. 61 1-33 grown to cultures! Roots 0.0144 0.0052 0.0200 1.58 1.27 flowering Sand /Tops 0.0144 0 . 005 2 0.0200 1.38 1-25 (early cultures\Roots 0.0180 0.0104 0.0050 1-47 1. 16 growth period) . Buckwheat ' Water cultures- [Tops 0.0108 0.0130 O.OIOO 1 .40 1-35 grown to Roots 0.0108 0.0130 O.OIOO '•5° 1-39 maturity Seeds 0.0108 0.0078 0.0200 1.28 1-74 (late Sand Tops 0.0108 0.0130 O.OIOO 1.40 1.26 growth cultures^ Roots 0.0108 0.0130 O.OIOO 1.27 1.26 period) I ^Seeds 0.0108 0.0130 O.OIOO 1.24 1. 17 It is interesting to see Shive's solution far superior to Knop's and to Tottingham's for growth of buckwheat. In former work by Shive the maximum growth of wheat tops appeared in the following molecular propor- tions: KH2PO4, 0.0180; Ca(N03)2, 0.0052; MgS04, 0.0150. The Ca(N03)2 needs for buckwheat rises greatly in the late growth period, while the KH2PO4 and MgS04 requirements both fall. Many of the excellent methods used in these 3-salt nutrient solution studies were worked up by McCall. McCall and Richards'" summarize their "> Shive, J. W., and Martin, W. H., A comparison of the salt requirements for young and for mature buckwheat plants in water cultures and sand cultures. Amer. Jour. Bot. 5:186-191. 1918. '" McCall, A. G., and Richards, P. E., Mineral food rsquirements of the wheat plant at different stages of its development. Jour. Amer. Soc Agron. 10:127-134. 1917. igig] CURRENT LITERATURE 177 results on wheat in the following table, showing the mean molecular proportions of the 3 component salts and the ionic ratios for the culture solutions giving the best and the poorest growth of wheat during the different periods of development. They promise a full discussion of data with the publication of Series Atfo growth rank Mean molecular proportions in tenths OF total concentration KHaPO, CaCNOj), MgSO. Series i FBest 9 3-4 3-8 35 2.7 2.0 6.1 4.7 .2.0 4.6 1-5 S-S 1.8 1.9 4.2 1.9 5-8 2-5 2 I first period \Poorest 9 Series 2 fBest 9 second period \Poorest 9 . - Series 3 fBest 9 third period \Poorest 9 results of experiments, now being carried on at the Maryland Experiment Station on the nutrient requirements of soy beans and buckwheat. .\11 this work indicates that Sachs, Knop, and others had not nearly exhausted the subject of desirable nutrient solutions for water and sand cul- tures, but that we still have much to learn. The concentration of the nutrient salts used in these solutions are far above those existing in the general soil solution, and upon the whole the immediate significance of this work in prob- lems of soil fertility is not evident. In the soil the dissolving power of well developed root hairs in contact with nutrients of low solubility introduces a new and important feature. It must be stated, however, that Hibbard's work offers indications of an important bearing of this method on questions of soil fertility. — Wm. Crocker. Respiration and age of plant organs. — Nicolas" has studied the respira- tion of very young leaves and leaflike structures in comparison with that of the corresponding fully developed organs taken from older parts of the same plants. In a few cases he included also a comparative study of the respiration of sections of the stem or branch bearing the young and old leaves. His material was selected from 15 diflferent species, including annuals, biennials, and perennials. The studies were conducted by the method of confined atmospheres in absence of light at temperatures from 15 to 24° C, the gas analysis being made with the Bonnier-Mangin apparatus. Respiratory intensities were calculated on the basis of the fresh weight of the materials used. The "internal respiration" also was determined in a few cases, using hydrogen atmospheres. Respiratory quotients CO2/O2 and the ratio I/N between "internal respiration" and normal " Nicolas, G., Contribution a I'etude des variations de la respiration des vege- taux avec I'age. Rev. Gen. Botanique 30: 214-225. 1918. 178 BOTANICAL GAZETTE [February respiration are given, the former varying from o. 58 to i .06 and the latter from 0.13 to 1.38. In every case the young organ, whether leaf, cladode, or branch, had a greater respiratory intensity, a larger respiratory coefhcient, and a lower intra- molecular respiration than the corresponding older organ. This is true whether very young organs or organs in a somewhat later stage of development are compared with fully developed organs of the same year's growth or fully developed organs of the current year's growth are compared with organs of the previous year's growth (leaves of Olea euro pea L.). The magnitude of the differences in respiratory intensity between young and old organs varied rather widely in the different species studied, and was evidently related to the relative differences in age. In general the respiratory intensity of the young organs was from 3 to 7 times that of the older organs. The author reviews the work of previous investigators, all of whom agree that the respiratory intensity of young organs is greater than that of the corre- sponding older organs. Especially interesting in this connection are the studies of Bonnier and Mangin,^^ of Maige,'^ and of Mme Maige/^ as cited by Nicolas in his article. Bonnier and Mangin found two maxima for respira- tory intensity in the seasonal development of a plant, one at the opening of the leaf buds or at the germinative period, the other at the time of flowering. Maige found that, while the respiratory intensity of flowers decreases with age when calculated on either wet or dry weight, it increases when stated in terms of amount of gaseous exchange per individual flower, and Mme Maige pointed out a decrease in respiratory intensity in each organ of the flower except the gynecium, where it sometimes increases with age. The author gives reasons why it is thought that the respiratory differences observed between young and old organs cannot be explained by the absence of well developed cuticle in very young organs, the relative amounts of chlorophyll in the tissues, or the greater acidity of young organs, and raises the question whether they may not be referred to the activities of diastase and oxidase. Finally, the author refers to a previous paper with IMme Maige,^5 in which it was shown that increase in turgescence increases both the respiratory quotient and the respiratory intensity, and concludes that the turgescence of young " Bonnier, Gaston, and Mangin, Louis, Recherches sur les variations de la respiration avec le developpement des plantes. Ann. Sci. Nat. Bot. VII. 2:315-364. 1885. '3 Maige, M. A., Recherches sur la respiration de la fleur. Rev. Gen. Botanique 19:1-28. 1907. '■t Maige, Mme G., Recherches sur la respiration des differentes pieces florales. Ann. Sci. Nat. Bot. IX. 14:1-62. 191 1. '5 Maige, Mme A., and Nicolas, G., Recherches sur I'influence des variations de la turgescence sur la respiration de la cellule. Rev. Gen. Botanique 22:409- 422. 1910. iqiq] CURRENT LITERATURE 179 growing organs is of importance in determining the character and amount of their respiration. The author's conclusions are as follows: "In young organs, principally leaves, intramolecular combustions are more complete than in older organs; young tissues consume much more oxygen than those completely developed, fix relatively less, and thus set at liberty greater quantities of energy, which they use in growth." — G. T. Harrington. Catalase, respiration, and vitamines. — Dutcher'* finds that the catalase activity of polyneuritic pigeons is ver>' low, and that it rises to normal W'hen the fowl is fed water soluble vitamine. His results are given in the following table: Catalase activity of tissues Tissue _ Polyneuritic pigeons, percent- age of normal < Polyneuritic pigeons receiving water soluble I vitamine, per- centage of normal Liver Kidney. ... Pancreas. . . Heart Breast Lung Blood A\erage no 102 "5 86 152 84 56 lOI The author says: "It is probable that polyneuritis is accompanied by in- complete or partial oxidation, with accumulation in the tissues of products of incomplete oxidation. It is also probable that water soluble vitamines func- tion, directly or indirectly, in stimulation of oxidation processes, thereby clear- ing the tissues of toxic materials. When pigeon tissues are arranged in the order of their catalase content (as measured by the oxygen liberated from hydrogen peroxide), tissues group themselves in the order of their metabolic activity and also in the order of their content of water soluble vitamine." Appleman'7, in a recently published article, says: "Respiration in sweet corn in the milk stage is very high when the corn is first puUed. This high rate of respiratory activity falls off rapidly with storage. Catalase activity in a collateral set of ears showed a decline with storage, which is almost directly proportional to the decline in respiratory intensity after a like period of storage. The catalase activity of the expressed juice from both sweet corn and potato '^ Dutcher, R. Adams, Vitamine studies. I. Observations on the catalase activ- ity of tissues in avian polyneuritis. Jour. Biol. Chem. 36:63-72. 1918. '" Appleman, C. O., Respiration and catalase acti\-ity in sweet corn. Amer. Jour. Botany 5:207-209. 1918. i8o BOTANICAL GAZETTE [February tubers is a fair index of the comparative intensity of respiration in the tissues. The data from both plant and animal tissues available at the present seem to justify the general indication that catalase action is invariably correlated with the oxidative processes involved in respiration." — Wm. Crocker. Respiration of stored wheat. — Bailey and Gurjar'^ have done an excel- lent piece of work on the respiration of stored wheat. Significant literature is well presented and related to the work in hand, and the methods used in the work are clean cut and exact. The contribution has a very important applica- tion in the shipping and storage of grains. They worked with moisture con- tents ranging from 12 to 18 per cent, such as appear in the practical handling of grains. The following are the more important results. Respiration gradually and fairly uniformly rises with moisture content up to 14 . 5 per cent in case of plump spring wheat. With the rise of moisture above this percentage the respiration is markedly accelerated. The soft starchy wheats respire more rapidly than the hard vitreous wheats containing the same percentage of moisture. With more than 14 per cent moisture shriveled wheat respires 2 to 3 times as fast as plump wheat of the same water content, due to a larger percentage of embryo in the shriveled grains; with less than 14 per cent moisture there is little difference. Freshly dampened wheat respires more slowly than wheat of the same water content that has been dampened for a long time or that has been naturally dampened. The difference is noticeable at 13 per cent moisture, and rises as the moisture rises. Wheat stored at room temperature respires more rapidly than that of the same moisture content at lower out-door temperatures. Un- soundness of wheat caused by the freezing of unripe plants increases respira- tion. This is attributed to the accumulation of glucose in the frosted grains. Increased temperature inci;eases the respiration up to 55° C. When seeds are stored in closed chambers and the respiration taken by 4-day periods, the rate is highest for the first period and diminishes materially in successive periods as the carbon dioxide content rises. The respiration is also reduced in an oxygen free atmosphere, the ratio to that occurring in a normal atmosphere being about 1:2.5. Many will think the author's evidence for their viscosity conception of limited respiration is insufficient. They will also question whether the amount of glucose present limits respiration when low moisture has already run respira- tion to so low an ebb. — Wm. Crocker. Relation of host and parasite among fungi. — An excellent service has been rendered by Reed'^ in bringing together the extensive and scattered data regarding the susceptibility of more or less related hosts to physiological strains '* Bailey, C. H., and Gurjar, A. M., Respiration of stored wheat. Jour, .\gric. Research 12:685-713. igi8. '9 Reed, George M., Physiological specialization of parasitic fungi. Mem. Brooklyn Bot. Gard. 1:348-409. 1918. 1919] CURRENT LITERATURE 181 among various fungi. Some 68 species of fungi, the majority of them belonging to the Uredinales, have been reported to show such specialization. The first known and best studied species is Puccinia graminis, producing the destructive stem rust of wheat and of other cereals and grasses. A few species having a wide range of hosts, like P. subnitens, appear not to be specialized. The cita- tion of literature includes 174 titles, supplied by 67 writers, indicating the prominence which this line of investigation has attained within the last few years. Eriksson's studies on the specialization of the grain rusts, reported in 1894, introduced the subject, but the fixed and unchanging character of physio- logical strains has first been shown definitely in the present paper, since being confirmed by Stakman and others.^" It is pointed out that so far the data do not indicate that bridging species are capable of altering the physiological nature of the parasite so as to enable it to extend the range of its natural hosts, as has heretofore been assumed. In fact, it appears that among fungous parasites there are definite strains or races not distinguishable morphologically, but only by their physiological behavior in infecting certain hosts, and that these strains retain the same characters through all the metamorphoses of the fungus, and when tested by use of any kind of reproductive body that the particular species produces. The specializa- tion of the same fungus in widely separated regions may possibly be different, but the data are scanty. The relation of physiological specialization to mor- phological variation is barely mentioned. The whole subject of specialization is one of great scientific and economic interest, making the present admirable summary particularly timely. — ^J. C. Arthur. Heath and grassland. — Continuing the investigations already noted" of certain English heaths and grassland. Farrow" has accumulated more data upon the eflfects of a rabbit population upon vegetation retrogression. It is demonstrated that the presence of rabbits alone is sufficient at times to change a pine forest through Calluiia heath and Carex arenaria associations to a dwarf grass or a Cladonia heath. Experiments with irrigation and with the applica- tion of manure tend to show that both sterile soil and lack of soil moisture are factors in limiting the rate of growth and the luxuriance of the vegetation. This increased growth with imt)roved conditions results in a decrease in the number of species in the area, since the more rapid growth of certain plants, like Agrostis vulgaris, smothered less vigorous ones, such as Festuca ovina. 20 Stakman, E. C, Parker, J. H., and Piemeisel, F. J.. Can biologic forms of stem rust on wheat change rapidly enough to interfere with breeding for rust resist- ance? Jour. Agric. Res. 14:111-123. ph. 13-17. 1918. ^' BoT. Gaz. 64:263. 1917. " Farrow, E. P., On the ecology of the vegetation of Breckland. III. General effects of rabbits on the vegetation. IV. Experiments mainly relating to the available water supply. V. Observations relating to competition between plants. Jour. Ecology 5:1-18, 104-112, 155-172. 191 7. l82 BOTANICAL GAZETTE [February Evidence is also presented that such plants as Pleris aquilina and Pinus often succeed in competition owing to their dead foliage excluding the light from their competitors, causing etiolation and decay. In a more recent paper Farrow^^ has examined the retrogression begun by rabbits and continued by sand blasts. This retrogression shows exactly the reverse order of the succession inaugurated by irrigation, being particularly noticeable in the Agrostis vulgaris giving place to Festuca ovina wherever the sand blast became intensive. Once begun, bare areas tend to increase, the sand assisting in destroying the vegetation both by direct attack and by remov- ing the substratum, leaving clumps of grass upon the tops of small hummocks which are being constantly undermined. With the checking of wind erosion in such bare areas Polytrichum and Cladonia become agents of stabilization and revegetation. — Geo. D. Fuller. Photosynthesis. — Osterhout and Haas^'' summarize as follows a piece of work on the dynamics of photosynthesis. " Ulva which has been kept in the dark begins photosynthesis as soon as it is exposed to sunlight. The rate of photosynthesis steadily increases until a constant speed is attained. This may be explained by ass,uming that sunlight decomposes a substance whose products catalyze photosynthesis or enter directly into the reaction. Quanti- tative theories are developed to account for the facts." The rate of photo- synthesis was determined by the rate at which a portion of Ulva rendered sea water basic to phenolphthalein. Since the dissociation of carbonic acid is very slight, change of reaction is a very crude way of measuring the amount present. There is also the possibility of other exchanges of more strongly dissociating materials that could modify the reaction of the water. In the face of excellent and very accurate methods for the quantitative determination of carbon dioxide it seems hardly justifiable to use this questionable method for a study of either respiration or photosynthesis. It is also doubtful whether sufficient regard has been given to other possible limiting factors of the rate of photosynthesis in these experiments. If, in spite of the defects of experimen- tation, the general conclusion proves true, it is a contribution of great signifi- cance and aids in confirming Willstatter's view that the presence of a catalyzer is a common internal limiting factor to the rate of photosynthesis. — Wm. Crocker. Organic plant poisons. — Brenchley^^ finds hydrocyanic acid very toxic to pea and barley seedlings in water cultures. Hydrocyanic acid in concentra- tions of I part to 100,000 proved rather quickly fatal for peas and somewhat '3 Farrow, E. P., On the ecology of the vegetation of Breckland. V. Character- istic bare areas and sand hummocks. Jour. Ecology 6:144-152. 1918. '4 Osterhout, W. J. V., and Haas, A. R. C, Dynamical aspects of photosynthesis. Proc. Nat. Acad. Sci. 4:85-91. 1918. ^5 Brenchley, Winifred E., Organic plant poisons. I. Hydrocyanic acid. Ann. Botany 31:447-456. 1917. iQig] CURRENT LITERATURE 183 less toxic for barley. Dilutions as great as i part to 4,000,000 to 10,000,000 proved somewhat toxic. Hydrocyanic acid showed no stimulation and the cyanogen radicle is the toxic agent. Brenchley^ has also studied the effect of various phenols (phenol o-cresol, m-cresol, p-cresol, resorcinol, pyrocatechol, pyrogallol, phloroglucin, orcinol) upon the growth (as indicated by increased dry weight) of barley and peas in water cultures. The purpose was to learn the direct effects of these phenols on the plants, so that it could be considered in using the phenols as partial soil sterilizers. The following concentrations were used: M/ioo, M/100X1/5, M/IooXI/5^ and M/iooXi/6^ The general physiological effect was the same for all the phenols, but the concentration at which these effects showed varied considerably with the different members. The highest concentration was quickly fatal with all the phenols, and the next to highest concentration with o-cresol, pyrocatechol, and pyrogallol, but there was a slight recovery in the others. The lowest concentration showed no injury in any. None of the solutions showed any stimulus effect in any concentrations. — Wm. Crocker. Regeneration in Phegopteris. — Miss Brown^' has recorded the results of some experiments on regeneration in Phegopteris polypodioides. Near the base of the petiole of a detached leaf regeneration took place in contact with sand moistened with Knop's solution in moist air. A prothallium-like growth appeared, and from this were developed rhizoids, structures intermediate between leaves and prothallia, and true leaves. The possible determining factors are enumerated, and among them the separation of the leaf from the parent body was evidently necessary; at least it seems evident that "some phase of nutrition must be an important factor in regeneration, if not the most important factor." — ^J. M. C. Selaginella. — Van Eseltine^* has begun a series of contributions dealing with the American species of Selaginella allied to 5. rupestris. The group is in need of critical revision, and the results will be of interest to the morphologist as well as the taxonomist. The first paper deals with the representatives of the group occurring in the Gulf Coastal Plain and the territory immediately adjacent to the northeast. In this region 8 such species are recognized, 2 of which are described as new, and an additional one was described by the same author recently. The numerous drawings and photographic plates supple- ment well the full descriptions. — J. M. C. ^*Brenchley, Winifred E., Organic plant poisons. II. Phenols. Ann. Botany 32:259-278. 1918. "7 Brown, Elizabeth W,, Regeneration in Phegopteris polypodioides. Bull. Terr. Bot. Club 45:391-397. figs. j. 1918. ** Van Eseltine, G. P., The allies of Selaginella rupestris in the southeastern United States. Contrib. U.S. Nat. Herb. 20:159-172. pis. 15-22. figs. 6^-70. 1918. 1 84 ■ BOTANICAL GAZETTE [February Vegetation of a glacial plunge basin. — In certain rock basins of glacial origin near Syracuse, New York, low soil and air temperatures prevail through- out the year, the difference between the rim and bottom of the depressions often amounting to 30° F. These temperature depressions have been shown by Petry^9 to be the controlling factors in the development of plant associa- tions characterized by distinctly northern species, such as Cornus canadensis, Pyrola asarijolia, Coptis trifolia, and Ribes lacustre, whose local distribution coincide exactly with areas of low soil and air temperature. — Geo. D. Fuller. Geotropism and phototropism. — Van Ameijoen-^" finds that neither geo- perception nor photo-perception or reaction occurs in the seedlings of Avena saliva or Sinapis alba in complete absence of oxygen. Contrary to Correns and Kenkel, he finds that, on complete or partial withdrawal of oxygen, the reaction of seedlings to a geotropic stimulus does not differ from their reaction to a phototropic stimulus. — Wm. Crocker. Rusts of Costa Rica. — Arthur^' has studied the rusts of Costa Rica based chiefly upon collections made by Holway, and this first presentation of Costa Rican rusts includes 118 species, 22 of which are described as new, and 12 others are new to North America. The indications are that the rust flora of Costa Rica will be found to be of exceptional richness and importance. — J. M. C. Aquilegia. — Payson^^ has published a revision of the North American species of Aquilegia. In addition to the keys, descriptions, and discussions, there is an unusually full list of stations. He recognizes 25 species, 3 of which are described as new, and also 9 subspecies or varieties, 2 of which are new. — J. M. C. New African plants. — Moore,33 in connection with his studies of African Compositae, has described a new genera {Emiliella) of the Senecionidae and 8 new species of Senecio. — J. M. C. ^' Petry, Loren C, Studies of the vegetation of New York State. II. The vege- tation of a glacial plunge basin and its relation to temperature. Bull. Torr. Bot. Club 45:203-210. 1918. 30 Van Ameijden, U. P., Geotropism and phototropism in the absence of free oxygen. Recueil Trav. Bot. Neerl. 14:149-218. pis. 15-19. fig. i. 1917. 3' Arthur, J. C, Uredinales of Costa Rica based on collection by E. W. D. HoLWAY. Mycologia 10: 111-154. 1918. 3^ Payson, Edwin Blake, The North American species of Aquilegia. Contrib. U.S. Nat. Herb. 20:133-157. pis. 8-14. 1918. 33 Moore, Spencer LeM., Alabastra diversa. Part XXIX. Jour. Botany 56: 225-233. 1918. Cambrttjge ©nitoersitp ^nss On Growth and Form. By D'arcy W. Thompson, C.B., D.Litt., F.R.S. With 408 illustrations. Demy 8vo. $6.50. "Professor Thompson has written an interesting and valuable book on a topic of enduring interest Why does the animal kingdom present to us a picture of such infinite variety? .... And why, running through the diversities, are there so many resemblances ? . . . . The purpose of Professor Thompson's book is to explain the 'must be' of structure." — Times. Fossil Plants. A Textbook for Students of Botany and Geology. By A. C. Seward, M.A., F.R.S. , Professor of Botany in the University of Cambridge. Volume III. Pteri-- dospermeae, Cycado61ices, Cordait'ales, Cycadophyta. With frontispiece and 253 illus- trations. Demy 8vo. $5.50. Cambridge Biological Series. British Grasses and Their Employment in Agriculture. By S.F.Armstrong, F.L.S., of the School of Agriculture, Cambridge. With 175 illustrations. DemySvo. $2.00. Plants Poisonous to Live Stock. By Harold C. Long, B.Sc. (Edin.), of the Board of Agriculture and Fisheries. With frontispiece. Royal 8vo. $2.00. Cambridge Agricultural Monographs. Science and the Nation. Essays by Cambridge Graduates, with an Introduction by the Right Hon. Lord Moulton, K.C.B., F.R.S. Edited by A. C. Seward, F.R.S., Master of Downing College, Cambridge. Second impression. Crown 8vo. $1.50. Published by the Cambridge University Press, England Q. P. Putnam's Sons, Agents in the United States 2, 4 and 6 West 45th Street, New York, city The Anatomy of Woody Plants (With 306 figures and illustrations) By E. C, JEFFREY Professor of Plant Morphology in Harvard University The author is easily the leading authority on the aiiatomy of plants. His book aims to describe, analyze, and discuss the anatomical structure of the most important families of plants. It does for the anatomy of plants what so many textbooks have done for the human anatomy. The book will be of great service, not only to botanical instruction everywhere, but also to paleobotany, for the study of vascular anatomy has revolutionized that subject. x-{-4^8 pages, cloth; $4.00, postage extra {weight 2 lbs. 14 oz.) THE UNIVERSITY OF CHICAGO PRESS CHICAGO ILLINOIS The Home Typewriter as well as The Business Typewriter 15 the Self-Starting All the family can tts'e'it. Write for descriptive matter. REMINGTON TYPEWRITER COMPANY Incorporated 374 BROADWAY NEW YORK CITY FINE INKS AND ADHESIVES For those who KNOW Higgins' Drawing Inks eternal Writing Ink Engrossing ink Taurine Mucilage Photo Mounter Paste Drawing Board Paste Liquid Paste Onice Paste Vegetable Qlue, Etc. Are the Finest and Best Inks and Adhesives Emancipate yourself from the use of corrosive and ill-smelling inks and adhesiyes and adopt the Higglns Inks and Adhesives. They will bea revelation to you, they are so sweet, clean, well put up, and withal so eflScient. At Dealers Qenerally CHAS. M. tllQQINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street Brooklyn, N.Y. AN ANNOUNCEMENT ^ Our activities for the past year have been nearly 100 per cent devoted to war production of vital importance to the government, compelling us to discontinue regular manufacture in practically all of our lines. ^ The cessation of hostilities has now enabled us to begin the readjust- ment to a peace production basis. While this adjustment must be gradual, we are pressing it with all possible speed in order to satisfy the requirements of our normal trade at an early date. ^ We appreciate the patience and understanding with which our patrons have reacted to the situation and trust they may continue to exercise such consideration until our manufacturing facilities are once more on a pre-war footing. Bausch X^ [pmb Optical (g. 41 2 ST. PAUL STREET ROCHESTER, N.Y. NEW YORK CHICAGO SAN FRANCISCO WASHINGTON Leading American Makers of Microscopes, Projeclion A pparatui (Balopticons J , Photographic Lenses, Ophthalmic Lenses and Instruments, Range-rinders and Gun Sights /or Army and Navy, Searchlight Reflectors, Binoculars, and Other High-Grade Optical Products » f *^' '\ • Volume LXVII Number 3 THE Botanical Gazette Editor: JOHN M. COUU'ER MARCH 1919 Relation of Minimum Moisture Content of Subsoil of Prairies to Hygro- scopic Coefl5cient - - F. J. Alway, G. R. McDole, and R. S. Trumbull 185 Notes on North American Trees. IV - - - - C. S. Sargent 208 Possible Correlations Concerning Position of Seeds in the Pod Byron D. Halsted 243 Embryo and Seedling of Dioon spinulosum - Sister Helen Angela Dorety 251 (With Plates X, XI) Development of Strcpharia epimyces - - - - W. B. McDougall 258 (With ten figures) r Briefer Articles Hybrid Perennial Sunflowers - - - - T. D. A. Cockerell 264 Relationships within the Rhodosporeae - - - - Geo. F. Atkinson 266 Current Literature Minor Notices --------___ 268 Notes for Students - - - - - - - _ _ -269 The University of Chicago Press CHICAGO, ILLINOIS, U.S.A. THE CAMBRIDGE UNIVERSITY PRESS, London and Edinburgh THE MARUZEN-KABUSHIiCMCAISHA. Tokyo. Osaka. Kyoto. Fukuoka. Sendai THE MISSION BOOK COMPANY. Shanghai Volume LXVII Number 3 The Botanical Gazette A MONTHLY JOURNAL EMBRACING ALL DEPARTMENTS OF BOTANICAL SCIENCE EDITED BY JOHN M. COULTER With the assistance of other members of the botanical staff of the University of Chicago Issued March 18. I9!9 The Botanical Gazette is published monthly by the University of Chicago at the University Press, 5750 Ellis Avenue, Chicago, HI. "[[The subscription price is $7.00 per year; the price of single copies is 75 cents. Orders for" service of less than a half-year will be charged at the single-copy rate. ^Postage is prepaid by the publishers on all orders from the United States> Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawaiian Islands, Phihppine Islands, Guam, Samoan Islands, Shanghai. ^[Postage is charged extra as follows: For Canada, 35 cents on annual subscriptions (total $7.35), on single copies, 3 cents (total 78 cents); for all other countries in the Postal Union, 84 cents on annual subscriptions (total $7.84), on single copies II cents (total 86 cents). ^[Patrons are requested to make all remittances payable to The University of Chicago Press in postal or express money orders or bank drafts. The following are authorized to quote the prices indicated : For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. 4. Yearly subscriptions, including postage, £1 12^. 6d. each; single copies, including postage, 3J. 6d. each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage, Yen 15.75 each; single copies, including postage. Yen 1.75 each. For China: The Mission Book Company, 18 Peking Road, Shanghai. Yearly sub- scriptions, $7.00; single copies, 75 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 84 cents, on single copies II cents. - Claims for missing numbers should be made within the month following the regular month of publication. The pubUshers expect to supply missing numbers free only when losses have been sustained in transit and when the reserve stock wiU permit. DeUveries are not guaranteed within the war zones of Europe, nor in other countries where transportation is disturbed by con- ditions incident to war. Business correspondence should be addressed to The University of Chicago Press, Chicago, lU. Communications for the editor and manuscripts should be addressed to the Editor of the Botanical Gazette, The University of Chicago, Chicago, 111. Contributors are requested to write scientific and proper names with particular care, to use the metric system of weights and measures, and in citations to follow the form shown in the pages of the Botanical Gazette. Papers in excess of thirty-two printed pages are not accepted unless the author is willing to pay the cost of the additional pages, in which case the number of the pages in the volume is correspondingly increased. Illustrations are furnished without cost to author only when. suitable originals are supplied, A copy of the suggestions made in the January number, 1907, wUl be sent on appUcation. It is advisable to confer with the editors as to illustrations required in any article to be offered. Separates, if desired, must be ordered in advance of publication. A table showing approxi- mate cost of separates is printed on an order blank which accompanies the proof; a copy will be sent on request. Entered as second-class matter August 21, 1896, at the Post-Office at Chicago, Illinois, under the Act of March 3, 1879. AccepUnce for mailing at special rate of postage provided for in Section 1 103, Act of October 3, 1917, authorized on July IS, igrS, VOLUME LXVII NUMBER 3 THE Botanical Gazette MARCH igig RELATION OF MINIMUM MOISTURE CONTENT OF SUBSOIL OF PRAIRIES TO HYGROSCOPIC COEFFICIENT^ F. J. Alway, G. R. McDole, and R. S. Trumbull Introduction It has long been recognized that the maxima and minima percentages of water found in well drained soils in the field are in general roughly dependent upon the relative fineness of texture, but very few data have been published in such form as to permit of any attempt to compute the actual relations which these extremes bear to the physical constants. In a previous paper (3) we have reported laboratory experiments and field observations showing that when loams, after rains sufficiently heavy to moisten them thoroughly, are protected from losses by evaporation and tran- spiration, they lose water by downward movement until the ratio of moisture content to hygroscopic coefiicient reaches a value between i . 8 and about 2.5; while with coarse sands the ratio is as high as 6.0 or 7.0; and fine sands occupy an intermediate position, the ratio rising with a decrease in hygroscopicity. While the maxima under field conditions are easily ascertainable anywhere, it being necessary only to await heavy rains or to irrigate a small area, the corresponding minima are developed only when a very scanty rainfall or a prolonged absence of precipitation is ' The work reported in this paper was carried out in 1907-1913, while the authors were members of the staff of the Nebraska Agricultural Experiment Station. 18s l86 BOTANICAL GAZETTE [march accompanied by weather conditions which stimulate both evapora- tion and transpiration, and so favor a reduction of the soil moisture content. These favoring conditions are high temperature, low atmospheric humidity, high wind velocity, and a high degree of insolation. Even in a semi-arid region several years may pass without the concurrence of the necessary conditions, while in humid regions such intervals are of still greater length. The results of the greenhouse experiments of Briggs and Shantz (6) would suggest that after periods of extreme drought the prairie subsoil might be expected to show a moisture content which either approximated the so-called wilting coefficient (1.47 times the hygroscopic coefficient) or which was somewhat lower than the former value but bore no distinct relation to the latter. In the opinion of these authors the wilting coefficient "practically marks the cessation of growth," and after this point has been reached the soils continue to lose water through the tissues of the plants, even after they are dead, the final moisture content of the soil being as low as though the soil and air had been in direct contact. However, pot experiments of any kind, and especially those employing shallow vessels, appear ill adapted to answer the question as to how dry a particular soil may become under field conditions. Accordingly the data obtained in the field at such times as when the weather conditions have been favorable to an extreme reduction of the subsoil moisture should prove of especial interest, provided they are accompanied by determinations of the hygro- scopic coefficient or wilting coefficient of the soils. In regions of winter rains and summer droughts, such as Cali- fornia, one may safely count upon the continuance of hot rainless weather with clear skies for many weeks after the dry season has once set in, but in those with summer rains one never knows when to make preparations for studies that are dependent upon extreme brought conditions, and it may happen that, after all arrangements have been completed, several years may pass before weather con- ditions favoradle for the work occur. For this reason data on the soil moisture content during unusual droughts in humid regions are most apt to be secured in the course of some less specialized investigation. iQig] ALWAY, MCDOLE, b- TRUMBULL—SUBSOIL 187 Location of prairies sampled The fields from which we secured the following data form two groups, one in the southwestern corner of Nebraska adjacent to the towns of McCook, Wauneta, Imperial, and Madrid; and the other close to the Nebraska Experiment Station at Lincoln. As all the former were at a distance of 200 miles or more from the experiment station, the sampling of them was feasible only at long intervals, and many of the data from these were secured incidental to the collection of samples for chemical studies. The western group of fields is beyond question well within the semi-arid region, while those at Lincoln He ahnost as far to the west as the strictly humid climate extends on the American prairies. Some may consider that even Lincoln falls within the eastern limit of the great semi- arid region, but the composition of the soil (4, p. 414), the growth of vegetation and its agricultural history, as well as the moisture conditions of the subsoil distinguish it from the drier country west of Holdrege. All the factors which determine the difference in climate alter so gradually from east to west that it is impossible to place any definite line of demarcation between the humid and the semi-arid regions, the most that we are justified in assuming being that for every advance of a few miles to the westward of Hastings there is a nearer approach to strictly semi-arid conditions. At Hastings we appear to be still within the humid region, while at Holdrege, 50 miles farther west, most of the characteristics of semi- arid regions are discernible. Also the distribution of carbonates in the subsoil indicates that the district between Hastings and Holdrege is the region of most rapid transition (4, p. 414). Favorable weather conditions The weather of the period covering our work proved extremely favorable for the development of dry subsoils in both localities. At Lincoln it included the driest two-year period (1911-1912) of the past 20 years, 1897 to 1916, although in two years, 1895 ^i^d 1 901, there had been a lower annual precipitation than in either of these (table I) . Accordingly the soil moisture conditions we found there may be considered to include those representing the eft'ects of extreme drought. i88 BOTANICAL GAZETTE [march In southwestern Nebraska our work was begun in seasons, 1907 and 1908, forming the conclusion of a series of wet years. This was followed by a prolonged dry period, reaching its climax in TABLE I Relation of annual precipitation at Lincoln, year by year, to normal (=100), 27.51 INCHES, showing RELATIVE DRYNESS OF PERIOD OF observation (1906 TO 191 2). Year Percentage Year Percentage Year Percentage Year Percentage 1895.... 60 1901 .... 80 1907 99 1913- ••■ 95 1896 138 1902. . . . 150 1908 130 1914 145 1897.... 93 1903.... 126 1909... 126 1915.... 128 1898 102 1904. . . . lOI 1910. . . . 114 1916. . . . 80 1899 1900. . . . 82 1905.... 1906. . . . 129 124 1911. . . . 1912. . . . 89 81 123 1910-1911, the precipitation in 1910 being the lowest recorded since observations were begun at North Platte 42 years ago (table II). By 191 1 the subsoil moisture had probably been reduced to as TABLE II Annual precipitation in inches at stations in southwestern Nebraska Length of record in years* McCook 16 Wauneta 27 Imperial 26 H. O. Ranch II North Platte 42 Normal 1905 1906 1907 1908 1909 1910 1911. . .' 1912 1913 1914 1915 1916 Maximum. . Minimum. . . *To end of 1916. 19.71 33-97 20.59 19.32 18.08 22.54 9-34 12. 15 14.69 18.26 18. 24 30 -95 15-35 33-97 9-34 18.70 32.24 22.82 20.18 24.77 18.46 13-82 18.82 20.00 16.05 17.26 27.04 14-95 32.24 13-82 20. 79 33-05 26. 23 16. 76 26. 27 20.03 11.77 17-37 24.58 16.60 16.94 37-14 19-33 37-14 11.77 17.80 20.14 12.02 21 .01 16.89 7.62 12. 76 20.74 14.99 19.42 35-84 14.60 35-84 7.62 18.88 26.81 27.99 19.61 19.96 22.41 10. 70 17-43 18.69 19. 10 15-79 32.70 12.96 32.70 10.70 low a point as is ever experienced in southwestern Nebraska. The climate of the Nebraska portion of the Transition Region, including both groups of fields, has been discussed in some detail in a previous paper dealing with the composition of its loess soils (2). iqiq] ALWAY, McDOLE, b- TRUMBULL—SUBSOIL 189 Experimental methods In taking the samples we used augers provided with extensions, commonly employing two sizes, one i . 5 inches in diameter with which to take the sample, and another 2.0 inches in diameter to enlarge and clean out the hole preparatory to sampling the next lower section. In many of the borings in western Nebraska the subsoil in part or in all the levels sampled was too dry to be remov- able by the ordinary auger, sliding off the bit as this was being withdrawn. In such cases we employed a Tinsley "auger with casing" (11), the sleeve on this retaining the soil loosened by the bit. Except where otherwise indicated, the samples were com- posites from 3 borings 10-20 yards apart. Composites were made from the first 3 borings only where it could be seen from the behavior of the soil toward the auger that the general moisture conditions in all 3 were similar, but not necessarily identical. Extremes under semi-arid conditions After a prolonged drought.— As already stated, 1910 proved the driest year in southwestern Nebraska since observations were begun, the precipitation amounting to scarcely half the normal (table III) . The autumn of this year and the following winter and spring were practically without snow or rain until April, the total precipitation at McCook from the end of August 19 10 to the first of the following April amounting to only i . 60 inches (table IV), and this fell in such small amounts as to influence the soil moisture content through only a negligible distance. At Imperial and Wauneta the weather was not quite so dry, but the difference was not sufficient to cause an appreciable difference in the moisture content of the subsoil. The samplings made in fields near McCook and Wauneta near the end of October 19 10 showed such low ratios of moisture content to hygroscopic coefficient (table V) that some undiscovered source of error was suspected, and for this reason 6 weeks later we resampled two of them, A and B at McCook, and took sets from two additional fields. These confirmed the correctness of the extremely low ratios. The concordance of the moisture content with the hygroscopic coefficient was very striking, as though the 190 BOTANICAL GAZETTE [march plant roots, while not recognizing the wilting coefficient, practically ceased to withdraw water as soon as the hygroscopic coefficient had been reached. There was little difference in moistness between TABLE III Monthly precipitation in inches at McCook, Wauneta, Imperial, and THE H. O. Ranch, showing dryness of seasons Month Normal McCook Wauneta Imperial 1910 McCook Wauneta Imperial H. O. Ranch January. . . February. . March April May June July August .... September . October November . December. . Annual 0.21 0.62 0.73 1.89 2.82 3-29 3 09 2.SS 1.72 1.03 0.56 O.S7 26 69 03 04 54 34 47 74 1-3S I 13 0.39 0.57 0.44 0.69 1-33 2.27 2.82 3-34 2.91 2-73 1-34 1. 10 0.50 0.72 0.0 0.0 0.0 0.76 2.77 1 .12 0.0 0.0 0.0 0.82 2.00 •44 70 93 72 17 o 17 77 64 20 o 10 0.8s 0.40 O. 10 0.38 0.71 1.98 2.51 0.72 2.82 i.S8 T T 0.57 19. 08 18. SS 20. 19 9-34 13.82 11.77 35 06 36 60 25 24 0.34 0.97 1.28 0.0 0.03 0.14 7.62 Month 1911 1Q12 McCook Wauneta Imperial H. 0. Ranch McCook Wauneta Imperial H. 0. Ranch January February. . . March April May June July August September. . October .... November. . December. . 0.05 0.47 0.12 1.72 1-25 0.66 0.84 4-34 0.59 0.96 0.05 1. 10 O.IO 0.70 0.50 3-4S I -75 1-35 130 307 1.80 330 0.0 ISO 0.42 0.37 0. 22 2-55 2.19 1.29 1. 10 3-4S 1.44 2.92 O.IO 1.32 O.IS 0.31 0.03 1.94 1.29 0.92 0.84 1-97 0.86 2.8s 0.05 1-55 0.0 0.24 i-SO 2.01 0.0 2.77 2.29 2. II 2.13 1.09 0.50 T 0.20 l-Sl 2-35 2.82 0-95 1.89 3.26 2.78 2.61 1-43 0.20 0.0 0.52 1.08 3-61 2.85 1. 41 1.82 . 509 4.28 2.01 i-SS o-iS 0.21 O-IS 1. 00 2.70 1-74 IS3 1.80 S-16 2.49 2.41 1-45 0.0 0.05 Annual. 12.15 18.82 17-37 12.76 14.69 20.00 24. S8 20.44 the surface foot and the succeeding 2 or 3 ft., and even the deeper subsoil was but little if at all moister. At no level and in none of the fields was there any growth water, the moisture content being below the computed wilting coefficient, which corresponds to a ratio of 1.47, or approximately 1.5 (6). I9I9] ALWAY, MCDOLE, & TRUMBULL— SUBSOIL 191 m < o- o 2 O H O (=] n H o « ui o o U u < z o H < H a o < c 3 >— > t/5 W lO fO '. t^ 0 . ^ M . CO N 0 • • • ■ 0 0 0 0 • 0 0 • 0 0 0 0 d < M 0 . m • •°s • ■ • ■ 00 • • 0 • - 0 . M . • • 0 Vi M Ph"" ,~ ^n X :: ^ ::::::::: : : J3 d ••:::;■ :: •^ ::::::::::;: .^ ::?/:: : : : : :h : C 0 d M M a' 0 0 M M : : :h . 0 . . 10 . • 0 ■ • 0 • 0 d ■ ■ : : : : ; : : ::::,„:::::::::;::::: ! ! I !H ! I ! ! •tH • . 0 • • • • . . . . . . :' ::;; 0 ;;;■•••;••■•■- • : : : : : : : : : :;:: ° :::::::::::..: . u 0 « 0 • • • do-' d : : : : :^ : : : 4-> a t/2 0 ID d * • • M • • ■ M . OS • • M • . -CO . . C4 • - 0 • 0 • • ■ 0 ti 3 < ^ t^ I r«. ro N 0 • • : : : :5 1 to • • • 0 to 4 - 0 • 0 0 0 0 - • ■ 0 0 • • ; • 0 : ° : • - 0 "a 1—1 . . . r* fN. . . . r^ . . . 0 0 ■ ■ • N 0 g : ; 00 d ... 0 0 ... 0 0 0 • • 0 • c 3 1—1 0 0 d . •\o • . fO CI . . . .00 • • ; • • • • 0 0 • • . . 0 ■ • >> ID c» w a < M • * *f^ n * * * ' : 0 ; M d fa 4 . d • . v,CO • c" to 0 d Y^ '''.:'.::. I '.'.'.'.'.', \ '.'. \ '.'.'.'.'. I '^ •. '• 0 s M d ^ . . . . . . > 0 0 d 4J u 0 M d • r* . . . . ir> • . . • d . . . . 0 - • • ■ • ■ 0 • 0 . •• . M N re -^ irjvO t^OO 0 0 "-I (^ M M M tH H HI 0 0 M M TRUMBULL— SUBSOIL 195 surface soil. January and February together had a precipitation somewhat below normal at McCook, but above at both Wauneta and Imperial, while March at all 3 places had a precipitation 2 or 3 times the normal. April was rainless until the 20th, between which date and the 28th from 2 to 3 inches of rain fell. Samphng was carried out at McCook on May 7 and 8, no rain having fallen since April 28; at Imperial on May 11, 13, and 14, 0.45 inch having fallen there in 4 light showers; and at Wauneta on May 16 and 17. At the last place the only rain since April 28 had been one of o. 10 inch on May 10. Thus conditions had been ideal for the downward movement of the water into the subsoil, while at each place an interval of 8-19 days had elapsed between the last good rain and the date of sampling. The generally favorable weather of autumn, winter, and spring was evidenced by the circumstance that in the early spring the outlook appeared unusually promising for the farmers. Wheat had come through the winter in fine condition and preparations were being made for seeding a large acreage to spring grains, the prospects being considered so favorable that local merchants were willing to furnish seed grain in return for a reasonable share of the crop. Conditions appeared ideal for a study of the degree to which the ratio in the surface soil had to be raised before water could pass downward into the deeper portions of the subsoil, where during the previous year the moisture had been reduced to the hygroscopic coefficient or even slightly below. In the fields with heavier soil we found that the moisture content had been distinctly affected at McCook (table VII) to only 2 ft., at Wauneta in the one field to 3 ft., in the other to 4 ft. or more, and in the only one sampled at Imperial to 5 ft. In normal seasons. — That the low ratios prevaihng throughout the subsoil of the prairies after severe droughts, as illustrated in the preceding tables, are not entirely absent even in favorable seasons, may be seen from table VIII reporting conditions at the H. O. Ranch. There, as at McCook, Wauneta, and Imperial, after periods of drought the ratio was found not far from i . o at all depths, while under more favorable conditions, as in July 1908, the low ratio was still to be found at some level within the first 6 ft. 196 BOTANICAL GAZETTE [march While after protracted droughts and probably also after extremely wet periods the moisture conditions in the subsoil are quite uniform, they vary much from place to place under more normal weather conditions, as illustrated by table IX. TABLE VII Moisture conditions in western Nebraska in May 191 2, after wet winter and spring Depth McCooK Wauneta Imperial Foot May 7 May 7 Mays Field E May 16 Field B May 17 May II May 13 May 14 Field A Field B Field c Field E Field G Field H Field I Field J Field K HYGROSCOPIC COEFFICIENTS I. . 8.4 9.6 8.1 8.7 8.4 7.2 9.6 10.3 8.3 7-5 7.8 7.6 10. 1 10. 1 7.3 9.2 10.4 10. 0 9.6 8.4 7.6 7-5 9.1 9.0 9-5 8.4 6.8 8.2 10.2 8.9 S-3 4-7 4-9 1.6 2.6 1.9 1-5 1-3 1-3 2.6 3.7 35 3-5 1.6 1-3 7.1 9.0 5.8 6.3 7-1 .■i-i 3-2 3-2 5-4 3-7 ^ .4. 2 3-. 4-. Average 6 3 0 8.4 8.S 9.2 8.4 7.0 1-7 2.7 3.7 RATIOS I. . 7 2.6 1.4 I.I 1 .0 I.I 1-3 2.1 1-4 1 .1 1 .1 1 .0 1 .1 2.1 1-3 1 .0 2.0 1.9 1-4 1 .0 1 .1 1 .2 2.3 2.2 1.6 1-3 1 .2 1-3 2.0 1-7 1-3 1 .2 1.2 1 .0 2-S 2.1 4.2 4.6 4-S 4-7 2.9 2.8 2.7 2.3 31 3-4 2.4 2.4 i-S I.I 2. 1 2.3 1-3 1.2 2.4 35 2.0 I.I 1 .1 3.. 4.- Average 6 1 . 2 1-4 1-3 1-4 1.6 1-4 3.8 2.9 1 .9 Computations from data of Shantz and of Burr The only data reported by other investigators that may be used for comparison with our own appear to be those secured by Shantz at Akron, Colorado, in 1909, and by Burr at North Platte, Nebraska, in 191 2. While neither of these authors reports the hygroscopic coefficients of the soils, each gives the wilting coeffi- cients for a representative set of samples, these having been com- puted from the determined moisture equivalents. From these data we have computed the hygroscopic coefficients by means of the Briggs-Shantz formula (6, p. 65) : hyg. coef. = wilt. coef. X0.68. I9I9] ALWAY, MCDOLE, &■ TRUMBULL— SUBSOIL 197 For the period June 7-September 27, 1909, Shantz determined the moisture content twice a day in a grama-buffalo grass associa- tion, recording it in 6-inch sections to a depth of 3 ft. and in foot sections through the succeeding 3 ft. From his data we find that on August 7-8 and lo-ii the ratio of the moisture content to the TABLE VIII Data from H. O. Ranxh in different years, rNCLUDiNG the favorable SEASONS OF 1907 AND I908 Defih 1907 1908 IQIO 1911 1913 Foot Nov. 22 April 30 July 29 March 24 Sept. 21 April 21 July 13 HYGROSCOPIC COEFFICIENTS I 7-7 10.3 10. 1 7-7 71 7.2 6.4 8.5 9.8 9,8 8.3 6.9 7-4 8 r, 8 '' 8.8 "■5 9.2 7-9 7-5 6.5 8.9 II l^ 8.4 2 9 II 7 6 6 5 5 7 3 7 4 3 9 I 9 10 8 7 7 6 7 4 7 9 5 8 I z IO-5 8 6 4 8.6 c 8.5I 7-5 7.2I 6.8!- 7 -7] 6 7 8 6.4 0 8.3 Average i-6 8.3 8.4 8.2 8.8 8.6 9.2 7.6 RATIOS 1 1.9 l.I I .1 1-3 i-S 1.8 1.8 1.2 I.I 0.9 I.O I . I 1 .1 2.1 2.0 I . I 1.6 1.8 1-5 1.6 1-3 2. 1 1 .2 1 .0 1.0 1 .0 1 .1 I.I I . I 1 .1 1.2 0.8 0.8 0.9 0.9 I.I 0 7 0.9 2 0.8) 0.8 0.9 1.0 I . I. I . I 1 . 2 1 ^ 4 ' 0.9 c 6 7 8 I 2 • 9 ' Average i-6 i-S I.I 1-7 1.2 0.9 1 .0 1.0 hygroscopic coefficient in the 7-12 and 13-18 inch levels fell to approximately i.o (10, p. 35); while from July 22 to September 9 the ratio at the latter depth was almost continuously much below 1.5. From August 7 to 13 the sixth foot, and to a less extent the fifth, showed ratios close to i . o. In each of the 4 months included in the study the rainfall was much above the normal for Akron, the excess varying from 25 to more than 100 per cent (8), and 198 BOTANICAL GAZETTE [march averaging at least 50 per cent above the normal. There was no actual drought at any time during the season, but there were two rather dry periods, June 14- July 6 and July 11-25, ^^ which light rains gave totals of o. 20 and o . 09 inch respectively. When the subsoil at Akron, even in that unusually wet summer, had its moisture content reduced to such a low point, it is probable TABLE IX Data from 6 individual borings on the H. O. Ranch, November 22, 1907, illustrating variations from boring to boring Depth No. I No. 2 No. 3 No. 4 No. s No. 6 Aver- age Maxi- irnun Mini- mum Foot HYGROSCOPIC COEFFICIENTS I 2 3 4 5 6...- 7 Average 1-7 I 2 3 4 S 6 7 Average 1-7 6 9 7-4 7.8 7.6 7.2 6 I 7 I 7 8 8 4 8.6 10.4 10.2 8-3 6 7 8 8 10 4 7 3 6.7 10.2 10. 0 8.1 7 5 8 4 10 2 8 5 5-1 7.0 8.4 9.9 7 7 7 9 9 9 7 3 4.2 7.0 71 II. 8 6 7 7 3 II 8 7 2 5-3 7.8 6-5 9.9 5 4 7 0 9 9 7 9 C.7 6.9 6.0 9.2 5 2 7 3 9 2 7 6 6.5 8.1 8.0 9.2 6 5 7 6 9 2 6.1 6.7 6.7 4.2 5-3 5-2 6.3 1.6 1-7 1.8 1-9 1-7 2.0 1.8 2.0 1-3 1 .1 1 . 2 I.I 1.2 1-5 1.2 i-S 1-3 I.I I.I I.I 1 .2 1 .1 1.2 1-3 1.2 I.I 1-5 I.I I.I 1.2 1.2 i-S 1.2 1-3 1-7 1-3 1-4 1 . 2 1-3 1-7 1-3 1-4 1-7 2.0 1-4 1-3 1-4 2.0 1-3 1-7 1.9 1.8 1.2 1-4 1-5 1.9 1-3 1-3 1.6 1-5 1-3 1-4 1-4 1.6 1.6 1 .1 1 .1 I.I 1.2 1-3 I .2 1-3 that in a really dry season the ratios would be found as low as those we encountered in southwestern Nebraska. The root systems of the native plants were studied by Shantz, but the penetration of the grama and buffalo grasses he indicates (10) would not account for the removal of available moisture from below the first foot or two. Burr (7) reports data from a prairie sampled in the spring, summer, and early autumn of 191 2 (table X). In the spring high ratios were shown in the first 2-3 ft., but by the end of June the igig] ALWAY, MCDOLE, cp- TRUMBULL— SUBSOIL 199 ratios at all levels sampled had fallen to practically i.o and so remained through the remainder of the season, there being no evidence of further drying of the subsoil. Thus the data of both Shantz and Burr confirm our findings regarding the stage of dryness to which the subsoil may be reduced in a dry season by the short grass vegetation, while those of the latter author agree also with our view that after the subsoil moisture content has been reduced to approximately the hygroscopic coeffi- cient it suffers but httle, if any, further lowering through a continua- tion of the drought conditions, and not with that of Briggs and Shantz that the subsoil continues to lose water through the plant tissues until it approaches an air-dry condition (6, p. 8). TABLE X Ratio of moisture context to hygroscopic coefficient in a prairie field AT North Platte in 191 2, computed from data of Burr Depth foot Hygroscopic coeflScient* 6.8 6.8 6.8 6.8 7-S 7-5 April 18 2.5 9 2 I I I 29 May June 10 1-4 1.3 1.6 1-4 1-3 1.2 29 July 25 26 Aug. 0.8 o," o o o o Sept. * Computed from wilting coefficients of a representative set of samples (7). Extremes in eastern Nebraska The periods of extreme drought at Lincoln were not numerous, and usually when these came an examination with the soil auger showed that the moisture content of even the surface foot or two was well above the hygroscopic coefiicient, and as it was only the minimum moisture content that we were seeking in these prairie fields we report data from only a few sets of samples. On only 3 occasions in the 6-year period (1906-1912) did we find in the surface 2-3 ft. the dry condition which indicates the approaching exhaustion of available moisture (table XI). The first sampling, on August 23, 1909, had been preceded by a comparatively dry period of 42 days, during which only 1.57 inches of rain had fallen; none had fallen in the last 20 days, while the 200 BOTANICAL GAZETTE [march weather had been unusually hot and windy. The time of the season was that at which the draft upon the subsoil moisture might be expected to show the most marked effect. In the soil of the first foot we found a ratio of i .o and in that of the next 3 ft., an average ratio of i . 4 ; but the sixth foot, with a ratio of i . 9, appeared to have lost but little of the moisture which it could retain against downward movement. TABLE XI Moisture conditions in prairie fields near Lincoln after unusually dry periods August 23, 1909 August 3, 1911 June 7, 1912 Depth foot Hygroscopic coefficient Ratio Depth foot Hygroscopic coefficient Ratio Depth foot Hygroscopic coefficient Ratio I 2 3 4 5 6 10.9 10.8 "5 14.0 12. S II. 8 I.O 1-3 i-S 1-5 1.6 1.9 I 2-S . • • 6-8... 9 13-I 12.3 12.8 1-4 1-3 2.0 2.1 I 2 3 4 5 6 134 lS-2 14. 1 13-9 12.5 12.2 1-5 1-3 1.2 1.2 1.5 1.9 On the second occasion, August 3, 191 1, a dry period of 65 days had just been ended by a rain of 0.84 inch. Compared with a normal precipitation of 9.0 inches for this period, only 2.68 inches of rain had fallen, and this in light showers, while both the mean temperature and the wind velocity had been somewhat above the normal. As the subsoil of the second to the fifth foot appeared uniformly dry it was combined into a single sample, the ratio proving to be 1.3, but in the sixth to ninth foot it was 2 .0 to 2 . i. The moister condition in the surface foot indicated in the table was due to a shower of the day before having moistened the immediate surface layers. From the time of the preceding to the next and last sampling, June 7, 191 2, the weather on the whole was very unfavorable to the accumulation of any moisture in the subsoil, and the spring of 191 2 was exceptionally favorable to the exhaustion of whatever available water was within reach of the plant roots. The moisture conditions found were quite similar to those on the preceding occasion. Thus the samplings, taken at times of drought when one might have expected almost the lowest moisture content in the subsoil I9I91 ALWAY, McDOLE, a- TRUMBULL—SUBSOIL 201 ever to be found in prairie fields at Lincoln, showed dry subsoil onl}- within the first 5 ft., below this depth the ratios lying between extremes of 1.9 and 2.7, or in general between 2.0 and 2.4, the moisture retaining capacity of the subsoil. On only one occasion, and then only in the first foot, was the moisture content found reduced as low as the hygroscopic coefficient, and there it is to be attributed to the surface few inches of the foot section having been dried by evaporation to a point much below this value, with the result that the average for the whole foot section shows a low ratio. In this connection it is of interest to know the ratios which normally prevail in the deeper 'subsoil of the eastern prairies. In April 191 1, a field situated on a gentle slope and 50 ft. or more above ground water was sampled to a depth of 18 ft. (table XII). Below the fifth foot ratios ranging only between 2.1 to 2.4 were found. TABLE XII Moisture conditions in prairie near Lincoln, April 13, 1911, showing NORMAL condition OF DEEPER SUBSOIL Depth foot Hygroscopic coefficient Ratio Depth foot Hygroscopic coefficient Ratio Depth foot Hygroscopic coefficient Ratio I 2 3 4-.... s 6 XI. 8 15-3 14-3 14.2 13-6 131 2.5 2.0 1.8 1.8 1.9 2.1 7.... 8.... 9.... 10. . . . II. . . . 12. . . . 13-0 12.8 13.6 12. I 13-4 130 2.2 2.2 2.1 2.4 2.1 2.2 I3--- 14.... IS ■•• 16.... 17.... 18.... 12.3 12.0 12.3 10.3 9.8 10.9 2.2 2-4 2-3 2.3 2.3 2.2 Thus as near the surface as the sixth foot, when conditions were such as to develop the driest subsoil, the ratio was not much below that found in the deep subsoil under normal conditions. This failure of the natural vegetation of the prairies of eastern Nebraska to exhaust the free water of the deeper subsoil is in sharp contrast with the conditions found on the short-grass prairies of the southwestern part of the state, as previously described. That this moist condition is due to a difference in the conduct of the native plants and not to any peculiar properties of the humid subsoil is evident from the fact that in the alfalfa fields adjacent to the prairies kept under observation the ratios were quite com- monly found reduced as low as i . 2 to i .4 to a depth of 15 or 20 ft., 202 BOTANICAL GAZETTE [m.'VECH or even more (table XIII). In an oak grove planted on the prairie some 30 years before and sampled on practically the same dates TABLE XIII Moisture conditions in eastern Nebraska alfalfa field, adjacent to pr.a.irie reported in table xii, showing failure of prairie veget.4.ti0n to reduce MOISTXIRE content OF DEEPER SUBSOIL NOT DUE TO ANY PECULIARITY OF SUBSOIL April 13, 1911 September 12, 191 2 Hygroscopic coefficient Ratio Boring i Boring 2 Depth foot Hygroscopic coefficient Ratio Hygroscopic coefficient Ratio I II. 6 134 II. 0 8.5 II-S 2.4 1-7 i-S i-S 1-4 13.2 13-7 12.4 lO-S II. 8 1-4 I.I I.I I.I 1.2 12.9 13-S 12. I 10.8 12. 1 1-3 1. 1 2-6 7-12 1. 1 13-18 19-21 I.I I.I as the prairie fields, the subsoil moisture was found to be affected to a greater depth than in the latter, the drying effect extending apparently to at least 15 ft. (table XIV). TABLE XIV Moisture conditions in an oak grov'e near Lincoln August 23, 1909 August 2, 1911 Julys, 1912 Depth foot Hygroscopic coefficient Ratio Depth foot Hygroscopic coefficient Ratio Depth foot Hygroscopic coefficient Ratio I 2 3 4 s 6 10. 1 II. 7 14.2 14. 1 13-9 13-4 I.I 1-3 1-3 1.2 1.2 1.2 I 2-9... 10-14. iS---- 9-1 12.4 II. 9 12.0 1.8 1-3 1.6 1.6 1. . . . 2. . . . 3--- 4.... 5---- 6.... 10. 0 II. 8 14.2 14. 1 14.0 13-5 0.7 1-7 1-4 1.2 1 . 2 1 . 2 Discussion The moisture conditions in the deeper subsoil of the prairies are very dissimilar according to whether we deal with humid or with semi-arid fields. In the former at depths below 6 ft. the sub- soil appears always moist, even after the severest drought, while in the latter the extreme dryness indicated by ratios of i . o-i . 2 is in general persistent in the deeper subsoil, extending to a depth of igig] ALWAY, McDOLE, &> TRUMBULL— SUBSOIL 203 1 2 ft. or more after prolonged droughts, and even in wetter seasons is commonly found in one or more foot levels within the first 6 ft. That the lack of dryness in the deeper subsoil of the humid prairies is not due to any peculiarity of the subsoil is evident from the observation that a fair stand of alfalfa may in the course of a few years reduce the moisture content almost to the hygroscopic coefiicient to a depth of 20 ft. or more. In our deep cyhnder experiments (i) the exhaustion of free water was observed only within the zone of root development, and in our recently reported study of the movement of water in the absence of plants (3) we found no appreciable transfer of water from a moister to a drier portion of a soil when the ratio in the former was as low as i . 5 and that in the latter between i . 5 and i . o. If we assume that movement of water through a soil ceases when the ratio in the moistest portion has fallen as low as 1.5; that the deeper subsoil loses water through upward movement only when it is penetrated by plant roots; and, lastly, that plants are able to develop roots into a soil layer only when this has a moisture content above the computed wilting coefficient (5), the ratio 1.5, we must conclude that the roots responsible for the dry condition (indicated by ratios of 1.1-1.4) encountered in any subsoil level either will be found surviving or that they have died only since this level of the subsoil was last reduced to the dry condition. In order to explain how the dry condition of the deeper subsoil is first established and how it is renewed after wet periods, it seems necessary to assume that among the shallow rooted grasses there are distributed a considerable number of very deep rooted peren- nials. After this dry condition of the deeper subsoil has once been established it may be maintained through a dry period of several successive years without the presence of any roots in it, the moisture from the rains and snow being held near the surface until it either evaporates or is transpired by the shallow rooted plants, while the upward movement of water from the moist zone beyond the extreme reach of plant roots is at least too slight to show a distinct effect. The absence of the dry condition in the deeper subsoil after pro- longed droughts, such as illustrated by field E at Wauneta (table 204 BOTANICAL GAZETTE [m.4rch VI), may be attributed to a temporary absence of the deep rooted perennials or to their fewness. The factors just mentioned are sufficient to account for the maintenance of a dry upper subsoil through which no roots could develop into the moist zone. The question of whether the living roots are to be found in the deeper subsoil only during each successive wet period, they follow- ing the downward extension of the moist zone, continuing to withdraw water until the ratio approximates i . o, and then dying off, or whether they continue alive but withdrawing practically no moisture throughout the dry periods of several years which inter- vene between the successive wet periods, is to be answered only by detailed field investigations, involving the use of pits or trenches 12-20 ft. deep. The present moisture conditions of the deeper subsoil of the prairies, like their plant population, are to be regarded as the result of a slowly established equilibrium, and any alteration of the plant cover may greatly affect the subsoil moisture conditions. The complete suppression of plant life over an acre or more, a condition approached in young orchards and groves kept in clean cultivation, might during a series of wet years raise the moisture content of the deeper subsoil to its water-retaining capacity and maintain this with little change during the ensuing dry period. If such a field were neglected, however, it would soon be taken possession of by many species, most of them shallow rooted annuals, but some deeper rooting perennials, which, meeting little competition for moisture in the deeper subsoil, could develop an extensive root system there and gradually reduce the moisture to approximately the hygroscopic coefficient. Then, as on the prairie, this dry con- dition would be maintained except at such times as unusually wet seasons extended the moist zone far below its normal limits. While it is evident from table VI that the lower limit of the dry zone in the deep loessial soils in the semi-arid region is more than 12-15 ft. below the surface we have no data showing its maximum depth. RoTMiSTROV (9), from his studies near Odessa, concluded that there permanently moist subsoil in waste land occupied by weeds, etc., is first encountered at 14-30 ft. The depth to which the root systems of the deeper rooted prairie plants indicated by igig] ALWAY, McDOLE, &• TRUMBULL—SUBSOIL 205 Shantz (10) extend would not suffice to explain the dry condition of the deeper subsoil which we encountered. The persistently moist condition of the deeper subsoil of the humid prairies is to be attributed to the fewness of the roots developed in them. When deep rooted perennial plants such as alfalfa or forest trees are introduced, their subsoil moisture is utilized to a much greater depth. It is evident that on these a forest once established should be able to maintain itself if pro- tected from fires. The subsoil moisture conditions in general would indicate that the natural condition of grassland in eastern Nebraska is due to other causes than soil moisture conditions, while in western Nebraska it may be fully accounted for by those alone. The distribution of carbonates in the first 6 ft. of soil in the prairies at McCook and Wauneta indicates that in prehistoric times the climate was similar to that now prevailing (4). Carbonates are found in the surface foot or two only in almost negligible quantities, while in the fourth, fifth, and sixth feet they constitute from 3 to 6 per cent of the weight of the soil. Summary 1. During a 6-year period, in which the weather was exception- ally favorable for a study of the minimum moisture content of the subsoil, moisture studies were carried out on Nebraska prairies, both in the buffalo-grass formation in the southwestern part of that state, where the climate is typically semi-arid, and in the prairie- grass formation near Lincoln, which Hes within the limits. of the humid region. The fields were sampled to a depth of 6 ft. or more, and in the case of every sample the hygroscopic coefficient as well as the moisture content was determined, and the moisture condition is expressed as the ratio of moisture content to hygroscopic coeffi- cient, this having the advantage of expressing the relative moistness while at the same time indicating whether either free water (i . i or above) or growth water (1.6 or above) is present, and if so the amount of each. 2. The subsoils of the semi-arid prairies were characterized by their persistent dryness. Usually throughout more or less of the 2o6 BOTANICAL GAZETTE [march first 6 ft. a ratio of i . 5 or lower was found, and commonly in one or more of the foot sections a ratio as low as i . i was encountered. After droughts of unusual severity the whole of the subsoil to a depth of 6 ft., and in some cases of 12 ft., showed a ratio of approximately I. o. 3. There was no appreciable further reduction of the moisture content when, after the subsoil had been reduced to this very dry condition, there followed a 4 or 5-month period of practically rain- less autumn and winter weather. After such droughts the surface foot was found but little drier than the subsoil. 4. The subsoils of the humid prairies, on the contrary, showed no distinct reduction of the moisture content through a greater depth than 5 ft., and even in this a ratio as low as 1.2 or 1.3 appeared only under the severest drought conditions. The normal moisture condition in the deeper subsoil (6-20 ft.) appears to correspond to a ratio lying between 2 . o and 2 . 4. 5. The dry condition of the deeper subsoil so common in the semi-arid prairies is to be attributed to the presence of perennials with a vertical root range of 15 ft. or more, while the moist con- dition characteristic of that of the humid prairies is regarded as evidence that the roots of the native vegetation are but little developed below the fifth foot. The occurrence of areas in the semi-arid prairies, even after a severe drought, in which the subsoil below the sixth foot is quite moist, is to be attributed to the absence or fewness of deep rooted perennials in such places. 6. After the subsoil at any level has been exhausted of the water in excess of the hygroscopic coefficient it remains in this dry condi- tion until the precipitation conditions are sufficiently favorable to raise the ratio to 2.0 or upward throughout the whole distance from the surface down to the level in question. Accordingly during many wet periods following droughts the upper moistened portion of the subsoil will be isolated from any deeper lying moist layer by a zone in which the subsoil is too dry to permit of the penetration of plant roots. 7. While in the semi-arid prairies after protracted droughts the moisture conditions in the first 6 ft. are quite uniform, under more iqiq] ALWAY, McDOLE, 6- TRUMBULL— SUBSOIL 207 normal weather conditions they vary much from place to place, thus rendering the results obtained in single borings unrehable as an index of the general moisture conditions. University Farm St. Paul, Minn. LITEIL^TURE CITED 1. Alway, F. J., Studies on the relation of the non-available water of the soil to the hygroscopic coefficient. Nebr. Agric. Exp. Sta. Research Bull. 3. pp. 122. figs. 37. 1913. 2. Alway, F. J., and McDole, G. R., The loess soils of the Nebraska portion of the Transition Region. I. Hygroscopicity, nitrogen, and organic carbon. Soil Science 1:107-238. pi. j. figs. 2. 1916. 3- . Relation of the water-retaining capacity of a soil to its hygroscopic coefficient. Jour. Agric. Research 9:27-71. figs. 4. 1917. 4. Alway, F. J., and Rost, C. 0., The loess soils of the Nebraska portion of the Transition Region: IV. IMechanical composition and inorganic con- stituents. Soil Science 1:405-436. figs. 2. 1916. 5. Briggs, L. J., Dry-farming investigations in the United States. Rep. 84th Meeting Brit. Assn. Adv. Sci., 1914. pp. 263-282. pl.j. fig. 7. 1915. 6. Briggs, L. J., and Shantz, H. L., The wilting coefficient for different plants and its indirect determination. U.S. Dept. Agric, Bur. PI. Ind. Bull. 230. pp. 83. pi. 2. figs. g. 1 91 2. 7. Burr, W. W., The storage and use of soil moisture. Nebr. Agric. Exp. Sta. Research Bull. 5. pp. 88. 1914. 8. Grace, 0. J., The effect of different tunes of plowing small grain stubble in eastern Colorado. U.S.D.A. Bull. 253. 1915. 9. RoTMiSTROV, V. G., The nature of drought according to the evidence of the Odessa Experiment Field, pp. 48. figs. 20. Odessa. 1913. 10. Shantz, H. L., Natural vegetation as an indicator of the capabiUties of land for crop production in the Great Plains Area. U.S. Dept. Agric, Bur. PI. Ind. Bull. 201. pp. 100. pi. 6. figs. 23. 1911. 11. Tinsley, J. D., and Vernon, J. J., Soil and soil moisture investigations in the season of 1901. N.M. Agric. Exp. Sta. Bull. 38. pp. 94. pi. 11. fig. i. 1901. NOTES ON NORTH AMERICAN TREES. IV C. S. Sargent PiCEA GLAUCA (Mocnch) Voss, Mitt. Deutsch. Dendr. Ges. 16:93. 1907. — Although Abies canadensis Miller is the oldest name for the white spruce, Picea glauca, according to the rules of the Vienna Congress, must be adopted for this tree, for there is a Picea canadensis, which is a valid name for the hemlock spruce under the genus Picea. The Rocky Mountain variety then becomes Picea glauca var. albertiana, n. comb. — Picea canadensis var. albertiana Rehder, Mitt. Deutsch. Dendr. Ges. 24:213. 1916. JuNiPERUS UTAHENSis var. mcgalocarpa, n. var. — Juniperus megalocarpa Sudworth, Forestry and Irrigation 13:307. figs, i and 2. 1907; Wooton and Standley, Contrib. U.S. Nat. Herb. 19:37. 1915. — Sahina megalocarpa Cockerell, Muhlenbergia 3:143. 1907. — Differing from the type in its larger fruit sometimes i . 7 cm. in diameter, and in its habit of sometimes forming a single erect stem. A tree 10-14 n^- high, with a straight trunk occasionally i-i .2 m. in diameter. Valley of the San Francisco River near Alma, southwestern New Mexico, W. R. Mattoon, September 1906, Alfred Rehder, August 13 and 14, 1914 (nos. 285, 285b, 289); Arizona: rim of the Grand Canyon, C. 5. Sargent, Septem- ber 9, 1904; Angel, near Flagstaff, Percival Lowell, September 4, 1910. Mr. Rehder visited the type station of this tree in New Mexico in August 1914 and obtained a large amount of material which shows that it must be considered a variety of /. utahensis, for he found trees with fruit intermediate in size between that of the typical /. utahensis (6-7 mm. in diameter) and the largest fruits of /. megalocarpa, that the trees with single stems did not always produce large fruit, and that the large-fruited trees sometimes had the charac- teristic habit of /. utahensis. Like that of /. utahensis, the fruit of the variety ripens at the end of the second season. A specimen of /. utahensis in the herbarium of the Arboretum collected by E. Bethel at Radium, Colorado, in November 1908 at an elevation of 2300 m., has fruit 1.3 cm. in diameter and should perhaps be referred to the variety. PoPULUs TREMULOiDES var. vancouveriana, n. var. — Populus vancouveriana Trelease apud Tidestrom in Piper and Beattie, Fl. Botanical Gazette, vol. 67] [208 19 19] SARGENT— NORTH AMERICAN TREES 209 Northwest Coast, 118. 191 5.— Differing from the type in its thicker, more coarsely serrate leaves, densely hoary tomentose below when they unfold, villose aments, and in its pubescent or puberulous branchlets and tomentose winter-buds. Leaves broadly ovate to semiorbicular, rounded or slightly cordate at broad base, abruptly short-pointed or rounded at apex, coarsely crenately ser- rate and sometimes obscurely crispate on the margins, when they unfold covered below and on the petioles with a thick coat of long matted pale hairs and slightly villose above, soon glabrous, and at maturity thick, dark green, lustrous, and scabrate on the upper surface, paler on the lower surface, 8-1 1 cm. long and broad, with prominent midribs and primary veins; petioles slender, compressed, becoming glabrous, 5-8 cm. in length. Rachis of the staminate inflorescence slightly villose, the pedicels pubescent; disk of the flower puberulous toward the base, flowers otherwise as in the species. Pistillate inflorescence 5-6 mm. long, the rachis, pedicels, and slightly lobed disk of the flower densely villose, becoming pubescent or glabrous on the fruit; ovary conic, pubescent, with a short style, and stigmas divided into narrow divergent lobes. Fruiting aments 8-9 cm. long, the fruit oblong-conical, pubescent or glabrous, 5 mm. long; pedicels not more than i mm. in length. A tree 10-12 m. tall, with a trunk 30-40 cm. in diameter, stout spreading branches forming a round topped head, stout, reddish brown pubescent or puberulous branchlets often becoming glabrous during their first summer, and acute tomentose pubescent or glabrous winter-buds. Borders of salt marshes on the coast of Vancouver Island, British Columbia, near Sidney, J. Macoun, May 3. 1913 (staminate flowers), May 13, 1913 (pis- tillate flowers), June 13, 1913, April 1914 (pistillate flowers), April 13, 1914 (fruit), May 6, 1914, C. 5. Sargent, July 27, 1913; by shingle mill near Sidney, /. Macoun, April 2, 1913; old brickyard, Sidney, June 22, 1914; near Victoria, British Columbia, Engelmann and Sargent, August 19, 1880, /. Macoun, August 21, 1893 (no. 2131), May 27 and August 31, 1893 (no. 2232), May 28, 1908 (nos. 85714, 8805a), June 4, 1908 (no. 88059), May 5, 1915; Dead Man's Creek, near Victoria, /. Macoun, July 23, 1908; Esquimo, Vancouver Island, J. Macoun, June 23, 1887; Cape Laza, near Comox, Vancouver Island, /. Macoun, July 7, 191 5. Extreme forms of this tree certainly appear distinct from P. tremuloides Michaux and its western variety aiirea Daniels, but the shape of the leaves is not constant ; the branchlets and the young leaves are sometimes glabrous or nearly glabrous; on some branches the winter-buds are not tomentose but are 210 BOTANICAL GAZETTE [march pubescent or glabrous; and the only constant character I can find m the Van- couver Island trees is the presence of the dense covering of hairs on the rachis of the male and female inflorescence and on the disk of the pistillate flower. Populus arizonica, n. nom. — Populus mexicana Sargent, Silva N. Am. 14:73. pi. 733 (not Wesmael). 1902; Man. 162. fig. 136. 1905. — A photograph of the type specimen of P. mexicana Wesmael collected by Berlandier between Tampico and Rial del Monte in May 1827, kindly sent to the Arboretum by the late Casimir DeCandolle, shows clearly that the tree with small fruit which is common in the neighborhood of Tucson, Arizona, was wrongly referred to Wesmael' s species. For the Arizona tree I now suggest the name of P. arizonica. It is well distinguished from the other cottonwoods of the United States by the small fruit, which does not exceed 5 mm. in length. Arizona. — Common up to 2200 m. above sea level. Santa Catalina Mountains, Pima County, C. G. Pringle, 1881, A. Rehder, August S, 1914 (no. 256) ; Bear Creek, foot of Santa Catalina Mountains, ^. iJeMer, August 30, 1916 (no. 452); Sabino Canyon, Santa Catalina Mountains, /. W. Tourney, February 20, 1894, A. Rehder, August 7, 1914 (no. 233), September i, 1916 (no. 500) ; near Tucson, Pima County, Engelmann and Sargent, September 30, 1880, /. W. Tourney, February 27, 1894, February 1898, C. S. Sargent, February 27, 1894, March 27, 1916; Douglas, Cochise County, A. Rehder, August 27, 1916 (no. 447a, a planted street tree); Yavapai County, near Clarkdale, A. Rehder, September 11, 1916 (no. 555), near Camp Verde, September 9, 1916 (no. 542), Beaver Creek, near Camp Verde, September 8, 1916 (no. 540), banks of Gannett Creek, Prescott, September 4, 1916; on Salt River, near the Roosevelt Dam, Gila County, W. H. Goddard, June 191 7; Coconino County, Sycamore Canyon, Percival Lowell, September 191 5, A. Rehder, September 14, 1916 (no. 573); Hermit Creek, Grand Canyon of the Colorado, Alice Eastwood, April 10, 1917 (no. 6002); Turkey Creek, near Flagstaff, C. 0. Lampland, March 191 7; Canyon Diablo, C. 0. Lampland, April 5, 191S (these specimens have only very young flower buds and may belong to another species). California.— A sterile branch collected by A. Rehder July 23. 1914 (no. 128), on Mill Creek above Forest Home, San Bernardino Mountains, is doubtfully referred to this species. New Mexico.— Near Silver City, Grant County, E. L. Greene, 1880 (dis- tributed as P. Fremontii), 0. F. Arthur, March 16, 1918, M. W. Talbot, April i, 1918, /. A. Scott, May 18, 1918. In the neighborhood of Tucson, where this poplar has been planted in considerable numbers, it is a magnificent tree 20-28 m. in height with a trunk iQig] SARGENT— NORTH AMERICAN TREES 21 1 sometimes i m. in diameter, a broad head of wide spreading branches, slender branchlets, glabrous or puberulous, and lustrous yellow-green leaves often puberulous early in the season. The bark of the branches and young stems is nearly white and on old trunks it is pale green and slightly divided into broad flat ridges. PoPULUS ARizoNiCA var. Jonesii, n. var. — Differing from the t^-pe in the pubescent, not puberulous, young leaves, petioles, and young branchlets. Mexico, Valley of Palms, Marcus E. Jones, April 8, 1882 (no. 373, type); valley near Chihuahua, C. G. Pringle, March 31, 1886 (no. 885); Saltillo, C. G. Pringle, June 4, 1888 (no. 2098, with larger leaves and more pubescent branch- lets), C. 5. Sargent, March 1887 (a very large tree with pendulous branches); Valley of Mexico, C. G. Pringle, February 13, 1899 (no. 8019); Pedras Negras, C. S. Sargent, March 21, 1900 (a planted tree). Populus Palmeri, n. sp. — Leaves thin, ovate, cuneate or rounded at the broad base, gradually or abruptly contracted at apex into a narrow acuminate entire point, finely serrate with incurved teeth, ciliate on the margins when they unfold, otherwise glabrous, 6-10 cm. long and 4.5-8 cm. wide; petioles slender, glabrous, 3.5-6 cm. in length. Flowers not seen. Fruit in slender glabrous aments 12-15 cm. long, ovate, obtuse, slightly pitted, puberulous, thin- walled, 4-valved, 6-7 mm. long, the disk deeply lobed, 4-5 mm. in diameter; pedicels slender, 7-9 mm. in length. A tree 20-21 m. tall with a straight trunk i m. in diameter, erect, smooth, pale branches forming an open pyramidal head, the lower branches smaller, horizontal or pendulous, and slender, glabrous branchlets hght reddish brown €arly in the season, becoming pale grayish brown in their second year. Bark pale, 5 cm. thick, deeply divided by w^de fissures into narrow ridges. In most fertile soil near springs, at the base of high chalk bluffs of Nueces Canyon of the upper Nueces River, Uvalde County, Texas, growing with Salix nigra var. Lindheimerii, Carya pecan, Morus rubra, and Ulmus crassijoUa, E. J. Palmer, April 11 and September 1918 (nos. 13340, 14511). In the shape of the leaves and their serration, in the small fruit, and in the remarkably slender branchlets this poplar is so different from all other American species that, although it is stiU very imperfectly known, I venture to describe it. It is the only species seen by Palmer in Uvalde County. Populus texana, n. sp. — Leaves thin, glabrous, broadly ovate, truncate at base, gradually narrowed, long-pointed, acuminate at apex, coarsely crenately serrate below the middle, entire above, 212 BOTANICAL GAZETTE [march 7-8 cm. long and 6-7 cm. wide; petioles slender, compressed, 4-7 cm. in length. Flowers not seen. Fruit in slender, glabrous aments 7-8 cm. long, oblong-ovate, acute, deeply pitted, glabrous, thin-walled, 3-valved, 8-9 mm. long, the disk slightly lobed, 2 . 5-3 mm. in diameter; pedicels slender, 3-4 cm. in length. Seeds ovate, acuminate, 4 mm. long. A tree up to 20 m. high, with a trunk sometimes i m. in diameter, and stout more or less pendulous branches, stout, glabrous, pale yellow-brown branchlets, and acuminate, glabrous winter-buds. In canyons and along the streams of northwestern Texas, where it appears to be the only cottonwood. Low sandy banks of the Canadian River, Cana- dian, Hemphill County, E. J. Palme/, June 17, 1918 (no. 14107); creek banks, Amarillo, Potter County, E. J. Palmer, July 13, 1917 (no. 12541); canyon, Paloduro Creek, Randall County, E. J. Palmer, October 3, 1918 (no. 14591); river banks in canyon, Gamble's Ranch, Armstrong County, June 6, 1918 (no. 13959). "One of the largest trees found in Paloduro Canyon, growing in the protection of high bluffs. It usually grows in the protection of high bluffs or at the heads of canyons. The young trees here are slender and straight, but older specimens are very irregular or unsymmetrical in growth, with pale or dark ashy bark. It is rarely found in the more open parts of the canyon here, but near Canyon City it grows on the river margins" (E. J. P. in litt.); Post, Garza County, E. J. Palmer, May 31, October i, 1918 (nos. 13848-, 13853, 14575); along creeks. Sweet Water, Nolan County, E. J. Palmer, October 21, 1917, May 28, September 28, 1918 (nos. 13045, 13799 type, 13899, 14526). By the shape of the leaves and by the thickness and color of the branchlets this species cannot be distinguished from P. Wislizenii Sargent, but from that species it is well distinguished by the smaller fruit on much shorter pedicels and by the glabrous winter-buds. The range of the two trees is also quite different. PoPULUS Macdougallii Rose, Smithsonian Misc. Coll. 61:61. 1 9 13. — This species, of which I have not seen the flowers, is well distinguished from P. Fremontii by the minute disk of the fruit, which does not exceed 3 mm. in diameter. The fruit is borne on slender, glabrous pedicels 3-5 mm. in length, in racemes 5-6 cm. long; it is ovate and acute at apex to ellipsoidal and acute or acuminate at ends, glabrous, slightly pitted, thin-walled, 3-valved, 1Q-12 mm. in length. The seed is oblong-ovate, acuminate, 3 mrn. in length. It is probably always a small tree with erect branches and slender branch- lets pubescent or puberulous when they first appear, soon becoming glabrous and pale yellow -brown at the end of their first season. 1 9 19] SARGENT— NORTH AMERICAN TREES 213 This is the common and probably the only cottonwood of the valley of the lower Colorado River. It is common on both banks of the river at Yuma, and is planted in some of the towns of the Colorado Desert region like Yuma, Mecca, and Indio. It has also been planted at the Needles on the Colorado River in San Bernardino County, Cahfornia. PopuLUS Fremontii S. Watson. — This is the common and only cottonwood of the valleys of northern and central California west of the Sierra Nevada. The leaves are slightly cordate at the broad base and coarsely serrate often with few teeth. The fruit is ovate with a disk about 5 mm. in diameter, on pedicels 3-5 mm. in length. In San Bernardino County, California, Nevada, Utah, and Arizona poplar trees occur which, although the disk of the fruit is smaller or larger than that of the typical F. Fremontii, until better known are best considered perhaps varieties of that species. Three of these forms may be distinguished as follows : PopuLUS Fremontii var. Thornberii, n. var. — Leaves broadly ovate, abruptly contracted into acuminate points, slightly cordate at the wide base, coarsely crenately serrate with numerous teeth, glabrous, 6-8 cm. long and broad; petioles 3.5-4 cm. in length. Flowers not seen. Fruiting aments 5-6 cm. long, the capsules elhpsoidal, 3-valved, deeply pitted, 8-9 mm. long; disk 3 mm. in diameter; pedicels 2-3 mm. in length. A large tree with pale deeply furrowed bark and pale gray glabrous branchlets. Low ground near Tucson, Pima County, Arizona, C. S. Sargent, March 27, 1916. From the typical F. Fremontii this variety differs in the more numerous serratures of the leaves, in the ellipsoidal, not ovate, fruit with a smaller disk, and in the much shorter pedicels. This tree was shown to me by Professor J. J. Thornber of the University of Arizona, whose name I venture to asso- ciate with it. PopuLUS Fremontii var. pubescens, n. var. — Differing from the type in its more pubescent branchlets. This is a common tree in San Bernardino and San Diego counties, Cali- fornia, and extends into Nevada and southern Utah. The branchlets of the type specimen of F. Fremontii, which was collected by Fremont in the upper Sacramento Valley, are described as slightly pubescent, but on the other specimens of this tree \vhich I have seen from CaHfornia north of San Ber- nardino County they are glabrous, and as the range of the trees with the 214 BOTANICAL GAZETTE [march distinctly pubescent branchlets extends far beyond the region occupied by typical P. Fremontii it will perhaps be best to consider that they represent a geographical variety. I have seen the following specimens: California. — San Bernardino Mountains, G. R. Vasey, 1880; San Ber- nardino, C. C. Parry, April 1883, C. S. Sargent, March 30, 1916; Barstow, San Bernardino County, W. L. Jepson, May 1914 (no. 5894), March 8 and 22, 1916 (nos. 6610, 6611, 6626); Warner Hot Springs, San Diego County, Alice Eastwood, April 9, 1913 (no. 2619); Bernardo, San Diego County, Le Roy Abrams, May 2,- 1903 (no. 33701). Nevada.— " Kiernan, Meadow Valley, Wash.," L. N. Goodding, April 28, 1902 (no. 634); south base of Mount Grant, Mineral County, A. A, Heller, July 2, 1913 (no. 10909). Utah. — St. George, Washington County, M. E. Jones, March 30, 1880 (no. 1611). PopuLus Fremontii var. Toumeyi, n. var. — Differing from the type in the cordate-cuneate broad base of the leaves, and in the larger disk of the fruit. Leaves ovate, the base shallow cordate and gradually narrowed and cuneate to the insertion of the petiole, gradually narrowed and acuminate at the entire apex, coarsely and irregularly crenately serrate below, glabrous, 6-7 cm. long and broad. Fruit oblong-ovoid to slightly obovoid, acute or obtuse at apex, 8-10 mm. long, the disk 6-7 mm. in diameter; pedicels 4-5 mm. in length. Arizona. — Tucson, Pima County, /. W. Tourney, April 28, 1894 (type); Pima Canyon, Santa CataUna Mountains, Pima County, /. /. Thornher, March 2, 1913; Santa Cruz River bottoms, Pima County, /. /. Thornher, March 30, 1913; Nogales, Santa Cruz County, McPherson, April 15, 1915; Hermit Creek, Grand Canyon of the Colorado, Coconino Coimty, Alice East- wood, April 10, 1917 (no. 6002). Populus Parryi, n. hyb. {P. Fremontii Xlrichocar pa). — Leaves ovate, rounded or shghtly cordate at the broad entire base, gradu- ally narrowed, acuminate and entire at apex, finely crenately ser- rate below, sparingly villose and ciliate on the margins when they unfold, soon glabrous, and at maturity thin, dark green, and lustrous above, silvery white below, 6-8 cm. long and broad; petioles slen- der, slightly compressed, glabrous, 4-6 cm. in length; leaves on vigorous shoots sometimes oblong-ovate, truncate or rounded at base, acute at apex, more coarsely serrate, 9-12 cm. long and 8-10 cm. wide, with stout compressed petioles 3-4 cm. in length. iqiq] SARGENT— north AMERICAN TREES 215 Staminate aments densely flowered, 5-6 cm. long, puberulous, the bract of the flower broadly obovate, laciniate; anthers 10 or 12; aments of pistillate flowers villose, 6-7 cm. long, becoming at matu- rity 15-16 cm. in length. Disk of the flower broad, entire or erose on the margin; ovary broad, ovate, puberulous; stigma 3-lobed or occasionally 2-lobed. Fruit broadly ovate, rounded at apex, slightly pitted, puberulous, thin-walled, inclosed sometimes for one-third of its length in the enlarged disk, 5-6 mm. long, often abortive; pedicels puberulous, 2-2.5 mm. in length; seeds narrow- obovoid to ellipsoidal, 3 mm. long. A large tree with deeply furrowed bark, wide spreading branches, slender glabrous branchlets reddish brown in their first season, light orange-brown in their second year, and acuminate, lustrous, glabrous winter-buds. Streets of San Bernardino, San Bernardino County, California, C. C. Parry, March and April 1883 (type), C. S. Sargent, March 16, 1916, 5. B. Parish, October 15, 191 7; along Cottonwood Creek, west side of Owen's Lake, Inyo County, California, F. V. Coville and F. Funston, June 19, 1881 (no. 996); in the Canoda de las Uvas, about 2 miles north of Fort Tjon, Kern County, Cahfomia, F. V. Coville and F. Funston, July 5, 1891 (no. 1163). These trees appear intermediate in character between P. Fremontii and P. trichocarpa. The leaves resemble in shape those of the common Californian form of P. Fremontii, but are silvery white below like those of P. trichocarpa and the other balsam poplars, and their serration is much finer than that of the leaves of P. Fremontii, but coarser than that of the leaves of P. trichocarpa. The staminate flowers have fewer stamens than those of either of the supposed parents; the disk of the female flowers is very similar to that of both of them, but the ovary, which is glabrous in P. Fremontii and densely tomentose in P. trichocarpa, is pubescent. Parry, who first noticed this tree and who would have considered it a new species if he had seen it growing wild, thought that it might have been an exotic species introduced into San Bernardino. The leaves on the specimens of the wild plants from the western base of the Sierra Nevada are similar to those of the San Bernardino trees, but the fruit is rather longer and more acute. Of these specimens Coville wrote me November 4, 1892: "I send you by mail specimens of a poplar collected along streams flow- ing from the southern Sierra Nevada. Specimens of P. Fremontii, P. tremu- loides, and P. trichocarpa were collected by the Expedition (Death Valley), and these specimens show characters between P. Fremontii and P. trichocarpa.'' OsTRYA viRGiNiANA K. Koch.— The variety of this tree, on which the branchlets, petioles, and peduncles are covered with short erect glandular hairs, may be distinguished as 2i6 BOTANICAL GAZETTE [march OsTRYA VIRGINIANA var. glandulosa, n. comb. — Ostrya virginiana var. glandulosa Spach, Ann. Sci. Nat. II. 16:246. 1841. From Quebec and Ontario to southwestern New England and western New York, and in eastern Michigan this is the prevaihng variety. The two forms occur in New Jersey, Pennsylvania, Indiana, northern lUinois, south- western Missouri, Oklahoma, and on the high Appalachian Mountains. Betula Eastwoodae, n. sp. — Leaves broad-ovate to elliptic, crenately serrate except at the cuneate base, thick, glabrous, dark green above, pale below, rather conspicuously reticulate venulose especially on the upper side, 2 . 5-3 cm. long and i . 6-2 . 3 cm. wide; petioles slender, glabrous, often tinged with red, 5-7 mm. in length. Staminate catkins usually solitary or in pairs, sessile, 2-3 cm. long, 5 mm. in diameter, their scales broadly ovate, acute and apiculate at apex, pubescent, dark red. Pistillate catkins pendulous on peduncles 8-10 mm. long, cylindric, i . 5-2 cm. in length, 6-7 mm. in diameter, their scales longer than broad, the lobes rounded at the narrow apex, ciliate, the lateral slightly spreading, one-third shorter than the terminal lobe. A tree rarely more than 6-7 m. tall with a trunk not more than 15 cm. in diameter, covered with close chestnut brown lustrous bark about 5 mm. thick and marked by conspicuous horizontal white lenticels, and slender red branch- lets more or less thickly covered with circular white glands. Roadsides, Hunker Creek, Yukon District, /. Macoun, August 4, 1902 (no. 54412); swamps in the town of Dawson, valley of the Yukon, British Alaska, forming jungles with Betula glandulosa Michaux, occasional plants of Betula alaskana Sargent, and different willows, Alice Eastwood, May 22 and June 14, 1914 (nos. 88, 271 = 88 type, also nos. 6, 7, 58, 69, 533 = 69, 89, 272-89, 102, 282, 381). The relationship of this tree is with B. glandulosa Michx., from which it differs in the shape and venation of the leaves, in the pendulous fruiting catkins, and in its arborescent habit. Betula commixta, n. hyb. ? {B. alaskana X glandulosa?).— Leaves broadly ovate to elliptic, acute at apex, broad-cuneate or rounded at base, coarsely serrate with blunt or acute teeth, thin, glabrous, smooth, dark green and lustrous above, pale and lustrous below, 3-4.5 cm. long and broad; petioles 1.5 cm. in length. Flowers not seen. Fruiting catkins erect, 2 cm. long, 6-7 mm. in iqiqI SARGENT— north AMERICAN TREES 217 diameter, their scales puberulous, the terminal lobe acute, one- third longer than the rounded lateral lobes. A shrub 2-3 m. tall, with dark brown stems and slender gray-brown branchlets thickly covered with resinous glands. On the tundra with B. glandidosa in the neighborhood of Dawson, British Alaska, Alice Eastwood, Ten Mile House, June 25, 1914 (no. 367 type); Twenty- four ]\Iile House, June 27, 1914 (no. 400). The proper disposition of this plant is doubtful, and it should perhaps be considered a species. The glandular branchlets and slender erect fruiting catkins resemble those of B. glandidosa Michaux, but the larger, acute, sharply serrate leaves are not of that species, and if it is a hybrid the size and serration of the leaves can only have been derived from B. alaskana Sargent. Celtis occidentalis L. — On what is usually considered the t}TDe of this species the leaves are broadly ovate, acute or short- acuminate at apex, obliquely rounded at base, coarsely or finely serrate, smooth on the upper surface, glabrous or sparingly pilose along the midribs and veins below, thin, not conspicuously venu- lose; petioles glabrous or rarely puberulous. The fruit is borne on glabrous or rarely puberulous pedicels much longer than the petioles and is subglobose, ellipsoidal, or slightly obovoid, and 9-10 mm. in diameter; the stone is only slightly reticulate. The branchlets are glabrous or occasionally pubescent. C. occidentalis is distributed from New England to Virginia and westward to Iowa, southwestern Missouri, western and central Kansas, and eastern North Dakota. It is less common and usually a smaller tree than its varieties canina and crassifolia, and much less widely distributed than the latter. All the forms of C. occidentalis are well distinguished by the dark purple fruit, which is larger than that of the other American species; it is borne on longer pedicels than that of our other species, with the exception of that of C. Douglasii. Celtis occidentalis var. canina, n. var. — C. canina Rafinesque, Am. Monthly Mag. 2:43. 1817; Planchon, DeCandolle Prodr. 17:174. 1873; Britton and Shafer, N. Am. Trees, 355. 1908. — C occidentalis Sargent, Silva N. Am. 7:67 (in part, pi. 317, not Linnaeus). 1895; Hough, Trees N. States and Canada, 193 (in part, fig. 2iy). 1907. — Differing from the type in the usually narrower long-acuminate leaves. Extreme forms of this variety look very distinct, but trees with leaves intermediate between these and those of the typical form are common. The 2i8 BOTANICAL GAZETTE [makch fruit varies as in the type from subglobose to obovoid, and there seems Httle difference in the length of the pedicels, which are always longer than the petioles. The leaves are usually glabrous, but on some of Bushes Missouri specimens the midribs and veins are pilose on the lower surface and the petioles are pubescent, as in the variety crassifolia (Monteer, nos. 548, 4725; Christian County, no. 4664; Dumas, no. 5905). This variety is distributed from the Province of Quebec to Iowa, Nebraska, North Dakota, and southwestern Missouri, southwestern Oklahoma, New York, and Ohio, and to northwestern Georgia (Cobb County, R. M. Harper, no. 166 in Herb. Gray). More distinct is Celtis occidentalis var. crassifolia Gray, Man. ed. 2, 397. 1856. — C. crassifolia Lamarck, Encycl. Meth. 4:138. 1797. — Differ- ing from the type in its usually narrower, acuminate, thicker leaves, often more coarsely serrate or nearly entire, scabrate on the upper surface and pilose below along the midribs and veins. In this form the petioles are usually villose-pubescent, but occasionally are quite glabrous; the pedicels are slightly villose, and the branchlets are glabrous or pubescent. I have seen specimens of this variety from Viriginia and West Virginia; North Carolina, A. Gray, Painted Rock, French Broad River, 1843 (in Herb. Gray); river banks, Biltmore (ex herb. Biltmore no. 12 10, with nearly entire leaves); Nashville, Tennessee; southern Indiana, wooded bluff of Blue River, near Middletown, Crawford County, C. C. Deam, June 15, 1915 (no. 16423, with coarsely serrate leaves, "a flat-topped shrub about 8 ft. high"), June 25, 1915 (no. 16418, the leaves entire or furnished with occasional teeth, "a shrub 8 ft. high in the dense shade of walnut and buckeye trees"); wooded bluff of the Ohio River, 6 miles east of Cannelton, Perry County, C. C. Deam, June 29, 191 5 (no. 16627, a tree 8 m. high with nearly entire leaves); southern and western Illinois; Fort Snelling, Minnesota; northern and southern Mis- souri; central Kansas; eastern and northwestern Oklahoma; Thomas County, central Nebraska; Bigstone, eastern South Dakota; central North Dakota; from the Tongue River Canyon, Big Horn Mountain, Wyoming; from the Black Canyon of the Boise River and the valley of the Clearwater River, Nez Perce County, Idaho; from Berlin, Dallas County, Alabama, R. S. Cocks, 1913 (with nearly entire leaves), and from Larissa, Cherokee County, B. F. Bush, April 30, 1909 (no. 5561); and Livingston, Polk County, Texas, E. J. Palmer, October 9, 1914 (no. 6785). Celtis Douglasii Planchon, Ann. Sci. Nat. III. 10:293. 1848; Piper, Contrib. U.S. Nat. Herb. 2:221 (Fl. Washington). 1906; Britton and Shafer, N. Am. Trees, 359. fig. 31Q. 1908.— C. relicu- lata Howells, Fl. N.W. America 602 (not Torrey). 1897.— C. rugosa iQig] SARGENT— NORTH AMERICAN TREES 219 Rydberg, Bull. Torr. Bot. Club 39:304 (not Newberry). 1912.— C. rugidosa Rydberg, Fl. Rocky Mountains, 207. 1061. 1917.— This tree, which has sometimes been considered the same as the more southern C. reticulata Torrey, can be distinguished from that species by its rather thinner, oblong-ovate, long-acuminate, coarsely serrate leaves, cordate or obliquely cordate at base, glaucescent on the lower surface and glabrous or sparingly pilose on the under side of the midribs and veins, by the slightly pilose petioles, and by the much longer pedicels of the fruit sometimes up to 1.5 cm. in length. The fruit, which has been described as "black or brownish," is light orange-brown on the Oregon and Colorado trees, and is subglobose to ellipsoidal and 7-8 mm. in diameter. C. Douglasii is a shrub or a small tree rarely 10 m. high, with rough, red- brown, or in Colorado dark gray bark 2.5 cm. thick and irregularly ridged, glabrous or slightly pilose branchlets, and pubescent and tomentose winter- buds. Nowhere common, it is widely distributed on dr>'- ridges and the rocky banks of streams, and occurs in Oregon east of the Cascade Range in the valley of the Deschutes River, on the rocky banks of the Columbia near the Dalles and on Pine Creek, Gilliam County, and in western Washington ranges from the valley of the Columbia in Klickitat County to the rocky banks of Snake River in Whitman County, and to Big Willow Creek, Canyon County, Idaho; it inhabits the western foothills of the Wasatch Mountains of Utah, south- eastern Utah (Grand River Canyon below Moab, Grand County), the southern slope of the Grand Canyon, Arizona {A. Rehder, July 19, 1914, no. 103), "in sand at the mouth of a dr}- canyon one-half mile below Democrat Springs, Kern River," Kern County, California, Mrs. Leo Polkingham, September 1916 (in Herb. Dudley), and on the eastern foothills of the Rocky Mountains of Colorado. In the shape of the coarsely serrate leaves and in the long pedicels of the fruit C. Douglasii is related to C. occidentalis var. crassifolia, from which it differs in its thicker leaves with conspicuous reticulate veinlets and usually glaucescent and less pubescent on the lower surface, and in the color of the fruit. In the thick reticulate- venulose leaves rough on the upper surface it resembles C. reticulata. Geographically C. Douglasii is intermediate between C. occidentalis, which in its var. crassifolia reaches northern Idaho, and C. reticulata, which extends northward to the Grand Canyon in Arizona. Celtis Lindheimerii K. Koch apud Engelmann, Dendr. 2:434. 1872.— C. Helleri Small, Bull. Torr. Bot. Club 24:439. 1897; 220 BOTANICAL GAZETTE [march Britton and Shafer, N. Am. Trees 358. fig. ji8. 1908; Mackensen, Trees and Shrubs of San Antonio, 17. pi. j. 1909. — Koch's descrip- tion of C. Lindheimerii was made from a tree growing in the Botanic Garden in Berhn which had been raised from seeds gathered at New Braunfels, Texas, by Lindheimer and sent by Engelmann to Berhn. Koch's description leaves no doubt that C. Lindheimerii (Engelmann in Herb. A. Braun) is the tree with leaves pale and densely pubescent on the lower surface and pubescent branchlets which is common at New Braunfels and in the neighborhood of San Antonio. Of the specimens of this tree which I have seen the oldest was collected by Drummond in 1834 without locality but probably near Austin (nos. 343 ?, 334, 259 in Herb. Gray). It was collected by Lindheimer at New Braunfels, Comal County, in 1850 (no. 444 in Herb. Gray). Mohr collected it in the valley of the Comal River, New Braunfels, in 1850. I collected it at San Antonio, Bexar County, in 1881, and Bush also collected it at San Antonio October 1900 (no. 1246), September 1901 (no. 797), and March 1902 and 1903 (nos. 1 172, 3677). Palmer's collections of this tree are from Sutherland Springs, Wilson County (no. 9302), Goliad, Goliad County (no. 9128), San Marcos, Hays County (no. 13311), dry limestone banks. South Llano River, Telegraph, Kimble County (no. 10931). C. Lindheimerii is most abundant in the neighborhood of streams and springs and occurs less commonly on higher ground. I have no evidence that it grows on the Edwards Plateau or westward. Celtis reticulata Torrey, Ann. Lye. N.Y. 2:247. 1828; Planchon, Ann. Sci. Nat. III. 10:293. 1848. — C. occidenlalis var. reticulata Sargent, Forest Trees N. Am. loth Census U.S. 9:126. 1884; Garden and Forest 3:40. ^g. 12. 1890. — C. mississippiensis var. reticulata Sargent, Silva N. Am. 7:72 (in part). 1895; Man. 301. fig. 243. 1905. — C. reticulata, C. Lindheimerii, and C. Douglasii are similar in their thick leaves, rough on the upper surface and conspicuously reticulate venulose below, and in their pedicels longer than the petioles. The entire leaves green on the lower surface and the orange-red fruit on shorter pedicels of C. reticulata distinguish it from C. Douglasii. The shape of the leaves of C. Lindheimerii, pale and pubescent below over their whole surface, makes it easy to distinguish that species from C. reticulata. I have not seen the flowers and spring leaves of C. reticidala, which, although it was described 90 years ago, is still very imperfectly known. The young iqiq] SARGENT— north AMERICAN TREES 22 1 leaves, like those of the other species of this group, probably show little evidence of the prominent veinlets which are not conspicuous on them before early summer. In collections of Texas plants C. Lhidheimerii and more frequently C. laevigata var. texana have been confounded with C. reticulata. Seedlings raised at the Arboretum from seeds of a tree with rough, mostly entire leaves growing at the Montezuma Well in central Arizona {Relider, no. 537) have coarsely serrate leaves which are rough or nearly smooth above on the same branchlet. From western Texas I have seen specimens of C. reticulata collected in Uvalde, Kimble, Mitchell, Nolan, Callahan, Randall, Hartley, and Jeff Davis counties. It ranges into western Oklahoma, and through southern New Mexico to southern, central, and northeastern Arizona, and occurs on Cedros Island off the coast of Lower California {Veatch, 1872, in Herb. Gray). Celtis reticulata var. vestita, n. var.— Differing from the type in its more pubescent serrate leaves and more pubescent petioles. Leaves broadly ovate, acute or acuminate at apex, unsj-mmetrically rounded or subcordate at base, the margins thickened, ciliate, and sharply but irregularly serrate, thick, dark green and scabrate above, paler and coated below with short pale pubescence with longer hairs on the slender midribs, primary veins, and conspicuoulsy reticulate veinlets, 3.5-4.5 cm. long and 3-3.5 cm. wide; petioles densely tomentose, 4-5 mm. in length; leaves on vigorous shoots acuminate, mostly cordate at base, more coarsely serrate, rugose, and covered above with short white hairs and more densely pubescent below, 6-8 cm. long and 4-4.5 cm. wide, their petioles thickly covered with matted pale hairs, 8-9 mm. in length. Fruit orange-red, 6-7 mm. in diameter. A small tree with a trunk 20-25 cm. in diameter and slender pubescent branchlets, those of \dgorous shoots stouter and densely villose. Near Canton. Blaine County, Oklahoma, in low ground along the North Fork of the Canadian River, D. M. Andrews, August 15, 191 5 (nos. 21, 49 type). Specimens of vigorous sterile branches collected by Palmer at Canadian, Hemphill County, Texas, June 17, 1918 (no. 14109), may prove a distinct form of this variety. The leaves are acuminate, obliquely cordate at base, coarsely serrate, more pubescent below, 6-9 cm. long and 3 . 5-5 cm. wide; the petioles and branchlets are densely tomentose. Celtis laevigata Willdenow, Berl. Baum. ed. 2. 81. 181 1.— C. mississippiensis Bosc in Encycl. Met. Agric. 7:577 {nomen nudum). 1821; Spach, Arm. Sci. Nat. II. 16:42. 1841.— C. Ber- landierii Klotzsch, Linnaea 20:541. 1847; Parlatore, DeCandolle 222 BOTANICAL GAZETTE [march Prodr. 17: 178 (in part). 1873. — For this tree of the southern states the name C. mississippiensis has usually been adopted. Bosc pub- lished a brief account, without a name or technical description, of Le Microcoulier de la Louisiana cultivated in France in the Nouveau Cours Complet d^ Agriculture (8:529. 1809), and republished it in the second edition of this work (10:41. 1822). As his plant came from near the mouth of the Mississippi River there is little doubt of its identity with the C. mississippiensis of Spach, for it was Spach who first described this tree as C. mississippiensis, Bosc's earlier C. mississippiensis being a nomen nudum and 10 years later than C. laevigata of Willdenow which, following K. Koch, must be taken up for our tree. A cotype of C. Berlandierii Klotzsch (no. 2318), collected by Berlandier at "Matamoras de TamauHpas," April 1 83 1, is preserved in the Gray Herbarium on a sheet with Berlandier^s no. 885, also collected at Matamoras in April 1831. A fragmentary specimen with flowers, collected by Berlandier in February 1828 (no. 1487-2271), a fruiting branch without locality or date and numbered 2429 on a label printed for the Gray Her- barium, and a vigorous shoot (no. 2429-999), collected by Ber- landier, May 1824, "De Goliad a Bexa," are mounted on another sheet. The leaves and fruit of these Berlandier specimens only differ in their smaller size from specimens of the leaves and fruit of C. laevigata grown on the rich bottom lands of the Mississippi Valley, a difference which the dryness of the region where Ber- landier collected them explains. C. laevigata, when it grows under favorable conditions, is a tree sometimes 30 m. high, with somewhat pendulous branches and slender, glabrous, red-brown branchlets. The leaves are thin, usually oblong-lanceolate, long-pointed and acuminate at apex, unsymmetrically rounded and often oblique or cuneate at base, frequently more or less falcate, entire or furnished with a few teeth usually toward the apex, green .on both surfaces, glabrous, smooth or occasionally scabrate above. The fruit is bright orange-red on pedicels shorter or slightly longer than the petioles. C. laevigata is distributed from the coast of Virginia to the Everglade Keys of southern Florida and through the Gulf states to the valley of the lower Rio Grande in Nuovo Leon, Mexico, and through eastern Texas and Arkansas to iqiq] SARGENT— north AMERICAN TREES 223 eastern Oklahoma and Kansas, northern Missouri, southern Illinois, and south- western Indiana; also in Bermuda. Trees occasionally occur with leaves more or less sharply serrate nearly to the base, and this variety may be distin- guished as Celtis laevigata var. Smallii, n. var.— C. Smallii Beadle, Small Fl. S. United States, 365. 1329. 1903.— Differing from the type only in its constantly serrate leaves. Celtis laevigata var. texana, n. var.— C. texana Scheele, Linnaea 22:146. 1847.— C. Berlandierii Planchon, DeCandolle Prodr. 17:178 (in part, not Klotzsch). 1873.— Differing from the t>T3e in the shorter ovate to lanceolate thicker leaves often pubescent on the midribs and veins below, in the often more prominent veinlets and pubescent petioles, and in its often pubes- cent branchlets. Leaves ovate to lanceolate, acuminate, unsym- metrically rounded or cordate at base, entire or sparingly and irregularly serrate, often subcoriaceous, dark green, smooth and granulate or rarely scabrate above, green below, with slender midribs and primary veins glabrous or sparingly villose- pubescent and furnished with small axillary tufts of pale hairs, and thin, only slightly raised, reticulate veinlets, 3.5-7 cm. long and 2-3.5 cm. wide; petioles slender, pale pubescent, 5-7 mm. in length. Flowers not seen. Fruit subglobose but rather longer than broad, dark orange-red, 6-7 mm. in length; pedicels glabrous or puberulous, slightly longer than the petioles. An arborescent shrub or small tree rarely more than S m. high, often grow- ing in groups, with pale or grayish rough bark rarely covered with wartlike excrescences, and slender, reddish, glabrous or gray-brown, pubescent branch- lets. I have taken up Scheele's name for the common Celtis of the Edwards Plateau and western Texas with some hesitation, for I have not seen his type specimen. His description perfectly applies, however, to a specimen collected by Lindheimer at New Braunfels in the herbarium of the Arboretum (no. 4 ex herb. Engelmann as C. texana). On other specimens collected by Lind- heimer at New Braunfels the lower side of the midribs and veins of the leaves are sparingly villose-pubescent, and such pubescence is found on the leaves of most of the specimens which I have referred to this variety. Texas. — New Braunfels, Comal County, F. Lindheimer, 1850 (nos. 1158, 1 1 59 in Herb. Gray, distributed by the Mo. Bot. Card, as C. Berlandierii); no. 4 (in Herb. Arnold Arboretum ex Herb. Engelmann, type) ; Comanche 224 BOTANICAL GAZETTE [march Spring, New Braunfels, C. Mohr, December 1880; San Antonio, Bexar County, B. F. Bush, October 2, 1900 (no. 1223), September 16, 1901 (no. 807), March 23, 1902 (no. 1 174), B. Mackensen, October and December 1910; Sutherland Springs, Wilson County, E. J. Palmer, March 30, 1916 (no. 9297); Pleasanton, Atascosa County, E. J. Palmer, September 23, 1916 (no. 10773); rocky river banks, Blanco, Blanco County, April 6, 1918 (no. 13294), Berlandier "entre le rio de las Nueces et Laredo," June 1820 (nos. 601, 201 1 in Herb. Gray); Boerne, Kendall County, E. J. Palmer, May 18, 1918 (no. 13644, with hoary tomentose branchlets); Kerrville, Kerr County, B. Mackensen, May i, 1910; Uvalde, Uvalde County, E. J. Palmer, April 10 and May 6, 1918 (nos. 13324, 13509, with coarsely serrate leaves scabrate on the upper surface) ; Sweet Water, Nolan County, E. J. Palmer, July 6, 1917 (nos. 12432, 12433); Barksdale, Edwards County, E. J. Palmer, May 7, 1918 (no. 13519) ; banks of Devil's River, Valverde County, E. J. Palmer, May 14, 1918 (no. 13601); Limpia Canyon, Jeff Davis County, Tracy and Earl, April 25, 1902 (no. 238 in Herb. Gray); Post, Garza County, E. J. Palmer, May 31, October i, 1918 (nos. 13837, 14565); banks Canadian River, Canadian, Hemphill County, E. J. Palmer, June 17, 1918 (no. 14110); Strawn, Palo Pinto County, E. J. Palmer, June 27, 1918 (no. 14256); Gamble's Ranch, Armstrong County, June 6, 1918 (no. 13969); along the Brazos, Graham, Young County, /. Reverchon, Octo- ber 27, 1902 (nos. 3262, 3267); dry rocky hillsides, Baird, Callahan County, E. J. Palmer, May 26, 1918 (no. 13680); Paloduro Canyon, Potter County, /. Reverchon, May 22, 30, 31, 1902 (nos. 3265, S2 f- f. 2930, 2930 bis), E. J. Palmer, June 3, 1918 (no. 13866), C. S. Sargent, April 1916; TivoH, Refugio County, E. J. Palmer, March 22, 1916 (no. 9250, with broadly ovate coarsely serrate leaves and pubescent branchlets); Austin, Travis County, G. W. Letterman, August 23, 1892; Dallas, Dallas County, /. Reverchon, August i860 (no. 854), B. F. Bush, September 30 and October 27, 1900 (nos. 1190, 1614, 1615), April 8, 1904 (no. 4254); rocky bluft"s, Dallas County, /. Reverchon, April 15, 1902 ; bottom lands Elmo, Kaufman County, /. Reverchon, October 22, 1902 (no. 3263); "sandy woods, Forks of the Trinity," Dallas County, /. Reverchon, (no. 3268); along the Brazos, Bryan, Brazos County, E. J. Palmer, April 26, 1918 (no. 13459); Velasco, Brazoria County, E. J. Palmer, March 21, 1918 (no. 13133); Columbia, Brazoria County, E. J. Palmer, September 29, 1914 (no. 66760) ; San Augustine, San Augustine County, E. J. Palmer, April i, 1918 (no. 13245); Houston, Harris County, E. J. Palmer, March 19, 1918 (no. 13 1 13); E. N. Plank, Eagle, Shelby County. OKLAHCMiA.— Sand hills, Ingersoll, Alfalfa County, C. 5. Sargent, May 6, 1902; near Page, Leflore County, G. W. Stevens, September 1913 (no. 2724); low woods along river, CUnton, Custer County, E. J. Palmer, July 16, 1917 (no. 1255). New Mexico.— yl. Fendler, without locahty, 1847 (no. 775 in Herb. Gray); Berendo Creek, Sierra County, C. B. Metcalf, May 23, 1904 (no. 926); bank of creek near Roswell, Chaves County, A. Rehd'er, August 16, 1916 (no. 352). igig] SARGENT—NORTH AMERICAN TREES 225 Kansas. — Woods, Barber County, A. S. Hitchcock, 1896 (no. 815 in Herb. Gray) ; vicinity of Huntsville, Walker County, R. A . Dixon, July 1909 (no. 387 in Herb. Gray). Missouri. — Willard, Greene County, /. W. Blankinship, August 2, 1893; Noel, McDonald County, B. F. Bush, August 7, October 12, 1908 (nos. 4977, 5255), E. J. Palmer, September 5, 1913, May 5, 1914, October 11, 1918, (nos. 4141, 5795. 14665, 14668); "along rocky banks," Carthage, Jasper County, E. J. Palmer, 191 7, Webb City, E. J. Palmer, September 28, 1908; Knight's Station, C. S. Sargent and E. J. Palmer, October 8, 191 1 (no. 3489), July 13, 1913 (no. 4019). Mexico. — Reynosa, Tamaulipas, on the lower Rio Grande, C. G. Pringle, August 7, 1888 (no. 2082); SaltiUo, Coahuila, Ed. Palmer, 1898. Celtis laevigata texana f. microphylla, n. f.— Differing from the variety in its smaller leaves with more prominent reticulate veinlets, and more densely villose-pubescent petioles. Leaves broadly ovate, acute, unsymmetrically rounded at base, smooth, dark green and granulate above, yellow-green below, with villose pubescent midribs and veins and conspicuous reticulate veinlets, 2-2.5 cm. long and 1.5-2 cm. wide. Flowers and fully grown fruit not seen. A shrub with slender, red-brown branchlets densely pubescent in their first season, becoming puberulous in their second year. Rocky banks of streams, Sweet Water, Nolan County, Texas, E. J. Palmer, May 27, 1918 (no. 13751 type). Celtis laevigata var, brachyphylla, n. var. — Diflfering from the type in the shorter thicker leaves. Leaves ovate, acuminate and long-pointed at apex, very oblique and rounded or cordate at base, entire, occasionally furnished on one side near the base with a broad rounded lobe, glabrous, thick and firm, green on the two surfaces, 3.5-4 cm. long and 2-3 cm. wide, with slender midribs and veins; petioles slender, glabrous, 5-6 mm. in length. Fruit subglobose to short-oblong, bright orange-red, 6-7 mm. in diameter; pedicels glabrous, longer than the petioles. ^ A tree about 10 m. tall, with slender, glabrous, dark red-brown branchlets. Rocky banks of the canyon of the Nueces River near Uvalde, Uvalde County, Texas, E. J. Palmer, September 26, 1918 (no. 14517 type). Celtis laevigata var. anomala, n. var. — Differing from the type in its oblong-ovate, acute leaves, cordate or unsymmetrically 226 BOTANICAL GAZETTE [march cordate at base, and dark purplish fruit covered with a glaucous bloom. Leaves entire, glabrous, dark green and scabrate above, paler and conspicuously reticulate venulose below, 2-4 cm. long and 1.5-2 cm. wide; petioles sparingly villose pubescent, 4-6 mm. in length. Fruit short-oblong, 6-7 mm. in length, the pedicels as long as or slightly longer than the petioles. A shrub from i . 5-2 cm. tall with slender pubescent branchlets. "In deep sands, among shin oak and Quercus marylandica," Clyde, Callahan County, Texas, E. J. Palmer, September 30, igi8 (no. 14550, type). Celtis laevigata var. brevipes, n. var. — C. brevipes S. Watson, Proc. Am. Acad. 14:297. 1879. — Differing from the type in its yellow fruit on shorter puberulous pedicels and in its puberulous petioles. Leaves ovate, acuminate, unsymmetrically rounded or cuneate at base, entire or rarely furnished with occasional teeth, glabrous, dark green and smooth or slightly roughened on the upper surface, yellow-green below with small clusters of pale hairs in the axils of the slender veins, inconspicuously reticulate venulose, 3.5-5 cm. long and 1-2.5 cm. wide; petioles slender, puberulous, 5-7 mm. in length. Flowers not seen. Fruit short-oblong, canary yellow, 5-6 mm. long; pedicels puberulous, shorter or slightly longer than the petioles. A small tree with slender, glabrous, red-brown branchlets. Arizona. — Camp Grant, Pinal County, T. J. Rothrock, July 1874 (no. 337, type of C. brevipes in Herb. Gray), J. G. Lemmon, 1880 (no. 68 in Herb. Gray) — ; along irrigating ditches, Tucson, Pima Coimty, Engelmann and Sargent, Sep- tember 24, 1880; banks of Rillito Creek near Tucson, A. Rehder, August 7, 1917 (no. 240 type); Apache Trail, Fish Creek, Alice Eastwood, April 19, 1917 (no. 6250); Crow Creek Canyon, Chiricahua Mountains, Cochise County, J. W. Tourney, July 1894 ("tree i8°X2o'-3o"'); Tonto Basin, Gila County, /. W. Tourney, June 19, 1892. Utah. — River Canyon, southeast Utah, Alice Eastwood, June 1892. California. — ^Laguna, San Diego County, D. Cleveland, July 10, 1885. The specimens of C. brevipes in Herb. Gray and the specimen from San Diego County, CaHfornia, have only partly grown fruit, but in the shape of the leaves they so closely resemble the yellow-fruited Arizona trees that there can be little doubt that they are of the same variety, although on the CaU- fornian specimen the leaves are distinctly rough on the upper surface, as they are in the specimen collected by Engelmann and Sargent at Tucson. 191 q] SARGENT— north AMERICAN TREES 227 Celtis pumila Pursh, Fl. Am. Sept. 1:200. 1814; E. J. Hill, Bull. Torr. Bot. Club 27:497. pi. jj. 1900. — C. occidentalis B pumila Muhlenberg, Cat. 100 (nonien nudum). 1813; Gray, Man. ed. 2. 397. 1856. — C. niississippiensis var. pumila Mackensen and Bush, Man. Fl. Jackson County, Missouri, 72. 1902. — Often considered a variety of C. occidentalis, C. pumila can be separated from that species by its smaller, usually entire, rather thicker leaves, by its small, dark reddish purple fruits on pedicels shorter or only slightly longer than the petioles, by its more deeply pitted nutlet, and shrubby habit. The color of the fruit and its short pedicels indicate a nearer relationship with C. laevigata than with C. occidentalis. C. pumila has been found in Pennsylvania, Delaware, and the District of Columbia, western New York, northern Indiana and Illinois, middle Tennessee, northeastern Mississippi, near Augusta, Georgia, near Lake Okeechobee, Florida (Ed. Palmer, 1874, no. 515 in Herb. Gray), Missouri, and northeastern Arkansas (Eureka Springs, Carroll County, £. /. Palmer, no. 4409). A branch without fruit, with small, nearly entire, glabrous leaves, collected by Mohr on uplands west of Franklin, Alabama, October 8 (no year, no. 66), and described as "a low spreading tree of slender growth," appears in spite of its habit to be C. pumila. A glabrous specimen with pedicels as long as or longer than the petioles collected on the sandy seashore at Hillsboro, Florida, hyp. A. Marten (no. 6506 in Herb. Gray) is probably from a depauperate form of C. laevigata. Celtis pumila var. georgiana, n. var. — C. occidentalis Abbot and Smith, Insects of Georgia, pi. j6 (not L.). 1797. — C. occidentalis var. pumila Chapman, Fl. 417 (not Muhlenberg). 1865. — C. geor- giana Small, Bull. Torr. Bot. Club. 24:439. 1897; Britton and Shafer, N. Am. Trees, SSltfiS- 316. 1908. — Differing from the type in the rugose upper surface of the leaves, more or less densely pilose along the midribs and -veins below, in the pilose petioles, and puberulous pedicels. A shrub or small tree occasionally 10 m. tall, with slender pubescent branchlets sometimes becoming glabrous by the end of their first season. The fruit was described as "tan" color by Small, and by Britton and Shafer as "red-purple to yellowish." The fruit when fully grown in early summer is dull yellow in color, but by the middle of October it becomes reddish purple like that of C. pumila, and is often covered with a glaucous bloom. 228 BOTANICAL GAZETTE [march C. pumila var. georgiana occurs on rocky blviffs, Franklin Furnace, Sussex County, New Jersey {K. K. MacKenzie, August 22, 1909, no. 4331), and ranges from the Piedmont region of North Carolina (Raleigh, Wake County, T. G. Harbison, June 11, 1918) to western Florida, Autauga and Dallas counties, Alabama, and occurs in southern Missouri (rocky hills and bluffs, B. F. Bush, Swan, Taney County, September 23, 1905 [no. 5040], Monteer, Shannon County, August 18, 1901 [no. 703], Noel, McDonald County, August 9, 1908 [no. 5040]). The oldest specimen of this plant which I have seen was collected near Augusta, Georgia, by Olney and Metcalf in 1855 (in Herb. Gray). On the rocky wooded slopes and ridges of Lawrence, Orange, Washington, Crawford, Perry, Floyd, and Harrison counties in the extreme southern part of Indiana a dwarf Celtis occurs in a few isolated stations. In the general outline of the leaf it resembles C pumila, but the pedicels of the lighter-colored fruits are much longer; the leaves, which are smooth or nearly so on the upper surface and rather thicker than those of C. pumila, are slightly pubescent along the under side of the midribs and veins and on the petioles; the branchlets are usually puberulous. Judged by the present inadequate information now accessible concerning this plant it appears intermediate between C. pumila and its variety georgiana, although the nutlets are smoother than those of C. pumila, and it may be distinguished as Celtis pumila var. Deamii, n. var. — Leaves broadly ovate to oblong-ovate, acuminate and often long-pointed at apex, unsym- metrically rounded at base, entire or occasionally sharply and irregularly serrate above the middle, thick, dark green on both surfaces, smooth or slightly roughened above, 3-nerved, reticulate venulose, slightly villose pubescent on the prominent midribs and veins, 6-8 cm. long and 3.5-4 cm. wide; petioles slender, villose pubescent, 8-10 mm. in length; leaves on vigorous shoots acute or acuminate, obliquely rounded at base, thicker, entire, scabrate above, often 10 cm. in length. Flowers not seen. Fruit sub- globose, ellipsoidal or slightly obovoid, tan color or orange until after midsummer, becoming when fully ripe dark orange-red, 7-8 mm. in diameter; pedicels 6-15 mm. long. An arborescent shrub 2-4 m. high, with a stem 3 . 5-5 cm. in diameter, covered with dark, rough, deeply fissured bark and slender, reddish brown, slightly pubescent branchlets. iqiq] SARGENT— north AMERICAN TREES 229 C. C. Deani, 6 miles from Derby, Perry County, July 4, 191 2 (no. 11 502), near Mitchell, Lawrence County, August 16, 1912, September 2, 1915, Au- gust 13, 1918 (nos. 12052, 12055, 18474, 18479, 26218), wooded bluffs of the Ohio River west of Leavenworth, September 11, 1915 (nos. 18586, 18589), on a bluff 3 miles south of New Middletown, Harrison County, September 6, 191 5 (no. 18727 type), near Big Springs, Washington County, September 12, 1915 (no. 8987), wooded blufif, Indian Creek, near Corydon, Harrison County, August 15, 1918 (no. 26232), near Elizabeth, Harrison County, September 26, 1918 (no. 26798). I take much pleasure in associating with this plant the name of Charles Clemon Deam, State Forester of Indiana, who for many years has industriously studied the trees of Indiana. Persea pubescens (Pursh) Sargent. — The first varietal name pubescens of Pursh (18 14) was taken up for this tree in The Silva of North America (7:7. 1895); but under the rules adopted by the Vienna Congress the first specific name must be used, and this is paliistris {Tamala palustris Rafinesque, Fl. Tellur. 137. 1838), and Persea pubescens (Pursh) Sargent should become P. palustris (Rafinesque) Sargent. Platanus occidentalis L. — The leaves of the northern plane tree are broadly ovate, 3-5-lobed by broad shallow sinuses rounded in the bottom, cordate or truncate at base, becoming glabrous except on the under side of the midrib and principal veins, the lobes broad, acuminate, serrate-toothed with long straight or curved remote acuminate teeth. Individual trees with leaves cuneate at base may be distinguished as Platanus occidentalis f. attenuata, n. f. — Dift'ering from the type in the long cuneate base of the usually less deeply lobed leaves. I have seen specimens of this form from Selma, Dallas County, Alabama, T. G. Harbison, June 31, 1916 (no. 51 type). Mississippi: T. G. Harbison, Pelahatchee, Rankin County, May 26, 1915 (no. 14), Jackson and Bolton, Hinds County, June 28, 29, 1916. Texas: E. J. Palmer, banks of Peyton's Creek, Matagorda County, May 6, 1916 (no. 9683), rocky creek banks, Boerne, Kendall County, May 19, 1916, Lacey's Ranch, Kerr County, May 31, 1916 (no. 9982). Oklahoma: G. W. Stevens, valley of the Chikaskia River near Tonkawa, Kay County. Missouri: B. F. Bush, near Monteer, Shannon Coimty, September i, 1911 (no. 663). Indiana: C. C. Deam, low banks of the Wabash River, near Murray, Wells County, May 26, 1916 (no. 19817). 230 BOTANICAL GAZETTE . [march Platanus occidentalis var. glabrata, n, var. — P. racemosa Hemsley, Bot. Biol. Am. Cent. 3:162 (not Nuttall). 1882.— P. glabrata Fernald, Proc. Am. Acad. 36:493. 1901. — P. densicoma Dode, Bull. Soc. Dendr. France 7:67. 1908.— Differing from the type in the 3-lobed leaves, truncate, broad-cuneate or rarely slightly cordate at base. Leaves usually broader than long, truncate, broad-cuneate or rarely cordate at base, 3-lobed by sinuses acute or rounded in the bottom, the lobes long-acuminate, entire, the lateral often furnished near the base with one or rarely with two small acuminate incurved secondary lobes, occasionally found also on the terminal lobe; when they unfold hoary tomentose below and pubescent above; pubescent when the flowers open toward the end of March, and in early summer pubescent along the under side of the midribs and veins but otherwise glabrous, usually 7-14 cm. long and 8-9 cm. wide, their petioles pubescent, becoming glabrous; peduncles with one or occasionally two heads of flowers and fruit. Described from specimens collected in the Provinces of Coahuila and Nuovo Leon, Mexico, this Platanus has been found not to be uncommon in western Texas, where it has been collected by 5. B. Buckley near Austin (with- out date or number), by A. A. Heller, Kerrville, Kerr County, April 1894 (no. 1622), and by E. J. Palmer on the banks of the Llano River at Llano, Llano County, June 23, 1916 (no. 10279), rocky banks of upper Seco Creek, Bandera County, May 18, 1916 (no. 10241), gravel bank, Nueces River, Uvalde, Uvalde County, September 24, 1918 (no. 14480), Fredericksburg Junction and Boerne, Kendall County, June 5 and May 19, 1916 (nos. 9817, 9826, 10069), rocky banks of the Guadalupe River, Kerrville, Kerr County, May 29, 1916 (no. 9921), rocky banks of upper Seco Creek, June 18, 1916 (no. 10241), Utopia and Sabinal, Uvalde County, April 10, 191 7, June 7, 1916 (nos. loioo, 11523), rocky banks, Devil's River, Valverde County, October 18, 1916, March 26, 191 7 (nos. 11084, 11371), Palliam, Zavalla County, March 21, 1917 (no. 11332). The close connection of this variety with typical P. occidentalis is shown by the appearance on leading shoots of 5-lobed leaves with serrate lobes (Palmer, Sabinal, Uvalde County, no. loioo, and Devil's River, Valverde County, no. 110841). More significant perhaps is the fact that occasionally trees occur growing with P. occidentalis north of Texas which cannot be dis- tinguished from the Mexican types of P. glabrata. Such specimens are those of E. J. Palmer, Choctaw County, Oklahoma, July 13, 1916 (no. 10463), the type of P. densicoma collected in the Maquoketa River, Jackson County, igig] SARGENT— NORTH AMERICAN TREES 231 eastern Iowa, preserved in Herb. Mus. Paris (photograph Herb. Arnold Arbore- tum), Biltmore, North Carolina (Herb. Bilt. no. 1271b), and of John Robinson, Brookline, Massachusetts, June 1880. Magnolia virginla.na L. — M. glatica L. — This species was based on the Tulipijera virginiana Plukenet, Ahn. Bot. 379. pi. 68, and the Magnolia foliis ovato-lanceolatis Linnaeus, Hort. Cliff. 222; Gronovius, Fl. Virg. 61; and the Magnolia Lauri folia siibtiis albi- cante Catesby, Nat. Hist. Car. 1:39. pi. 30. There are two distinct forms of this tree, one with glabrous branchlets and pedicels and usually narrow leaves, and one with branchlets and pedicels thickly clothed with long, silvery white hairs and often broader leaves. Specimens preserved in the British Museum show that the former is the type of Linnaeus' species. Tulipijera virginiana, etc., of Plukenet is represented in the Sloan Herbarium by 2 specimens, one in Plukenet's own herbarium, the other one of a number of plants collected in Maryland by Dr. Krieg and a Mr. Vernon. The latter, Dr. Rendle tells I'ne, agrees with Plukenet's figure and has a glabrous pedicel. Of the Hort. Cliff. specimens Dr. Rendle writes: "We have a specimen labeled glauca. The species' names, however, in Hort. Cliff, are rarely in Linnaeus' handwriting and glauca is in the usual hand, but there is also written in what may be Linnaeus' own hand Magnolia Catesby, with a reference to t. 39; this suggests that Linnaeus regarded the Hort. Cliff, plant as identical with Catesby's. The specimen is in flower and has a glabrous pedicel." Catesby's specimen in the British Museum has no flower, but his plate plainly shows that the pedicel is glabrous. The tj^ical Magnolia glauca is a small tree rarely more than 10 m. high, or often a shrub. It is the only form which grows from Massachusetts to south- eastern Virginia. Farther south it is rare and I have only seen specimens from Newbern, North Carohna, Darlington, Andrews, Bluffton, Georgetown, and Yemassee, South Carolina, and Meldrin, Georgia. Specimens collected in Florida in the vicinity of Eustis Lake, Lake County, by G. V. Nash (no. 575) and in the neighborhood of Orlando, Orange County, by C. H. Baker and T. G. Harbison have petioles, pedicels, and branchlets puberulous. For the form with the pubescent pedicels and branchlets I suggest the name Magnolia virginiana var. australis, n. var. — Differing from the tj^e in the silky white pubescence on the pedicels and branchlets. 232 BOTANICAL GAZETTE [march Leaves remaining on the branches until spring without change of color, elliptic to oval, oblong-obovate or rarely lanceolate, with puberulous, pubescent, or tomentose petioles and varying in width from 2.5 to 9.5 cm., trees with the broadest leaves being confined to western Louisiana and eastern Texas. This southern variety of M. virginiana is a tree often 20-30 m. high with a tall trunk occasionally i m. in diameter, covered with pale smooth bark and short small branches forming a narrow round-topped head, and branchlets more or less thickly covered during their first season with white silky pubescence, usually gradually disappearing in their second year; in southern Florida often much smaller, and on the Everglade Keys, where it is "very common, a shrubby tree up to 3 m. high" {E. A. Bessey). Swamps in the neighborhood of Wihnington, North Carolina, is the most northern station from which I have seen specimens of this tree; it is common in the coast region of South Carolina and Georgia and in all parts of Florida, and the only form of M. virginiana in the other Gulf states, where it occurs as far west as the valley of the Nueces River in Texas (San Augustine Coxmty) , but is much less common west of the Mississippi River than it is farther east. Although it crosses the Florida peninsula this Magnolia is most abundant in the coast region. It ranges inland, however, to Cuthbert, Randolph County, in western Georgia, to Tuskegee and Selma, Alabama, and to Tishomingo County in the extreme northeastern corner of Mississippi, and to Winn and Natchitoches parishes in western Louisiana. Magnolia acuminata var. ludoviciana, n. var. — Differing from the type in its broadly obovate, oval, or ovate leaves abruptly short-pointed at apex and rounded or cuneate at base, and in its much larger flowers. Leaves hoary tomentose below and slightly pubescent above when they unfold, becoming when the flowers open glabrous and yellow-green on the upper surface and pubescent on the lower surface with short pale hairs, 15-18 cm. long and 9-13 cm. wide; petioles puberulous, 2.5-4 cm. in length. Flowers 8-10 cm. long, the outer petals up to 4 cm. in width. A large tree. Rich woods, West FeUciana Parish, Louisiana, Dessert Plantation near Catalpa, Cocks and Sargent, April 12, 1916 (type); West Plantation near Catalpa, Cocks and Sargent, April 10, 1914; near St. Francisville, R. S. Cocks, May 15, 1915, and Catalpa, October 15, 1915. ipig] SARGENT— NORTH AMERICAN TREES 233 Acer saccharum Marsh.— The lower surface of the leaves of the northern sugar maple is usually green and glabrous, but it is sometimes glaucous or glaucescent and southward is slightly pubes- cent along the under side of the midribs and veins; and as the pale color of the lower surface of the leaves gives the trees a distinct appearance the varietal name adopted for them by some European dendrologists will probably be helpful. This form becomes Acer saccharum var. glaucum, nov. comb. — A. saccharinum var. glaucum Pax, Engler Bot. Jahrb. 7:242. 1886; Wesmael in Bull. Soc. Bot. Belg. 29:61. 1890. — A. palmifolium var. glaucum Schwerin, Gartenflora 42:455. 1893. — The leaves of this variety resemble those of the green-leaved variety in size and shape and are glabrous or in the southern states usually slightly pubescent on the under side of the midribs and veins. From the North, where it is much less common than the green-leaved form, I have seen specimens of this variety only from Isle-aux-Couvres in the St. Lawrence River, from Prince Edward Island, Nova Scotia, Lake St. John and St. Anne's, Quebec, northern V'ermont, Cooperstown, New York, western Pennsylvania, and Youngstown, Ohio; it is more common in southern Michigan and Indiana, and occurs in northeastern Iowa and central Tennessee. It is still more common in Missouri and northern Arkansas, and is the only form I have seen from South Carolina, Alabama, Mississippi, Louisiana, and southern Arkansas, where the sugar maple is not a common tree. Acer saccharum var. Rugelii Rehder, Cyclopedia Am. Hort. 1:13. 1900; Sargent, Man. /g. 5/5. 1905. — ^ . i^wge/ii Pax, Engler Bot. Jahrb. 7:243. 1886. — A. saccharinum subspec. Rugelii Wes- mael, Bull. Soc. Bot. Belg. 29:61. 1890. — In The Silva of North America this form of the sugar maple was confused, at least in part, with A . nigrum. As it is now understood, the leaves of this variety are usually broader than long and are cordate or rounded at base, 3-lobed with long acuminate lobes, usually entire or the lower lobes occasionally furnished near the base with a small rounded lobe ; the leaves are -3-nerved, thick, dark green above, green or glaucescent and glabrous on the lower surface, but on specimens collected by Palmer at Williamsville, Wayne County, Missouri (no. 6096), and from a large tree with short-lobed leaves at Campbell, Dunklin County, Missouri (C. S. Sargent, October 5, 1910), the lower surface of the leaves is thickly covered with loose pubescence. 234 BOTANICAL GAZETTE [march This variety appears to be rare and local and to occupy a comparatively restricted area. The tj^e station is at Dandridge on the Tennessee River, Jefferson County, Tennessee, and I have seen specimens from KnoxviDe, Tennessee, Eureka Springs, northwestern Arkansas, Williams ville, Campbell, and Allenton, Missouri, Lansing, Ingram County, Michigan, from Parry Sound, Georgian Bay (B. E. Fernow, 1908, a single tree), and Point Pelee in Lake Erie, Essex County, Ontario (C K. Dodge, 191 1). Acer saccharum var. sinuosum, n. var. — A. sinuosum Rehder, Sargent, Trees and Shrubs, 2:255. P^- ^95- 1913- — The distinctive character of A . sinuosum was found in a projection into the broad sinus at the base of the leaves formed by the nerves of the 2 upper lobes which form the base of the sinus. Since the species was described, large collections of this maple of the Edwards Plateau in western Texas show that this projection of the nerves is not a con- stant character and that A . sinuosum must be considered a small- leaved form of A . saccharum. This little Texan tree is known only on the banks and bluffs of Cibelo Creek, near Boerne, Kendall County, on the rocky banks of the upper Seco Creek, Bandera County, and at the base of a high Umestone bluff near Utopia, Uvalde County. Its isolation is remarkable and interesting, for none of the group of sugar maples grow nearer to the Edwards Plateau than A . grandiden- tatum Nuttall on the mountains in the extreme western part of Texas, A. floridanum Pax at Marshall, Harrison County, Texas, and A . leucoderme Small and A. saccharum var. glaucum Sargent in the Red River VaUey in southern Arkansas. Acer floridanum Pax. — A. saccharinum Elliott, Sk. 1:450 (at least in part). 182 1. The range of this species can be extended northward from River Junction, Florida, which is the type station, through the Piedmont region of Georgia and the Carolinas to the banks of the Roanoke River, near Weldon, HaUfax County, North Carolina, and to Dinwiddle County, southeastern Virginia (river banks and low wet woods near McKenney, W. W. Ashe). It is common in the neighborhood of Raleigh, Wake County, North CaroUna, where it has been largely planted, and is the common and prevailing street tree. Recent collections show that the variety of A . floridanum with villose- tomentose petioles and usually pubescent branchlets is not uncommon. It is the Var. FiLiPES Rehder, Sargent, Trees' and Shrubs, 2:255. ^9^3- — A. hrachypterum Wooton and Stanley, Contrib. U.S. Nat. Herb. 16: part 4, 146. 1913; 19:411. 1915. iqiq] SARGENT— north AMERICAN TREES 235 The type station of this variety is Columbus, Muscogee County, Georgia. It has also been collected in Georgia at Cuthbert, near Milledgeville, Mayfield, and on Shell Bluff on the Savannah River below Augusta, at River Junction, Florida, Calhoun Falls, South Carolina, in the streets of Raleigh, North Caro- lina, at Campbell, southeastern Missouri, and on the San Luis IMountains in southern New Mexico (.4. brachypterum) . This isolated New Mexican station far to the westward of the region usually occupied by A.floridanmn is remark- able, but in the shape of the leaves, in their pubescence and in that of the petioles, pedicels, and branchlets, and in the length of the wings of the fruit, it appears identical with some of the specimens of the var. filipes from the southeastern states. Acer rubrum Linn. — A. carolinianum Walter, Fl. Car. 251. 1888. — A. stenocarpiim Britton in Britton and Shafer, N. Am. Trees 64^, Jig. ^gS. 1908. — The leaves of the red maple are usually green and glabrous or pubescent below early in the season, gener- ally soon becoming glaucescent or glaucous below and glabrous or they are usually rather longer than broad, generally cordate or sometimes rounded at base, 3-5-lobed by acute sinuses with serrate lobes and slender glabrous petioles; in the autumn they turn scarlet on some trees and bright yellow on others. The flowers are red or yellowish green (var. pallidijiorum Pax). The fruit on different trees is red, yellow, or brownish. The branchlets and winter-buds are glabrous. The red maple grows on the borders of streams, in wet swamps and in upland forests, occasionally on dry hills, and is foimd from Newfoundland to the banks of the Miami River in the extreme southern part of Florida, and westward to western Wisconsin, jMinnesota, eastern Oklahoma, and to the neighborhood of Houston, Harris County, Texas. A. stenocarpum Britton is based on a single small stunted tree growing on a dry hill of flint rock at Allenton, St. Louis County, Missouri, on which the samaras of the fruit vary in width up to 6 mm. This maple has been growing in the Arboretum for several years. No other North American tree ranges through so many degrees of latitude as separate Newfoundland from southern Florida. In the shape of the leaves and in their pubescence, and in the size of the fruit, A. rubrum shows much variation. The extremes of these varieties have sometimes been considered species, but they are connected by so many intermediate forms that a better idea of the red maple can perhaps be obtained by treating it as a species with the following varieties: Acer rubrum var. tomentosum Pax, Engler Bot. Jahrb. 7: 182. 1886. — A. Drummondii Small, Fl. Southern U.S. 741 (insomuch as 236 BOTANICAL GAZETTE [march relates to Georgia and Florida, not Hooker and Arnott). 1903. — This variety, which was based on trees cultivated in Europe, is dis- tinguished by the close pale pubescence which covers the lower surface of the leaves during the season. The leaves are 5-lobed, cordate or rarely rounded at base, and the petioles are glabrous or slightly pubescent early in the season. The branchlets are usually glabrous and the winter-buds are pubescent. I have seen specimens of this form of the red maple from Biltmore, North Carolina (Herb. Bilt. no. ii6b), from the neighborhood of Augusta, Georgia, from the top of Flagstaff Mountain, Barclay, Alabama, Panther Burn, Sharkey County, Mississippi, Larissa, Cherokee County, Texas (B. F. Bush, May i, 1909, no. 5579), near Page, Leflore County, Oklahoma (G. W. Stevens, no. 2617), and swamps near Little Rock, Pulaski County, Arkansas. A specimen of this variety with pubescent branchlets and winter-buds, and slightly pubescent petioles, collected by J. K. Small at the Altamaha River Swamp, Liberty County, Georgia, in June 1895, and specimens collected by Mohr in April 1895 at Mount Vernon, Mobile County, Alabama, with broadly ovate, 3-5-lobed, slightly cordate leaves with pubescent petioles, fruit only 3 . 5 cm. long, and glabrous branchlets, serve to connect the variety tomentosum with Acer rubrum var. Drummondii Sarg. — The leaves of this tree are often broader than long, cordate at base, usually 5-lobed, with stouter midribs and veins than those of the other forms of A. ruhrum. Until nearly fully grown the leaves are covered on the upper surface with scattered pale hairs and are clothed below with thick snow white tomentum which is more or less persistent during the season; the petioles are stouter than those of the other forms of the red maple, and are covered during the season with thick white tomentum similar to that on the under surface of the leaves. This gradually disappears and the petioles often become nearly glabrous in the autumn. The fruit, which ripens in early spring before or with the unfolding leaves, varies from 5 to 6 cm. in length. This maple, which is a smaU tree usually not more than 10-12 m. high, inhabits deep river swamps often inundated through the year. It is dis- tributed from the valley of the Hastchatchee River, Forrest County, southern Mississippi, through Louisiana to the valley of the Neches River, Texas (Beaumont and Concord). It is not rare in southern and eastern Arkansas, southeastern Missouri, and occurs in northwestern Mississippi (Morehead, iqiq] SARGENT— north AMERICAN TREES 237 Sunflower County), and in southwestern Indiana (in a cypress swamp 18 miles west of Decker, Knox County, C. C. Deam.) In the broad 3-5-lobed leaves cordate at base this maple is very distinct from other forms of A. rubrum, but trees occasionally occur with 3-lobed leaves rounded at base. This form may be described as Acer rubrum var. Drummondii f. rotundata, n. f. — Differing from the variety in the 3-lobed leaves rounded at base. Specimens of this form have been collected in Louisiana at Chopin, Natchi- toches Parish, E. J. Palmer, May 6, 1915 (no. 7553 type), and at Glen Gordon, Covington, St. Tammany Parish, R. S. Cocks, March 28, 1911; in Texas near Beaumont, Jefferson County, C. S. Sargent, April 11, 191 5; and in Missouri at Poplar Bluff, Butler County, G. W. LeUerman, September 3, 1882. The fruit of this form has not been collected, but the tomentum of the leaves, petioles, and branchlets is that of the var. Drummondii. The shape of the leaves shows a transition into Acer rubrum var. tridens Wood. — A. harhatum Michaux, Fl. Bor. Am. 1:252 (at least insomuch as related to the leaves). 1803; Elliott, Sk. 1:451 (at least in part). 1821. — A. carolinianiim Britton in Britton and Shafer, N. Am.' Trees 648 (not Walter). 1 8 13. — The leaves of this maple are obovate, narrowed from above the middle to the rounded or rarely cuneate base, 3-lobed at apex, coarsely serrate usually only above the middle, often ovate or oblong-ovate by the suppression of the lateral lobes, green or glaucous and glabrous, pubescent or tomentose on the lower surface. The flowers and fruit are red or yellow. It has been found from ^lassachusetts to Florida, ^lissouri, and in eastern Texas to Harden and Cherokee counties, but is most abundant southward and sometimes, as in Richland Parish, northern Louisiana, it is the prevailing form. The extreme forms of this variety are distinct, but the 3-lobed leaves often occur on trees with leaves of the normal form of the red maple, and the leaves on vigorous shoots of trees of this variety are often 5-lobed. Walter's specimen of his A. carolinianum is preserved in the British Museum and is a typical .4. rubrum, not the variety tridens. The leaves of A. barbatum as described by Michaux and Elliott seem to be those of A. rubrum var. tridens. Their "peduncuH solito pilosi" might apply to A. florida- num filipes Rehder. Acer Negundo L. — The box elder or ash-leaved maple, which is one of the most widely distributed trees in the United States, has 238 BOTANICAL GAZETTE [march assumed slightly different forms in different parts of the country. In what is considered the typical species, which occurs in the region east of the Rocky Mountains, the united part of the samaras of the fruit is more or less constricted at the base into a short stipe, and on the more western forms this constriction usually does not occur. This constriction and its absence, together with the absence or presence of pubescence on the leaves and branchlets, have some- times been used to separate A. Negundo into several species, but the characters on which these species have been based are not particularly important, and it seems better to treat A . Negundo as a species with a number of varieties which often intergrade, for the characters on which they are based are not always constant. In what is considered the typical species the branchlets are green and glabrous; the leaves are usually 3-foliate but occasionally 5-7-foliate; the leaflets are ovate to elliptic or oblong-obovate, acuminate and often long-pointed, rounded or cuneate and often unsymmetrical at base, coarsely and irregularly serrate usually only above the middle, and occasionally slightly lobed, slightly pubescent above and more or less tomentose below when they unfold, and at maturity glabrous above, usually villose-pubescent along the under side of the midribs and veins, and often furnished with conspicuous tufts of axillary hairs, otherwise glabrous or slightly pubescent below. The typical form is distributed from western New England and central New York to Minnesota, Iowa, and Missouri, and southward to central Florida, northern Alabama, western Louisiana, and eastern Texas. I have not seen specimens of wild trees from eastern New England, eastern New York, New Jersey, or Delaware. The box elder is common along the St. Lawrence River near Montreal and in eastern Ontario, but these trees are believed to have been naturalized in recent years. As here considered it passes into the following varieties: Acer Negundo var. violaceum Kirchner in Kirchner and Petzold Arb. Mosc. 190. 1864. — Ruhac Nuttallii Nieuwland, Am. Middl. Nat. 2:137. 191 1. — Negundo Nuttallii Rydberg, Bull. Torr. Bot. Club 40:55. 1913.— This variety is distinguished by its rather stouter, pale or bluish violet branchlets covered with a glaucous bloom, rather larger buds, and usually 3-7-foliate leaves iqiq] SARGENT— north AMERICAN TREES 239 with slightly thicker, lanceolate to oblong-ovate or obovate, often entire or irregularly dentate, occasionally lobed leaflets, the terminal leaflet occasionally 3-lobed, glabrous above and usually slightly pubescent over the lower surface; the base of the fruit is usually but not always constricted. This variety is distributed from western Massachusetts, through New York to Ohio, northern Wisconsm, Minnesota, Iowa, South Dakota, Dufferin, Manitoba, and Nez Perce County, Idaho; it is common in northern Missouri and occurs near Noel, JNIcDonald County, in the extreme southwestern part of that state (£. J. Palmer, no. 5479). Acer Negundo var. texanum Pax, Engler Bot. Jahrb. 7:212. 1886. — A. calif ornicum var. texanum Pax, I.e. 11:75. 1890. — Rubac texana Small, Fl. Southern U.S. 743 (in psiri).— Negundo texanum Rydberg, Bull. Torr. Bot. Club 40:56. 1913.— This variety is best distinguished by the 3-foliate leaves with broader ovate to obovate, coarsely serrate leaflets cuneate or rounded at base and covered below through the season with loose pubescence. The branchlets are pale pubescent or tomentose during their first season and the body of the fruit is usually puberulous and shghtly or not at all constricted at the base. This variety occurs in western (Jackson County) and southwestern Mis- souri, northeastern Kansas, through Arkansas to western Oklahoma and to the valley of the San Antonio River, Texas. It appears to have been collected first by Lindheimer near New Braunfels, Texas, in 1843 (no. 360 in Herb. Gray). Eastward it passes into Acer Negundo var. texanum f. latifolium, n. i.—A. Negundo var. latifolium Pax, Engler Bot. Jahrb. 11 : 75. 1890.— Only differing from typical var. texanum in its glabrous branchlets and usually glabrous fruit often slightly constricted at the base. This form occurs in eastern Texas, southern Arkansas, Louisiana, in the valley of the Black River, eastern Mississippi, at Nashville, Tennessee, on the banks of the Catawba River near Marion, North CaroKna, in Virginia, and southern Ohio. Acer Negundo var. interior, n. var.— ^. interior Britton in Britton and Shafer, N. Am. Trees, 6ss,fig- 608. 1908.— ?^. Kingii Britton, I.e. 1908. — Rulac interior Nieuwland, Am. Mid. Nat. 2:139. igu.— Negundo interius Rydberg, Bull. Torr. Bot. Club 240 BOTANICAL GAZETTE [march 40:56. 1913. — Rulac texana Small, Fl. Southern U.S. 743 (in part). 1903. — This variety, the box elder of the Rocky Mountains, differs from the variety texanum only in its narrower and usually more acuminate and more irregularly serrate, often lobed leaflets usually covered below with a closer pubescence, and in the more pubescent or tomentose petioles and rachis. It ranges from Manitoba, Saskatchewan, and Alberta southward through Wyoming, Montana, Colorado, Utah, New Mexico, and Arizona. The oldest specimen of this tree which I have seen was collected by A. Fendler in New Mexico in 1847 O^o- 102 in Herb. Gray). More distinct is the variety of Arizona and southern New Mexico, which may be described as Acer Negundo var. arizonicum, n. var. — Leaves thin, 3- foliolate; petioles slender, glabrous, 4.5-7 cm. in length, often turning bright red late in summer; leaflets oblong-ovate to rhombic, acuminate and long-pointed at apex, rounded or cuneate at base, coarsely serrate, often slightly lobed near the middle, glabrous at maturity with the exception of conspicuous tufts of axillary hairs, 6-10 cm. long and 3-5 cm. wide; petiolules slender, glabrous, usually bright red, those of the terminal leaflet 2-2 . 5 cm. in length, the others only 6-7 mm. long. Flowers not seen. Racemes of fruit glabrous, 8-10 cm. in length; body of the fruit spreading, glabrous, not constricted at the base. A tree 7-8 m. high with light gray fissured bark and slender glabrous branchlets thickly covered with a glaucous bloom. Arizona. — Cave Creek Canyon, east slope Chiricahua Mountains, /. W. Tourney, July 1894; Oak Creek Canyon near Flagstaff, Coconino County, A. Rehder, July 14, 1914 (no. 34); Sycamore Canyon near Flagstaff, Percival Lowell, October 191 5 and 1916; Santa Catalina Mountains, /. G. Lemmon, May 1881 (no. 128 in Herb. Gray); Mount Kellogg, Santa Catalina Moun- tains, altitude 2700 m., A. Rehder, August 31, 1916 (no. 463 type). New Mexico. — Kelley's Ranch, 7 miles north of Ahna, Socorro County, A. Rehder, August 13, 1914 (nos. 279, 280; growing near no. 2S1, with densely pubescent branchlets and leaflets pubescent below, and so referable to var. interior); Glenwood, 7 miles south of Alma, Socorro County, A. Rehder, Au- gust 14, 1914 (no. 300). This is the most glabrous of the forms of .4. Negundo, and in its thin leaflets, bright red, petioles, and glabrous branchlets thickly covered with a glaucous bloom one of the most distinct of them all. iQig] SARGENT— NORTH AMERICAN TREES 241 Acer Negundo var. californicum Sargent, Garden and Forest, 4:148. 1891; Silva N. Am. 2:112. /?/. p/. 1891. — Negundo californicum Torrey and Gray, Fl. N. Am. 1:250, 684. 1838; Rydberg, Bull. Torr. Bot. Club 40:56. 1913. — Acer californicum Dietrich, Syn. 2:1283. 1840; Pax, Engler Bot. Jahrb. 7:213 (in part). 1836; 11:75. 1890. — Negundo aceroides Torrey, Pacific R.R. Rep. 4:74 (not Moench). 1857. — Negundo aceroides var. californicum Sargent, Garden and Forest 2:364. 1889. — Acer Negundo subsp. californicum Wesmael, Bull. Soc. Bot. Belg. 29:43. 1890. — Rulac californica Nieuwland, Am. Mid. Nat. 2: 139 (in part). 191 1. — Leaves trifoliate with tomentose or nearly glabrous petioles, rachis, and petiolules; leaflets oblong-ovate to rhombic, acuminate and long-pointed at apex, cuneate or unsjTnmetrically rounded at base, coarsely serrate above the middle or nearly entire, occasionally deeply lobed, glabrous on the upper surface except along the midribs and veins, thickly coated on the lower surface with matted pale hairs and furnished with large axillary tufts. Fruit on pubescent pedicels, puberulous or nearly glabrous, not constricted or rarely slightly constricted at base. A large tree with dark bark, hoary tomentose branchlets. and winter-buds. Valley of the lower Sacramento River southward to San Bernardino County, California. The California box elder was discovered by David Douglas in 1833, probably in the neighborhood of Monterey. Fraxinus AMERICANA var. subcoriacea, n. var. — Differing from the t^pe in its thicker, entire or shghtly serrate leaflets, silvery white on the lower surface. Individual trees of F. americana occur with thick, oblong-ovate, acuminate, entire or slightly serrate leaflets dark green and lustrous above, silvery white below, glabrous or slightly villose along the midribs, and 7.5-13 cm. long. These trees are so distinct in appearance and in their more rapid and vigorous growth that it seems desirable to give them a varietal des- ignation. What may be considered the t^-pe of the variety has been growing in the Arnold Arboretum since 1874, where it was raised from seed sent by W. C. Hampton from Mount Victory, Harding County, Ohio, as '^Fraxinus C." The trees of this variety have grown more rapidly and are handsomer than any of the other American ashes in the collection. In 1910 and 191 2 I collected the 242 BOTANICAL GAZETTE [march same variety at Campbell, Dunklin County, Missouri, where I saw- only a single individual. F. americana no. 4206, A. and E. G. Heller, Texarkana, Texas, September 1898, with entire and equally thick leaflets but not so pale below, is probably the same form. Castanea alnifolia var. floridana, n. var. — Differing from the type in the glabrous, lustrous under surface of the mature leaves and in its arborescent habit. In sandy soil with Quercus myrtifolia Willd. on the shores of St. Andrews Bay near Panama City, Bay County, Florida, T. G. Harbison, May 28, 1917 (no. 10, type), December 10, 1918 (nos. 13, 14); Dover, Screven County, Georgia, T. G. Harbison, May 13, 1913. A tree occasionally 13-14 m. tall, or sometimes shrubby. Unfortunately I have not seen the fruit of this tree, but in the shape of the leaves and in their serration, in the inflorescence, and in the glabrous branchlets it is not distinguishable from C. alnifolia Nutt. The leaves when they first unfold are hoary tomentose below, and the tomentum is sometimes persistent during the season on the upper leaves of vigorous shoots. On a specimen of what appears to be the same form collected by Harbison near Jacksonville, Florida, the branchlets are slightly puberulous, and there are a few hairs on the under side of the midribs of the otherwise glabrous leaves. The leaves of a shrub with pilose branchlets collected by Harbison on the coast near Wilmington, North Carolina, are broadly obovate and green, lustrous and puberulous on the under surface, and quite different from, the leaves of the typical form of C. alnifolia, which are narrow-obovate to oblong-elliptic and thickly covered below with pale tomentum. Arnold Arboretum Jamaica Plain, Mass. POSSIBLE CORRELATIONS CONCERNING POSITION OF SEEDS IN THE POD Byron D. Halsted' This study was made with the Henderson Lima bean on a block of ground one-fortieth of an acre. Nine rows of lo hills each were planted 5 seeds to the hill in the manner shown in table I. TABLE I Row Position of seeds in pod Percentage of viability I 2-seeded base 2-seeded tip 3-seeded base 3-seeded middle 3-seeded tip 4-seeded base 4-seeded first middle 4-seeded second middle 4-seeded tip '^2 2 70 3 4 60 76 c 56 6 64 7 . 60 8 72 9 74 Total 584 Table I shows that seeds from the middle of 3-seeded pods had the highest viabihty, and that of those from the base of 2-seeded pods nearly one-half failed to produce plants. The average viability of the 3 rows planted with basal seeds was 58 . 6 per cent, the next lowest 66.6 per cent in the rows planted with tip seeds, and the highest viabihty 69.3 per cent, obtained in the rows planted with seeds from the middle of 3-seeded and 4-seeded pods. Table II shows that the pods are chiefly of the 3-ovuled type, and that the others are somewhat equally divided between the 2-ovuled and 4-ovuled pods. The yield of pods from seeds of 2-ovuled pods is 15.61 per cent less than the average yield from seeds of 3-seeded and 4-seeded pods. Pods from plants grown from seeds borne in the basal position numbered 1463, from seeds from middle position 1735, and in the ' This paper was received from the late Professor Halsted in January 19 18. A brief biographical sketch appears in the Botanical Gazette of February 1919. 243] [Botanical Gazette, vol. 67 244 BOTANICAL GAZETTE [MARCH tip position 1403. This is 23.66 per cent more pods from the 3 rows planted with seeds from the middle of the pod than from the same area planted with seeds from the tip position, and TABLE II Number and types of pods for each row, both harvests combined Row Position of seeds in pod 2-ovuled 3-ovuled 4-ovuled Total Average per pod type T 2-seeded base 2-seeded tip 3-seeded base 3-seeded middle 3-seeded tip 4-seeded base 4-seeded first middle 4-seeded second middle 4-seeded tip Total 44 37 55 82 30 44 16 47 60 343 449 487 540 288 401 517 388 456 17 21 36 34 17 36 46 65 45 404 507 578 656 335 481 579 500 561 2 3 4 s 6 455 -S 5230 7-: 530-2 9 415 3869 317 4601 18.58 per cent more than the corresponding rows planted with basal seeds. The percentage of each pod type in the crop for each pod type planted is given in table III. TABLE III Type planted 2-ovuled 3-ovuled 4-ovuled 2-seeded pods 8.88 9.09 9.48 86.95 84.87 81. II 417 3-seeded pods 5.54 4-seeded pods 9.41 Total 9-30 83.87 7.33 Table III shows but Kttle variation in the number of 2-ovuled pods, but in 3-ovuled pods there is a decided decrease in the per- centage, going from 2-seeded to 4-seeded, and, of course, a corre- sponding increase in the 4-ovuled pods, and where the percentages are low the differences are great, over 125.66 per cent between the crop from 2-seeded pods and that from 4-seeded pods. In other words, there is seen to be a tendency for seeds from 4-seeded pods to reproduce the mother type of pod. The number of pods for each iQig] HALSTED— POSITION OF SEEDS 245 harvest and the percentages for the 3 types of pods are indicated in table IV. TABLE IV Number OF PODS Percentage of each type of pod ■ 2-ovuIed 3-ovuled 4-ovuled First harvest Second harvest 2672 1829 7. II 11.86 84.73 82.87 8.16 527 Total 4501 930 83 -37 7-33 Table IV shows that the first harvest comprised nearly three- fifths of the total crop. It is further shown that the first harvest has its 2-ovuIed pods below the average and its 3-ovuled and 4-ovuled pods above the average, while in the second harvest the reverse is seen. In other words, as the season advanced the average number of seeds in the pods increased. TABLE V Number and average weight of seeds Pods without Pods with Both averaged 4BORTS ABORTS Row Position of seeds IN POD N»xm- Average Num- Average Num- Average ber weight ber weight ber weight I 2-seeded base 479 0 404 gr. 433 0.434 gr. 456 0.419 gr. 2 2-seeded tip 683 0.376 52Q 0.404 606 0.390 3 3-seeded base 881 0.376 539 O.411 710 0.394 4 3-seeded middle Q12 0.396 662 0.424 787 0.410 5 3-seeded tip 413 0.390 371 0.419 392 0.405 6 4-seeded base 617 0.392 512 0.420 565 0.406 7 4-seeded first middle 799 0.391 630 0.422 715 0.407 8 4-seeded second mid. 878 0.385 783 0.415 831 0.400 9 4-seeded tip 803 0.360 528 0.387 666 0.378 Table V shows that the largest number of seeds (4624) was taken from the rows planted with seeds from the middle of the pods, while the number of seeds planted from the basal and tip seeds were 3461 and 3377 respectively. The average weight of all seeds from each row approximates the general average, the only striking deviation being in the lightness of the seeds grown from seeds produced in the tip position of 4-seeded pods. 246 BOTANICAL GAZETTE [march Table VI shows under pod average (seed-weight for the 3 types of pods) that the heaviest seeds are produced in 3-seeded pods and the hghtest are found in 4-seeded pods. It is also shown that the seeds in the basal position are lighter than elsewhere in the t>T)e of pod considered, and that in the combined harvests the seed weights TABLE VI Relation of position of harvested seeds in pod to weight: i. seeds from PODS without aborts and average for all 9 ROWS Both harvests. Pod averages. . First harvest . . Pod average . . . Second harvest Pod average . . . -base 2-tip 0.338 0.366 0.352 0.314 0.345 0.329 0.379 0.402 0.390 3 -base 3 -middle 3 -tip - 0.362 0.392 0.413 0.391 0.344 0.37s 0.388 0.369 0.399 0.442 0.459 0.434 4-base 4-first 4-second 4-tip 0.305 0.341 0.341 0.345 0.333 0.268 0.319 0.313 0.326 0.307 0.361 0.374 0.384 0.375 0 ■374 All 0.38s 0.360 0.426 make a continuous rising series from the base to the tip for all types of pods. There is an exception in the records for separate harvests, but it is among 4-seeded pods, a type not largely repre- sented, and small deviations are here to be expected. It is noted that the first harvest yields seeds of lighter weight than does the second harvest, and with a single exception (4-seeded base) the difference applies to each position in the pod. TABLE VII II. SEEDS FROM PODS WITH ABORTS AND AVERAGE FOR ALL g ROWS All Both harvests. Pod averages . . First harvest . . Second harvest 2-base 2-tip 3-base 3-middle 3-tip 4-base 4-first 4-second 4-tip 0.404 0. 0.324 0.442 0.448 440 0.388 0.473 0.377 0.344 0.413 0.403 0.439 0.418 0.368 0.401 0.438 0.476 0.329 0.271 0.423 0.361 0. 0.338 0.433 0.37S 373 0.3S4 0.403 0.388 0.374 0.428 0.414 0.377 0.453 Table VII shows from the pod averages that the weight of the seeds decreases from 2-ovuled pods to 4-ovuled pods, and that the average weight of the seeds in 3-ovuled pods is very close to the average for all seeds. The averages for both harvests show that the lightest seeds are produced in the basal position, and that there is in all types of pods a uniform increase from the weights in the basal position to the weights in the tip position. In the first harvest the seeds in each position in each type of pod are lighter than the seeds for the same position in the second harvest. iqiq] HALSTED— POSITION OF SEEDS 247 A comparison of the weights of seeds for each pod position in the combined harvests of pods without aborts (table III) with those with aborts (table IV) shows that the former are uniformly lighter, as seen in the following statement : Pods without aborts. Pods with aborts First harvest 0.360 gr. 0-377 Second harvest 0.426 gr. 0453 Both harvests 0-385 gr- 0.414 The pod averages for seed- weights (tables III, IV) show that the seeds of 3-ovuled pods without aborts are heavier than are the seeds (average weights) in the other two types of pods, while among pods with aborts the average seed weight is greatest in 2-ovuled pods. This is shown in the following figures, in which both harvests are combined : 2-ovuled Pods without aborts. Pods with aborts 0.352 gr. 0.440 3-ovuled 0.391 gr. 0.418 4-ovuled 0.333 gr- 0.373 From these averages it is seen that the differences are con- siderable, and it may be assumed that it is due to local environment within the pod. For example, in a 2-ovuled pod the average weight of the seeds is 25 per cent more when one ovule aborts, 6 . 93 per cent in 3-ovuled pods, and 12.01 per cent in 4-o\'nled pods. It is noted that the differences are greatest in the extreme or, as they may be called, exceptional pods. The influence upon the remaining seeds when two or more ovules abort in the larger types of pods cannot be determined from the records. Percentages of gains in weight of seeds associated with aborts over the weight of seeds from pods without aborts for each position in the pod are as follows: Pod averages 2-base 2 -tip 22.40 .96 3-base 3-middle 3-tip 4-base 4-first 4-second 4-tip 12.46 IQ.2S 20 4.12 2.8i 6.79 4-57 7.87 5-87 12.17 9-59 AU 7-79 248 BOTANICAL GAZETTE [march It is seen that the gain of weight of seeds for pods with aborts over corresponding seeds in pods without aborts is greatest in 2-ovuled pods and least in 3-ovuled pods. It is further shown that the weight is most augmented in the tip position in all types of pods, and least in the position next above the base in 3-ovuled and 4-ovuled pods. TABLE VIII Percentage of abortiveness for each position of seeds in pod Position of seeds in pod First harvest Second harvest Both 2-seeded base 17-05 17-30 14-30 13.10 17.80 20.0 16.2 16.0 17.6 26.7 24.9 23-0 234 24.8 23.0 22.8 22.8 27.1 21.81 2-seeded tip 21 . 10 3-seeded base 18.65 18.25 21.30 21.50 19 -SO 19.40 22.3s 3-seeded middle 3-seeded tip 4-seeded base 4-seeded first middle 4-seeded second middle 4.-seeded tin Table VIII shows that the abortiveness for each of the 9 rows is less in the first than in the second harvest. When the two harvests are combined, the range is from 18.25 to 22,35 P^^ cent. It is further deduced that the average for the middle seeds (19.05 per cent) is the lowest, while the tip seeds yielded the highest (21.58 per cent) average abortiveness. From the standpoint of percentage of good seeds per pod, the middle seeds are the best for good crop productions. TABLE IX Percentage of abortiveness for each pod position in whole crop 2-base 2 -tip 3-base 3 -middle 3-tip 4-base j4-first 4-second 4-tip All Both harvests .... Pod averages First harvest Pod averages Second harvest . . . Pod averages 49. SS 11.88 30.74 31-77 8.83 20.31 62.99 14-17 •?8 i^S 40.11 47-34 "si-28' 12.95 19.02 7-96 15-77 13.74 23.17 13.2s 52.63 13.93 S.26 19.04 12.84 s-So 19.26 16.99 476 t8 .57 4-34 19-75 2.29 55. 50 3-21 16.83 4.48 46.66 6.66 24.12 Table IX shows that when the two harvests are combined the abortiveness is chiefly in the lower portion of the pod and is very large in the basal position. In all types of pods there is a regular decrease in the number of aborts from the base to the tip, with the iqiq] HALSTED— POSITION OF SEEDS 249 greatest range in 4-ovuled pods. The averages are more nearly alike than averages for position in the pod and are not strictly comparable. In the first harvest the aborts are nearly two-thirds the number of those in the second harvest, but their distribution among the 9 positions in the pods does not follow fully the rule given for the whole crop. This is shown in the pod averages, where in the first harvest the smallest average is with the 3-ovuled pods, while in the second crop the smallest average is with the 4-ovuled pods. A greater abortiveness in the second crop may be ascribed to the advanced age of the plants or to the lack of proper insect visitation, but this latter circumstance may not be significant, as Lima beans are understood to be self-fertilized. It is possible, of course, that the cause may be related to the atmospheric conditions prevailing at the time the o\Tjles were ready to set, and this suggests the importance of repeating the present test through a series of years. TABLE X Average weight of seeds and percentage of ABORTIVENESS FOR EACH POD POSITION 2 -base 2-tip 3-base 3-middle 3-tip 4-base 4-first 4-second 4-tip All Seed weight . . Abortive- 0.338 0.366 0.362 0.392 0.412 0.303 0.341 0.341 0-345 0.38s ness Seed weight . . Abortive- 49. SS 7.0 11.85 30 40. 11 4.0 12.95 2.0 3-25 1 .0 52.63 8.0 13 -93 6.0 5. 26 6.0 4-34 50 19-75 ness 2.0 6.0 30 5-0 9.0 1 .0 4.0 7.0 8.0 The ranking figures make it easier to compare the relationship of the average weights and abortiveness among their respective units. Table X shows at once that the position yielding the heaviest seeds has the lowest percentage of abortiveness, and, contrariwise, the position with the lightest seeds has the largest number of aborts. It is evident that the type of pod has much influence upon the weight of the seed, and it is only proper that the comparison here made should be within the pod. With this consideration in mind it is seen that the order from base to tip with all the pods is reversed, that is, the base bears the lighter (or lightest) seeds and with more (or most) abortiveness. 250 BOTANICAL GAZETTE [march On account of the influence of pod type upon seed-weight and abortiveness it follows that the relative seed-weight does not neces- sarily determine the amount of abortiveness. For example, 2-ovuled pods bear seeds at their tips that are of the same weight as those formed at the base of 3-seeded pods, but the relative abortiveness is 1 1 : 40. Summary 1 . The greatest viability in Henderson Lima beans is associated with the seeds that are borne in the middle of the pods. 2. Three-seeded pods make up more than four-fifths of the crop; 3-seeded and 4-seeded pods are more numerous in the second of the two field harvests of ripe pods. 3. Seeds from the middle of the pod produce a much larger number of pods than do seeds from the base or tip. 4. The heaviest seeds are produced in 3-seeded pods, and the lightest in 4-seeded pods. 5. The seed weights make a continuous rising series from the base to the tip for all types of pods. 6. The first harvest yields lighter seeds than does the second harvest in each pod position. 7. The seeds associated with aborts are heavier than are those in full pods in each type of pod, and each position in the pod. 8. The abortiveness is less in the first than in the second harvest, and is least in the rows grown from seeds from the middle of the pods. 9. Abortiveness is chiefly in the basal position and decreases regularly from the base to the tip of the pod. 10. The position of the pod that yields the greatest weight of seed is associated with the lowest percentage of abortiveness. EMBRYO AND SEEDLING OF DIOON SPINULOSUM Sister Helen Angela Dorety (with plates X, Xl) Dioon spinulosum Dyer, imperfectly and incompletely described by EiCHLER' in 1883, and by Dyer^* in 1885, but fully and carefully by Chamberlain^ in 1909, is, like the other 2 species of Dioon, endemic in Mexico. The embryos and seedlings which furnish the material for this study were grown from o\'ules collected by Dr. Chamberlain in the mountains about Tierra Blanca and Tux- tepec during his several trips to the Dioon country. The tree is described by him as a magnificent ornamental cycad 30-40 ft. high. Unlike D. edule, it grows rapidly, and in 2 years makes a handsome greenhouse plant with a crown of large, fernlike leaves. The unique appearance of the plant and the great size of its ovulate strobilus and ovules led to the expectation of great pecu- liarities in its vascular anatomy and cotyledonary arrangements. The investigation has verified these expectations only in part. The study of the vascular anatomy of the embryo and seedling of D. spinulosum merely serves to emphasize the general harmony which prevails among the cycads in this respect. Embryo The seed, like those of all cycads, is filled with a massive endo- sperm stored with starch. Upon this tissue the proembryo and embryo proper are nourished, apparently without any resting period. When the embryo has attained a length equal to that of the seed itself, pressure is exerted upon the stony coat, and the thin region near the micropyle is broken. Fig. i shows a seed with a young embryo borne on the twisted suspensor; fig. 2 represents ' EiCHLER, a. W., Ein neues Z?/o(?«. Gartenflora 2:411. 1883. " Dyer, Sir W. T. Thistletox, Biologia Central! Americana, Botany 3 : 190. 1885. 3 Chamberlain, C. J., Dioon spimdosum. Box. Gaz. 48:401-413. 1909. 251] , [Botanical Gazette, voL 67 252 BOTANICAL GAZETTE [march a later stage in which the embryo has attained its full length, has broken the seed coat, and pushed out the dried remains of the suspensor and archegonial wall. The embryo at this stage consists of cotyledons, plumule, and a basal part which in its upper portion is hypocotyl, and in its lower portion root sheath; for the root is endogenous, and is not pre- ceded by any structure which might be looked upon as a ''radicle" or "caulicle." The hypocotyl is extremely short, and the distinc- tion between it and the root sheath cannot be determined super- ficially, but only by study of the internal structure. The number of cotyledons varies from 2 to 4, and all stages of their union are represented (figs. 20-22, 24). The vascular cylinder of the hypocotyl is a protostele. It has 4 easily recognized protoxylem groups, in no way differing from the h3Apocotyl cylinders of Ceratozamia'^ and Microcycas.^ The coty- ledons are multifascicular, like those of Ceralozamia and Microcycas, and unlike those of Zamia and Dioon ediile. In the embryos which are dicotyledonous, the manner in which the cotyledonary traces are supplied from the hypocotyl cylinder is exactly the same as that described by Thiessen^ for Dioon edule, and which Coulter and Chamberlain^ have shown to be characteristic of the cycads. When the embryo has 4 cotyledons, each cotyledon is on a side of the quadrangular node, and receives a secondary bundle from each of the adjacent angles. Twelve out of 100 embryos had 4 coty- ledons, and in each case the vascular strands arose in this manner. In the tricotyledonous embryos (there were 4 of them in 100), one of the 3 cotyledons was supplied in the dicotyledonary manner, and the other two after the manner of the embryos with 4 cotyledons; and yet, at a level just above the tip of the plumule, the number of vascular strands was about equal in the 3 cotyledons. The strands ■• DoRETY, Helen A., The seedling of Ceralozamia. Box. Gaz. 46:203-220. pis. 12-16. 1908. 5 , Vascular anatomy of the seedling of Microcycas calocoma. Bot. Gaz. 47:139-147- pis. 5,6. 1909. * Thiessen, Reinhardt, The vascular anatomy of the seedling of Dioon edule. Bot. Gaz. 46:357-38o.*'/»/5. 23-29. 1908. 7 Coulter, J. M.,^and Chamberlain, C. J., Morphology of gymnosperms. Chicago. 1910. iQiQJ DORETY—DIOON 253 are endarch where they separate from the hypocotyl cylinder (fig. 28) ; they become mesarch just after they enter the base of the cotyledon (fig. 27) and maintain that character throughout the greater length of the organ; near its tip, however, they are exarch (fig. 26). In all cases the orientation is collateral ectophloic, although at levels where branching is effected there is always an apparent concentric arrangement where the 2 xylem masses are still in contact, and the phloem masses are swung to right and left of them. Transfusion tissue accompanies the metaxylem. Muci- lage ducts and tannin cells are abundant. The root meristem is plainly visible below the hypocotyl plate in embryos of the age shown in fig. 2, but no differentiation has taken place, and of course there is no vascular tissue. The 4 poles of the root are later developed in connection with the 4 protoxylem groups of the hypocotyl vascular plate. The plumule consists of 3 or 4 abortive scales inclosing the rudi- ments of the first and second true leaves, sometimes of a third leaf, and the stem tip. There is no means of distinguishing between the stem tip rudiments and those of a new leaf, because the leaf meri- stem grows much more rapidly, and soon overtops the stem tip (fig. 14). The vascular system supplying all these bracts and leaves is complicated by the well known habit of girdling, the details of which have fascinated and baffled many investigators. Although D. spinulosum differs in no way from the other cycads in this respect, its greater size makes a naked eye drawing possible and thus furnishes a means for solution and demonstration. Figs. 14 and 15 were drawn from macerated stems. Each node of the stem is, like the Jiodes on a first year stem of foxglove, telescoped within the older one instead of growing above it. The internodes are not elongated because the primary meristem of the stem tip is held in check by the more rapidly growing secondary meristem for each developing leaf. Since each leaf is supplied with strands from cauline bundles in different parts of the stem, those strands which come to it from the opposite side of the stem describe almost a semicircle to reach the leaf; those which arise on the same side as the leaf pass directly into it; and small arcs are described by those strands which arise in intermediate positions. In fig. 14 254 BOTANICAL GAZETTE [march the traces for the older leaf (i') are supplied from the cauline strands numbered 2, 7, 14, 17, and 20. The strands from 2 and 20 will girdle the stem cylinder, those from 7 and 17 will make a p3,rtial girdle, and that from 14 will enter the leaf petiole directly. Fig. 15 is an attempt at demonstrating the same condition lower down in the stem. For the purpose of making the condition clearer the vertical magnification was made greater than the horizontal one. Many angiosperms which have both "radical" and cauline leaves give illustration of this same condition. In the second year stems of such plants each node is located at some distance above the older one, and the leaf traces arising on the side of the stem opposite the leaf describe a spiral or obhque arc before entering the leaf. In the part of the stem from which the so-called "radical" leaves spring, however, the vascular strands destined for these leaves describe a horizontal arc similar to the leaf trace girdle in all parts of the cycad stem. A careful but vain search was made for cortical cambium, vestigial traces of the primitive polystele of the Cycadofilicales. Germination Germination is hypogean, like that of all other members of the order thus far described. When the embryo has grown to the full length of the seed, the thin portion of the stony coat surrounding the micropyle yields to the pressure exerted upon it by the base of the axis. This base, scarcely more than a root sheath, emerges, pushing before it the brown and withered remains of the suspensor and archegonial wall. The cotyledonary base elongates and bends downward, and the root tip' emerges from its sheath (fig. 3). If the embryo is a monocotyledonous one, that is, if its whole coty- ledonary apparatus is a single sheath surrounding the plumule, this sheath is split by the radial growth of the axis; if there are two or more distinct cotyledons, their petioles are separated by the same cause (fig. 4). The plumule then emerges from the seed and be- comes erect. Of course, when the seeds germinate while in a vertical position, the root and the plumule develop in the same axis (figs. 5, 6). In 3 of the seeds double embryos developed (fig- 29). iqiq] DORETY—DIOON 255 Seedling The primary root persists indefinitely as a tap root. Large quantities of starch are transferred to it through the cotyledons, and it becomes large and swollen. The small lateral roots arise in whorls, usually of 4, and become a matted mass of fibers. Almost all the greenhouse grown seedlings have their roots hyper trophied, and the root tips have the characteristic tubercles described by LiFE^ in Cycas. The first leaves are yellowish scales, although thick and fleshy. They bear stipules like those of the foliage leaves, and some of them manifest the typical circinate venation (figs. 10, 11), The true leaves have been described by Chamberlain in the work pre- viously cited. The first true leaves of several of my seedlings had as few as 6 pairs of leaflets (fig. 9), but most of them had 12 or more. There is great variation in the size of leaves of plants grown in the same conditions and from the same sized seeds. One plant bore small leaves with leaflets 2x0.3 cm., while another just beside it bore on its first leaf 12 pairs of leaflets, each one of them 6X1.1 cm. Imitation of the moist air conditions under which the first leaves of Ceratozamia became foliage leaves was unsuccessful with Dioon spinulosum. The stele of the primary root is tetrarch, changing to diarch in its later formed portions and in the lateral branches. The relative size of the vascular cylinder to the whole root in various levels is shown by figs. 16-19. I^ the hypocotyl the diameter of the vascu- lar cylinder is only about one-seventh that of the whole axis. The manner in which the radial position is achieved in the root is illus- trated in figs. 18 and 19. The leaf traces are always endarch and collateral, and their arrangement in the petiole as shown in cross-section presents the well known omega-shape. Branching and anastomosis are frequent throughout the petiole, but there is in general a diminution of traces toward the top of the leaf. The plant is a much more rapid grower than Dioon edule, and is far more graceful. Under favorable conditions in the greenhouse, ' Life, A. C, The tuber-like rootlets of Cycas revoluia. BoT. Gaz. 31:265-271. 1901. 256 BOTANICAL GAZETTE [march plants made 3 and 4 leaves, each i m. long and 14 cm. wide, in less than a year after emerging from seed. Summary 1. The cotyledons of Dioon spinulosum vary in number from 2 to 4, and they are often lobed and divided so as to appear greater in number. In rare cases the cotyledonary sheath is undivided except near the tip. 2. They are multifascicular, like those of Ceratozamia and Microcycas, rather than like those of Zamia and Cycas, which have but few strands. 3. The arrangement and orientation of the vascular strands of cotyledons, hypocotyl, stem, leaves, and root do not differ in any marked degree from the general cycad arrangement. 4. The stem is large enough to demonstrate the cause of the girdling habit and to bring it into alignment with certain angio- sperms of the same habit. 5. There is no extrafascicular cambium or any other vestige of polystyle. Grateful acknowledgment is due to Dr. Charles J. Chamber- lain for the generous supply of material from which this study was made. College of St. Elizabeth Convent, New Jersey EXPLANATION OF PLATES X, XI The drawings 1-13 and 29 were made with the unaided eye and are reduced to one-half the natural size; 14-19 are diagrammatic; the remainder were made with the aid of the Abbe camera lucida. The following abbrevia- tions have been used: c, cotyledon; /, leaf; s, suspensor; sc, scale leaf; si, stipule; vc, vascular cylinder; x, xylem; ph, phloem; px, protoxylem; mx, metaxylem; ep, inner epidermis of the cotyledons. Fig. I. — Ovule with young embryo and coiled suspensor. Fig. 2.— Seed with mature embryo, pushing out suspensor and archegonial wall. Fig. 3. — Beginning of germination. Fig. 4. — Separation of cotyledonary petioles. Fig. 5. — Seedling germinated in vertical position. BOTANICAL GAZETTE, LXVII PLATE X Siste-y Helen- Angzla dtU DORETY on DIOON BOTANICAL GAZETTE, LXVII / PLATE XI Shttr Helen Angela del. DORETY on DIOON igig] DORETY—DIOON 257 Fig. 6. — Germination of lower haK of seed. Fig. 7. — First scale leaves, side view. Fig. 8. — Cross-section of cotyledons and plumule, showing approximately opposite arrangement of leaves. Fig. 9. — Seedling 2 months after germination. Figs. 10-12. — First, second, and third scales respectively. Fig. 13. — Size of vascular cylinder to relative root. Fig. 14. — Diagram showing cause of girdling of leaf traces and apparent absence of stem tip. Fig. 15. — GirdUng of leaf traces lower down in stem ; vertical magnification greater than horizontal. Fig. 16. — Diagram of hypocotyl cylinder or "plate." Fig. 17. — Section i mm. below fig. 16. Fig. 18. — Transition to root arrangement; separation of phloem masses of 4 poles. Fig. 19. — Vascular cylinder of root. Figs. 20-22, 24. — Cotyledonary variations. Figs. 23, 25. — Details of arrangement of epidermis over inner sides of cotyledons and cotyledonary sheath. Fig. 26. — Exarch bundle from tip of cotyledon. Fig. 27. — Mesarch bundle from central part of cotyledon. Fig. 28. — Endarch bundle from base of cotyledon. DEVELOPMENT OF STROPHARIA EPIMYCES W. B. McDoUGALL (with ten figures) Stropharia epimyces (Peck) Atk. was first described by Peck (9) in 1884 as Panaeolus epimyces. It was redescribed by Atkinson (i) in 1902 and again in 1907 (2), and placed in the genus Stro- pharia because of the purpHsh tinge of the spores and the presence of an annulus. It is considered rare, but it has occurred in several localities north of Urbana during each of the 4 summers I have spent in Illinois, and was particularly abundant during the seasons of 1915 and 1916. As is well known, this plant always occurs as a parasite on another mushroom. The identification of the host plant was first published by Atkinson (i) as Coprinus atramentarius. Later a second host, C. comatus, was added by Sherman (id). All specimens collected at Urbana have been on C. comatus. Several photographs of Stropharia epimyces and its host were published by McDouGALL (8), and excellent photographs were also published by Atkinson (2). Material for the developmental study of this plant was obtained within the city park of Urbana in Spetember 1915. It was imbed- ded, sectioned, and stained with fuchsin. Development The smallest carpophore sectioned measured 0.9 mm. by 1.2 mm. (fig. i). In this the pileus and stem fundaments cannot be said to be differentiated, but the primordium of the hymeno- phore already appears as a patch of heavily stained hyphae on each side of the median longitudinal section. Aside from this hymenophore primordium there is no differentiation of the carpo- phore at this stage, except a layer of coarse and rather loose hyphae on the periphery, representing the universal veil or blematogen. The size of the carpophore is not always an index of its degree of Botanical Gazette, vol. 67] [258 I9I9] MCDOUGALL—STROPHARI A 259 development, since the carpophore shown in fig. 2 is considerably- larger, although little if any further developed than that shown in fig. I ; but in any case the first internal differentiation is the appear- ance of the hymenophore primordium. The rapid growth of the elements of the hymenophore as the carpophore enlarges, together with the cessation of growth, or at least very slow growth of the ground tissue below the hymeno- phore primordium, soon cause the formation of an annular gill cavity (figs. 3, 4). The presence of the annular gill cavity makes it easy to see in longitudinal section which parts of the carpophore "'MS^^W Fig. I Fig. 2 Figs, i, 2. — Stropharia epimyces: fig. i, small carpophore showing primordium of hymenophore as only dififerentiation ; fig. 2, larger specimen at same stage of devel- opment. in general belong to the pileus fundament and which to the stem fundament, but it is not until still later that these are clearly dis- tinguished. The gill cavity enlarges rapidly (fig. 5), but it does not become very large before the formation of the lamellae by the downward growth of hyphae from the hymenophore begins (figs. 6, 7). By this time also, if not earHer, the universal veil has disappeared, and the mature carpophore is without any trace of it. Atkinson (4) has explained in detail the development of the lamellae in 26o BOTANICAL GAZETTE [march Agaricus Rodmani. The development in our plant is similar and therefore the details need not be repeated here. The same author Fig. 3 Fig. 4 Figs. 3, 4. — Stropharia epimyces: stages in development of annular gill cavity. Fig. s Fig. 6 Figs. 5, 6. — Stropharia epimyces: fig. 5, late stage in development of annular gill cavity; fig. 6, early stage in development of lamellae. has explained the rather deceptive "stalls" or "pigeon holes" formed by the lamellae at the stem end in longitudinal sections of carpophores in which the lamellae are attached to the stem. iQig] MCDOUGALL—STROPHARIA 261 Such "stalls" are shown in figs. 8 and 9, and prove that the lamellae are attached to the stem at this stage of development. Fig. 7 Fig. 8 Figs. 7, 8. — Stropharia epimyces: fig. 7, late stage in development of lamellae; fig. 8, similar stage cut to show "stalls," indicating that gills are attached to stem. Fig. 9 Fig. 10 Figs. 9, 10. — Stropharia epimyces: fig. 9, portion of expanding carpophore show- ing "stalls," indicating that gills are attached to stem; fig. 10, portion of expanding carpophore showing the rather slight inner veil. The inner veil which develops from the ground tissue below the hymenophore primordium remains intact until the lamellae are 262 BOTANICAL GAZETTE ' [march well developed, but it usually ruptures before the stem is much elongated (fig. lo) and forms a rather slight ring on the stem. The subsequent elongation of the stem results in the position of the ring near the stem base. Discussion As Atkinson (5) has said, the few species of Agaricaceae with endogenous origin of the hymenophore, whose development has been studied, fall into 3 groups, based on the order of differentia- tion of the pileus, stem, and hymenophore fundaments in the carpophore. In the first of these groups the pileus primordium is differentiated first. This group is represented by Hypholoma sub- lateritium, H. fasciculare, Amanita rubescens, and Amanitopsis vaginata. A second group includes those species in which the stem primordium is first differentiated, this being followed by the differ- entiation of the pileus primordium and later of the hymenophore primordium. This group is represented by Lepiota cristata, L. semi- nuda, and Rozites gongylophora. In the third group the hymeno- phore primordium appears while the remainder of the carpophore is seemingly undifferentiated, and the distinction of pileus and stem comes later. In this group we find Agaricus campestris, A . arvensis, A. Rodmani, Armillaria mellea, and Stropharia ambigua. Stropharia epimyces, from its mode of development, is to be placed in the third group mentioned. This may be taken as additional evidence, if any further evidence is needed, that our plant belongs to the genus Stropharia and not to Panaeolus, as was first thought by Peck, since the type of development is the same as that of the only other species of Stropharia that has been studied (11) and of species of the closely related genus Agaricus. Stropharia epimyces as it occurs at Urbana agrees in detail with the description given by Atkinson (i, 2), except that it occurs on Coprinus comatus instead of C. atramentarius as did those collected by Atkinson. The recognition of the host plant was easy in this case since there were many uninfected plants growing along with the infected ones, and also many plants which had "pinhead" and "button" stages of the parasite on them but were not deformed to such an extent as to be unrecognizable. The spores are dis- tinctly purple-black in color. The annulus is often not very prom- iqiq] mcdougall—stropharia 263 inent, but is usually perfectly apparent. The lamellae in mature specimens are adnate to adnexed. Harper (6, 7) has suggested that Stropharia epimyces is identical with Pilosace algeriensis (Fries) Quel. While it is entirely pos- sible that this may be true, our plant cannot belong to Pilosace, as we understand that genus, since it has an annulus and the lamellae are not free. The fact that all species of Agaricus and Stropharia thus far studied {Agaricus comtulus, see Atkinson, 3, is a possible exception) develop in the same way is of interest as indicating a close relation- ship between these two genera. The main taxonomic characters that have been used to distinguish these genera are the free gills in Agaricus and the attached gills in Stropharia. Atkinson (4), however, has found that in developmental stages of Agaricus Rodmani the gills are often attached, and that even in mature specimens of A . Rodmani and A . campestris the gills are sometimes adnexed, thus indicating that such characters do not clearly dis- tinguish the genera. University of Illinois Urbana, III. LITERATURE CITED 1. Atkinson, G. F., Preliminary notes on some new species of fungi. Jour. Myc. 8:110-119. 1902. 2. , A mushroom parasitic on another mushroom. Plant World 10:121-130. 1907. 3- , The development of Agaricus aroensis and A. comtulus. Amer. Jour. Bot. 1:3-22. 1914. 4' , The development of Lepiota cristata and L. seminuda. Mem. N.Y. Bot. Gard. 6:209-224. 1916. 5' , The development of Agaricus Rodmani. Proc. Amer. Phil. Soc. 54:309-343. 1915. 6. Harper, E. T., The probable identity of Stropharia epimyces (Peck) Atk. vfiih. Pilosace algeriensis Yrits. Mycologia 5:167-169. 1913. 7« , Two parasitic mushrooms. Mycologia 8:65-71. 1916. 8. McDouGALL, W. B., Some interesting mushrooms of Champaign County. Trans. 111. Acad. Sci. 9:125-128. 1916. 9. Peck, C. H., Report of the State Botanist. N.Y. State Mus. Nat. Hist. Bull. 133. 1884. 10. Sherman, Helen, The host plants of Panaeolus epimyces Peck. Jour. Myc. 11:167-169. 1905. 11. Zeller, S. M., The development of Stropharia ambigua. Mycologia 6:139-144. 1914. BRIEFER ARTICLES HYBRID PERENNIAL SUNFLOWERS From time to time references have been made to supposed hybrids between the perennial sunflowers, but there has been no systematic investigation of the subject. Such hybrids, if formed, might in many cases reproduce vegetatively, and so give rise to an essentially uniform group of plants of considerable extent, having the aspect of a true species. At Boulder, Colorado, Helianthiis orgyalis and H. Maximiliani have been growing in close proximity for a number of years. There has appeared close to these plants a distinct form which can hardly be any- thing but a hybrid between the two. Possibly such hybrids will be found growing wild in Nebraska, Missouri, or Texas, if anywhere the ranges of the parent species overlap. In order to bring out the characters of the new plant it is necessary partly to redescribe the supposed parents, especially since the descriptions in the manuals omit several significant characters. The 3 plants involved will be distinguished by the following numbers: (i) H. orgyalis DC; (2) H. orgyaloides, nov. (the presumed hybrid) ; (s) H. Maximiliani Schrad. Stems: (i) very smooth and glaucous to top, much branched, the branches slender; (2) essentially smooth, but roughish to the touch above, nearly as stout as in Maximiliani, and with few branches or short peduncles as in Maximiliani; (3) stout, Httle branched, scurfy, with matted white hairs, thinly hairy at top. Leaves: (i) linear, crowded on stem, i-nerved, but with a strong marginal nervure; surface glabrous; margins slightly undulate, with mere traces of obsolete teeth; width of stem leaf 6 mm.; (2) linear, appearing as in orgyalis, but up to 12 mm. broad, rough to the touch, remotely and indistinctly subdentate; a continuous but looped sub- marginal nervure; (3) broadened, narrow lanceolate, grayish, more or less scabrous on both sides, margins remotely and feebly dentate; no continuous marginal nervure; width of stem leaf 26 mm. Peduncles: (i) slender; (2) stoutish; (3) stout. Disk: (i) dark; (2) yellow, pale green in bud; (3) yellow. Botanical Gazette, vol. 67] [264 iqiqI briefer articles 265 Phyllaries: (i) linear, long, and spreading; (2) long, narrow, and spreading, of the orgyalis type; (3) lanceolate, loose, and spreading. Rays: (i, 2, 3,) without pistils; (i) 11-14, rather more decidedly orange than in Maximiliani, and more or less bifid at end; (2) 14-21, clear bright orange-yellow, essentially Maximiliani color, many with tips deeply bifid; (3) 30, light yellow, more or less emarginate at end, and largely in 2 rows. Diameter oj disk: (i) about 10 mm.; (2) about 16 mm.; (3) about 17 mm. Disk bracts: (i) very hairy at end, covering disk buds at early stage; no produced naked tips; (2) very hairy at end, covering disk buds at early stage, and with short naked tips; (3) hairy at end, covering disk buds at early stage, their tips elongate and sharp, pale green, not hairy. Stigmatic branches: (i, 2, 3) orange. Achcnes: (i, 2, 3) entirely glabrous. Pappus scales: (i) 2, about or hardly half length of corollas; (2) 2, about as long as in orgyalis, sometimes with well defined intermediate squamellae; (3) over half length of corollas. The hybrid is on the whole intermediate. It is surprising that the dark disk is not dominant. A remarkable feature is the deeply bifid ends of the rays in the hybrid, greatly exaggerating the character of the parents. The appearance of intermediate pappus squamellae is a common feature in true orgyalis. Essentially this hybrid is evidently known to the trade, although not described. We have a purchased plant belonging to it, but differing from the one just described in the following particulars: stems freely branching as in orgyalis, and plant as tall as orgyalis .-"^ disk olive green in bud ; peduncles rather slender; rays 18-24, only slightly emarginate at end; disk about 15 mm. in diameter; lobes of disk corollas light orange, the extreme tips reddened, or whole lobe suffused with red. F. C. Heinemann, of Erfurt, Germany, advertised a supposed hybrid of this group, calling it H. perennis hybridus pyramid- alis; but from his figure it seems to be simply H. Maximiliani. It is clear that our hybrid does not agree with H. Dalyi Britton or H. Kellermani Britton. Dr. J. C. Arthur writes me that some years ago, near Madison, Wisconsin, he saw a considerable growth of H. Kellermam, and near by H. grosseserratus and H. orgyalis, or what appeared to be such. He had the idea that H. Kellermani was a hybrid between the latter plants, but on attempting to make the same cross 'C. PuRDY (1916) advertises a very tall form of H. Maximiliani, said to grow to II ft. in height. Is this perhaps a hybrid ? 266 BOTANICAL GAZETTE [march artificially at Lafayette, he was unable to get any seed. H. Dalyi, according to Farwell, is a variety of H. Maximiliani, related to it much as var. oppositifolius Farwell is related to H. giganteus. An effort should be made, however, to raise a giganteusY. Maximiliani hybrid. H. amhiguus (Gray) Britton is supposed to be a hybrid giganteusX divaricatus, or at any rate to have giganteus as one parent. Thellung records a garden hybrid laeiiJlorusXrigidus, and the plant called H. serotinus Tausch (i 828) is supposed to be strumosus X rigidus. Evidently there is a great deal to be done, both in the field and in the garden, before we can reach a fairly clear understanding of this subject. It seems possible that in this genus the origin, through hybridization, of distinct plants, with the attributes of species, may be demonstrated. — T. D. A. CocKERELL, University of Colorado, Boidder, Colo. RELATIONSHIPS WITHIN THE RHODOSPOREAE During the last 15 or 20 years the author has studied the structure of 30 to 40 species of Pluteus and 4 or 5 species of Volvaria. In all of these species, without a single exception, the trama of the lamellae presents a curious and interesting structure. In the majority of the Agaricaceae, the trama hyphae of the lamellae lie, in general, in a parallel direction, . as in Mycena, Tricholoma, Collybia, Inocybe, Entoloma, Leptonia, etc. In Russula and Lactarius many of the cells are so swollen that the trama of the lamellae presents a vesiculose appearance. In Amanita the hyphae show a strong divergence from the median plane toward the subhymenium as they descend in the trama. In Pluteus and Volvaria, on the other hand, the most prominent hyphae converge as they descend in the trama of the lamellae. Along the median plane of the lamella there can usually be seen, in section, a layer of hyphae (sometimes more slender) against which these prominent cells converge. Attention was called to this peculiar structure in Pluteus seticeps in 1902,^ but no interpretation was offered as to its origin or significance. During the summer of 191 7, Professor Leva B. Walker, of the University of Nebraska, while studying the development of Pluteus ^ See Leptonia seticeps Atkinson, Jour. Myc. 8:116. 1902. Further collections and studies of this species show that it is a Pluteus. While the gills are attached to the stem before the expansion of the plant, they become free, rounded behind, and distant from the stipe. The stipe also easily separates from the pileus, and other structural characters are clearly those of Pluteus. It is therefore Pluteus seticeps Atkinson, ined. iqiq] briefer articles 267 admirabilis Pk., in this laboratory, discovered the origin of this structure. These prominent h\-phae are long, stout, subcylindrical cells, which arise from the inner cells of the subhymenium and extend downward, con- verging by their free ends against the trama, often compressing it into a thin layer. In some species the free ends of these cells are subcapitate, with a slight constriction below the capitulum. It is clear, therefore, that these peculiar structures in the trama of the lamellae of PliUeus and Valvar ia are internal cysj:idia of a special kind. They are different in form from the other tN-pes of cystidia which project outward beyond the surface of the hymenium in many species. The origin and develop- ment of these internal cystidia will be described by Professor Walker in a forthcoming paper on the development of Pluteiis admirabilis. The presence of these numerous internal cystidia, giving a distinct structural aspect to the trama of the lamellae in Pluteus and Volvaria, indicates that there is a very close phylogenetic relation between these two genera. In both genera the stipe is separable from the pileus, and the spores are of the same t\^e, being smooth, globose, or slightly elongated. No true species of Annular ia have been examined in the fresh condition to determine whether or not internal cystidia are present. It cannot now be stated with confidence to which group it belongs, but other morphological features indicate that it is more closely allied to Pluteiis and Volvaria. In the other genera of the Rhodosporeae the stipe is not separable from the pileus, the spores are angular, and these peculiar internal cystidia are absent. In the Rhodosporeae, therefore, there are at least two distinct phyletic lines. The relation of the genera to these two phyletic lines may be represented as follows. I. Pluteinae. — Pileus easily separable from the stipe; lamellae with numerous internal cystidia converging as they descend and some- times nearly obliterating the trama by pressure; spores smooth. Plu- teus and Volvaria {} Anntdaria) . II. Entolomatinae. — Pileus not separable from the stipe; lamellae without numerous internal cystidia, trama normal; spores angular. Entoloma, Leptonia, Clitopilus, EccUia, Nolanea, and Claudopus. — Geo, F. Atkinson^ 3 This paper was received from the late Professor Atkinson in January 1918. CURRENT LITERATURE MINOR NOTICES British lichens. — The first edition of part first of this work appeared in 1894. .The catalogue was done by James M. Crombie, who accompHshed an excellent piece of work, viewed from the standpoint of the Uchenists of his day and the generation preceding him. Part second was to have appeared shortly from the pen of Crombie, but his death prevented. Later the task was continued by Annie Lorrain Smith, and the second part appeared in 1911, following in part the methods of the author of the first part, on which it was an improvement as a whole. In 1918 the revision of part first by Miss Smith^ appeared. The volume contains an introduction dealing with the nature of the lichen and a catalogue of the British lichens (pp. 520). The work of the original author has been thoroughly revised, and the volume contains much new material and marked changes in classification. There are usable keys, and the descriptions have dropped most of the antiquated phraseology of the lichenist of the recent past. They will be more acceptable, therefore, to the botanist of today, who necessarily has been somewhat appalled at the technical language of the lichenist. The 71 plates represent a large amount of work and will be found helpful to the student. Unfortunately, the author's conception of the nature of the lichen is wrong and therefore imfortunate in a work of first rank. In America, at least, and we believe in Europe as well, there has been a marked recent trend of opinion among students of lower plants to the eft'ect that the dual hypothesis regarding lichens is untenable, and that the lichen must be a fungus after all, parasitic on an alga. However, the introduction on the nature of the hchen is a very minor feature of the work, and the publication of this revision gives us from the pen of one person a complete, creditable, and very useful catalogue of the British lichens. Although the catalogue in two volumes deals primarily with the lichens of a limited area, the work will be found useful in the study of the lichens of America and other regions. — Bruce Fixk. American gall insects. — Felt^ has published an excellent and usable key to the American gall insects. It is arranged with reference to the host plants on which the galls occur and will be of great value to every botanist who has ' Smith, Annie Lorrain, A monograph of the British lichens. Vol. I. Svo. pp. xxiv+520. pis. 71. figs. II. The British Museum. 1918. ^ Felt, E. P., Key to American gall insects. N.Y, State Museum Bull. 200. pp. 310. ph. 16. 191 7. 268 1919] CURRENT LITERATURE 269 occasion to collect and determine plants or to study abnormal plant growths. It is the only reasonably complete publication of the kind in America. There are a total of 1439 species, most of which can very readily be recognized by means of the key, the 250 text illustrations, and the 16 full page plates. The pubhcation is indexed both with reference to the host and the parasite. — Mel. T. Cook. North American Flora. — The sixth part of volume 22 contains the con- clusion of Rosaceae by Rydberg, including genera 54 to 57, much the largest genus being Rosa, with 129 species, 23 of which are described as new. The part closes with 26 pages of additions and corrections to the volume. The first part of volume 32 contains the beginning of Rubiales by Standley, 8 genera of the Rubiaceae being presented. Much the largest genus is Rondektia, with 109 species, 8 of which are described as new. Among the remaining genera 8 new species are described. — J. ^M. C. NOTES FOR STUDENTS Secondary dormancy in seeds. — Kidd and West^ have continued the study of the controlling action of carbon dioxide on the germination of seeds of Brassica alba. In two previous papers by the senior author it has been shown that low concentrations of carbon dioxide inhibit the germination of seeds, and that temperature and oxygen pressure determine the concentration necessary to inhibit germination. In normal oxygen pressure 2-4 per cent carbon dioxide will inhibit germmation at 3° C, while at 20° C. it requires 20-25 per cent. At 17° C. it requires 9-12 per cent carbon dioxide to inhibit with 5 per cent oxygen pressure and 20-25 per cent carbon dioxide with 20 per cent oxygen pressure. All seeds studied, except Brassica alba, germinate normally as soon as the carbon dioxide is removed, while B. alba remains dormant after the carbon dioxide is removed. The authors term this "second- ary' dormancy," in agreement with the usage of this term by Crocker. In the production of secondary dormancy the authors note the following general facts: (i) secondary dormancy is not produced if oxygen is absent during the primary period of inhibition or if carbon dioxide has been used in too high concentration; (2) conditions during the primary period of inhibition which prevent subsequent occurrence of dormancy are the ones that exercise injury on the radicle; (3) 100 per cent dormancy is obtained only within narrow limits of carbon dioxide and oxygen pressure. Secondary dormancy is not produced by a change in the permeability of the coats to gases or water, or to an increase in their breaking strength, but by a change in the embryo 3 Kidd, F., and West, C, The controlling influence of carbon dioxide. The production of secondary dormancy in seeds of Brassica alba following treatment with carbon dioxide and the relation of this phenomenon to the question of stimuli in growth phenomena. Ann. Botany 31:457-487. 1917. 270 BOTANICAL GAZETTE [march by which it becomes less responsive to germinative conditions. The following conditions caused the secondarily dormant seeds to germinate: removal or partial removal of the testa; redrying of the soaked seeds; short exposures to high or low temperatures; treatment with acids (especially no. 01 HCl and propionic); treatment with high concentrations of carbon dioxide followed by germination in air. High partial pressures of oxygen had no effect on the germination of secondarily dormant seeds. The authors give the following interpretation of this work: "It wiU be seen that the main interest of this communication centers around the causes underlying the initiation of growth rather than in the conditions of dormancy. In considering this question of growth in the case of seeds of B. alba, our experi- ments show clearly that there is no question of limiting factors. We have been able to trace no limiting factor responsible for the non-germination of white mustard seeds showing secondary dormancy. We find ourselves rather in the presence of facts which emphasize a conception of stimulus. It has been seen that widely different treatments, quite unclassifiable in any feature other than that they all result in injury and death, if carried too far, excite germina- tion and growth of white mustard seed. It appears to us probable that some return will have to be made to this conception of stimulus in plant physiology generally, and that in any experimental analysis of the living plant, as a unit and in relation to its life-cycle, the idea of limiting factors, which has so long dominated the minds of plant physiologists, will have to be modified." — Wm. Crocker. Chondriosomes in plants. — Investigations dealing with chondriosomes have become so numerous that it seems worth while to make a brief summary of the results obtained. As might be expected, a few structures of different nature have been called by the same name; but a host of names have been apphed to the same structure, so that we have mitochondria, chondriosomes, chon- driomites, chondriokonts, chromidia, sphaeroblasts, histomeres, plasmosomes, cytomicrosomes, etc. The "chondr," meaning a small grain, was chosen because most of the bodies are in the form of small granules; the "mito," meaning thread, is often suggestive, because the granules have a tendency to become arranged in rows. The terms mitochondria and chondriosomes will probably survive, and if a choice should be made between these two, it should be the latter, since it is noncommittal; while the fact that the threadlike arrangement is by no means imiversal is an objection to the term mitochondria. The name chromidia was applied because the writer believed that the granules were portions of the chromatin extruded from the nucleus. Suth granules certainly occur in animals and possibly in plants, but they are not the same structures as the chondriosomes. A historical resume of the subject, with a very complete bibliography up to 1 9 14, was compiled by Cavers,'' and in an investigation upon the rela- 4 Cavers, F., Chondriosomes (mitochondria) and their significance. New Phytol. 31:170-180. 19 14. iqiq] current literature 271 tion between chondriosomes and plastids, Mottiers has brought the Hterature up to 1918. La Valette St. George, working upon the male cells of insects, gave the first description of chondriosomes. He introduced the term cytomicrosomes. Meves, in 1904, gave the first description for plants, using the tapetal cells of anthers of Nymphaea for material. Lewitski, in 1910, was first to claim that chondriosomes give rise to plastids. A little later he made a comparative study upon living and fixed material, showing conclusively that the bodies are present m hving cells. The investigation by Mottier, to which reference has already been made, proves that some chondriosomes give rise to chloroplasts and leucoplasts. He also believes that the chondriosomes are permanent organs of the cell, of equal rank with the nucleus. Of course he recognizes that chloroplasts and leucoplasts also multiply by division. His claim that chondriosomes are concerned in the transmission of hereditary characters does not seem to be so well supported. Some investigators have suggested that chondriosomes transmit characters of the cytoplasm and that the chromosomes transmit characters of the nucleus. It seems to be established that chondriosomes are not artifacts, that they multiply by division, and that some of them give rise to plastids. Their role in heredity, if they have any, still remains to be demonstrated.— C. J. Chamberlain. Trimorphism of Pontederia.— The family Pontederiaceae is notable as containing the only known heterostyled species among monocotyledons (with possibly one exception), and is further remarkable among heterostyled plants as furnishing the only recorded examples of distinctly zygomorphic flowers in such plants. Hazen^ has recently published interesting observations on Pontederia cordata L. Leggett had reported in 1875 that this species was trimorphic, and the present paper is a detailed study of the flower forms, pol- lination, insect visitors, etc. The tubular perianth is slightly zygomorphic and in all 3 flower forms presents 2 sets of stamens: a longer set of 3 on the anterior side of the flower, and 3 short-stalked stamens on the posterior side of the flower. In 2 of the flowers the upper stamens protrude beyond the open perianth. The long- styled stigma reaches a height of 12-13 • 5 mm., the mid-styled form 7-8 mm., and the short-styled form ^-:i . 5 mm. above the base of the ovary. The ratios of the average heights of the 3 lengths of pistils are approximately as 100, 60, and 22. While the arrangement of parts is different in each of the 3 flower forms, it results in 2 sets of stamens adjusted to each length of pistil. The 6 legitimate crosses which may take place between the 6 sets of stamens and the 3 different 5 MoiTiER, D. M., Chondriosomes and the primordia of chloroplasts and leuco- plasts. Ann. Botany 32:191-214. pi. i. 1918. * Hazen, Tracy E., The trimorphism and insect visitors of Pontederia. Mem. Torr. Bot. Club 17:459-484. 1918. 272 BOTANICAL GAZETTE [march pistils are such that each flower type may be pollinated by either of the other 2 flower forms. Moreover, the flowers are placed on the axis of the spike so nearly horizontal as to lessen the probability of self-pollination. The microspores are ellipsoidal in form and the different sets of stamens show marked differences in size of pollen grains, the higher anthers having the larger pollen, the middle ones intermediate, and the short-stalked stamens the smallest spores. This relation suggests a correspondence with the 3 types of stigmas. Averaging a large number of spores it was found that the mean diameters of the 3 sizes of pollen grains were as 100, 80, and 51, and their vol- umes respectively as 100, 53, and 14. Recalling Halsted's work on Eichhornia crassipes, in which he found that all sizes of poUen grains germinated if given sufficient time, but that the larger spores germinated much more promptly than the smaller, Hazen suggests that prompt germination would be of great advantage in the long-styled Pontederia flowers in which the flowers wither so quickly that a slow germinating spore might not have time to function. The author lists observed insect visitors, naming 10 Lepidoptera and 4 Hymenoptera, the least skipper, Ancyloxypha numitor Fabr., being the most frequent visitor. Experimental work by the author is in progress on the relative fertility of the dift'erent flowers with various pollen combinations, and its publication is awaited with interest. — -R. B. Wylie. Phototropism. — Miss Parr,' working in Hottes' laboratory of the Uni- versity of Illinois, has done an excellent piece of quantitative work on the response of Piloholus to light. The literature on phototropism has been full of conflicting statements and theories, very largely due to the lack of quanti- tative work of the type done by Miss Parr. This work does much to show the reasons for these diverse views and to lay the foundations for substantial progress. The physics department of the University assisted in the control of the delicate instruments used in the measurements of light. It is very desirable at this stage of plant physiology that we get the more general cooperation of well-trained physicists and chemists to aid in transforming plant physiology from a qualitative to a quantitative science. The results of the work can best be presented by quoting the summary: (i) Piloholus responds to the light of all regions of the visible spectrum; (2) the presentation time decreases gradu- ally from red to violet, and there is no indication of intermediate maxima and minima; (3) the presentation time does not vary in direct ratio with the measured value of the energy of the light in the different regions of the spectrum; (4) the presentation time varies in inverse ratio to the square roots of the wave- frequency; (5) the product of the square root of the frequency times the pres- entation time decreases with the decrease in the energy value of the spectral regions and is an approximate constant for a given light source; (6) the spectral energy in its relation to presentation time may be expressed approximately in 7 Pare, Rosalie, Response of Piloholus to light. Ann. Botany 32:177-205. 1918. iQig] CURRENT LITERATURE ■ 273 the Weber-Fechner formula, if the wave-frequencies be made a function of the constant; (7) the relation of the spectral energy to the presentation time may also be approximately expressed in the Trondle formula, the wave-frequency being made a function of the constant. — Wm. Crocker. Breeding for disease resistance. — It has been a popular impression that newly produced disease resistant varieties will gradually lose their immunity in later generations. The idea was that such new varieties might sometimes become slightly infected; this short sojourn of the disease organism in the normally immune host would enable the former to adapt itself to the new con- ditions and gradually acquire virulence, until finally a new biologic form was developed to which the host in question was quite susceptible. Evans* carried the same idea further when he found that a cross between resistant and sus- ceptible races of wheat produced a hybrid even more susceptible to rust than the susceptible parent. Furthermore, rust from the hybrid could now infect the immune parent. Such facts were very discouraging, since they indicated that the artificial breeding of resistant crop plants is rapidly overtaken by the natural breeding of new biologic forms of the disease organism. Particularly acceptable, therefore, is the work of Stakman, Parker, and PiEMEiSEL,9 who find that wheats resistant to rust remain resistant regardless of the previous history' of the rust; the gap between immune and susceptible varieties is not bridged by transitional varieties or by artificial hybrids. "Re- sistance is rather an hereditary character, which cannot be produced by the accumulation of fluctuating variations within a susceptible line, nor broken down by changes in the host or parasite." Acceptable as such a conclusion may be, both to commercial breeders and to academic geneticists, it is very questionable how widely it may be applied. It will be difficult, although not hopeless, to explain away much of the contrary evidence. — Merle C. Coulter. Nature of monocotyledonous leaves. — Mrs. Arber'" has presented the results of an anatomical investigation of the phyllode theory of the mono- cotyledonous leaf. According to DeCandolle, it is equivalent to the leaf- base and petiole of a dicotyledonous leaf, but Mrs. Arber believes that certain monocotyledonous leaves are still further reduced in that they are equivalent to leaf-bases only. In case the monocotyledonous leaf shows a distinction of petiole and blade, Henslow suggested that the blade is merely an expansion * Evans, I. B. P., South African cereal rusts, with observations on the problem of breeding rust resistant wheats. Jour. Agric. Sci. 4:95-104. 191 1. 9 Stakman, E. C, Parker, John H., and Piemeisel, F. J., Can biologic forms of stem rust on wheat change rapidly enough to interfere with breeding for rust resist- ance? Jour. Agric. Research 14:111-123. pls.i^ij. 1918. '" Arber, Agnes, The phyllode theory of the monocotyledonous leaf, with special reference to anatomical evidence. Ann. Botany 32:465-501. figs. 32. 1918. 274 ■ BOTANICAL GAZETTE [march of the apical region of the phyllode and not homologous with the blade of a dicotyledonous leaf. Such a blade among monocotyledons Mrs. Arber calls a "pseudo-lamina." Such theories have been devised to explain the parallel venation of monocotyledonous leaves. Attention is also called to Gray's suggestion that some gymnosperm leaves may be equivalent to petioles, and the further suggestion made that this may be applied specially to the Gnetales. These views were tested by Mrs. Arber in anatomical investigations, comparing scale-leaves, petioles, and phyllodes of dicotyledons with the leaves of monocotyledons, the conclusion being reached that the occurrence of inverted vascular bundles toward the adaxial face of a leaf may be an indication of "phyUodic morphology." Other indications of phyllodic anatomy are devel- oped, and its systematic distribution shows that it does not occur with any frequency outside the Helobiae, Liliiflorae, and Farinosae. This distribution is taken to confirm the view that phyllodic anatomy is an ancient character, reveaUng the origin of the monocotyledonous leaf. — J. M. C. Stomata. — Rehfous" has published a detailed study of the stomata of many groups. The details are too numerous for citation, but some of the general conclusions may be indicated. He is convinced that stomata are of first importance in indicating phylogeny and relationships. Their structure he claims is very constant within a group, numerous examples of this being given. For example, the structure of the stomata of the Amentiferae shows that they are nearer the level of the dicotyledons than of the gymnosperms or pterido- phytes. In the same way it is shown that the Polypodiaceae constitute a special group, and that the Osmundaceae, Gleicheniaceae, and Schizeaceae approach more nearly the higher plants. A close resemblance is found between the stomata of cycads and conifers, leading to the conclusion that these groups are of common origin. Numerous illustrations of claimed relationships within great groups are either confirmed or contradicted. Several new types of sto- mata are described, among which those of Polypodium, Platyceriiim, Cycas, and Casuarina may be cited. In connection with the last named genus it is pointed out that its stomata are related to those of certain monocytoledons, as the grasses and certain of the xerophytic Liliaceae. The contribution is a valuable assemblage of facts in reference to the structure of stomata, accom- panied by clear illustrations. The conclusions drawn from these facts are open to discussion. — J. M. C. Water conduction in trees and shrubs. — Farmer" has pubHshed the results of an investigation of the comparative efficiency of the wood as a water- conducting tissue in about 60 species of plants, chiefly trees and shrubs. The " Rehfous, Laurent, Etude sur les stomates. Univ. Geneve, Inst. Bot. IX, no. 6. pp. no. figs. 125. 191 7. " Farmer, J. Bretland, On the quantitative differences in the water-conductivity of the wood in trees and shrubs. Proc. Roy. See. B. 90:218-250. 1918. iqiq] current literature 275 intake of water by the roots and its transpiration from the leaves have been much investigated, but " the behavior of the wood as the intervening conducting channel has almost entirely been neglected." The method used was to measure the amount of water passing in a given time and at standard pressure through a definite length of twig, the area of the cross-section of the wood being carefully measured. The paper includes two parts, one dealing with evergreens and the other with deciduous plants. Some of the res.ults are as follows. The specific conductivity of evergreens is relatively low, while that of deciduous plants is relatively high, and with a higher fluctuation. Some of the deciduous trees are more influenced by environmental conditions than are others. Considerable difference, in a lowering of conductivity, was found between the adult wood of the tree and that of "leaders" of young trees, a difference which becomes "exaggerated" in the main shoot of most cHmbers. The wood of arborescent monocotyledons was found to be defective in water-conductivity. The facts suggest that the lower conductivity of evergreens may be attributed to their narrow and short vessels. — J. M. C. The Journal of General Physiology. — Many will welcome a new Journal of general physiology.'^ Both plant and animal physiology have suffered from being too little related and treated as distinct subjects. Such a publication will aid in bringing them into closer relation. This journal is sure of sufficient financial support and no doubt able editorship. Its aim is stated as follows: " The Journal of General Physiology is devoted to the explanation of Hfe phenom- ena on the basis of the physical and chemical constitution of living matter." The first number contains the following articles: On the dynamics of photo- synthesis, W. J. V. OsTERHOUT and A. R. C. Haas; A method of studying respiration, W. J. V. Osterhout; The antagonism between thyroid and parethyroid glands, E. Uhlenhuth; Difference in the action of radium on green plants in the presence and absence of light, C. Packard; Amphoteric coUoids, J. Loeb; A theory of the mechanism of disinfection, hemolysis, and similar processes, S. C. Brooks; The law controlling the quantity of regenera- tion of the stem of Bryophyllum calycinum, J. Loeb; Reversal of reaction by means of strychnine in planarians and starfish, H. R. Moore; Light and the muscle tonus of insects; the heliotropic mechanism, W. E. Garrey; Lutear cells and hen-feathering, Alice M. Boring and T. H. Morgan. — Wm. Crocker. Embryo sac and fertilization in Oenothera. — Ishikawa''' has investigated the behavior of the gametophytes and the fertilization phenomena in 0. nutans ^^ The Journal of General Physiology, editors, Jacques Loeb and W. J. V. Oster- hout. Published bimonthly by the Rockefeller Institute for Medical Research. Vol. I. No. I. September 1918. Subscription $5.00. '" IsHiKAWA, M., Studies on the embryo sac and fertilization in Oenothera. Ann, Botany 32:279-317. pi. 7. figs. 14. 1918. 276 BOTANICAL GAZETTE [march and 0. pycnocarpa and their hybrids, both of which species were formerly- included in O. biennis. Many valuable confirmatory details need not be cited, but the following may be mentioned. The embryo sac is 4-nucleate, lacking antipodals and one of the polar nuclei, and this condition was found not only in Oenothera, but also in Ludwigia, Gaura, Godctia, and Circaea. The author regards it as a diagnostic character of Onagraceae, and therefore would exclude Trapa, with its normal 8-nucleate sac, from the family. This condi- tion in Onagraceae he thinks may have been produced by mutation, but not by adaptation. The pollen tube enters the synergid and the "mixed plasma" flows out and spreads over the egg. The cytoplasm of the pollen grain was found to contain an immense number of minute starch grains, which migrate through the pollen tube, enter the synergid, and finally disappear. The male nucleus is inclosed in a distinct plasma sheath until it reaches the egg. The synergid and the upper two-thirds of the egg have a distinct cellulose membrane, the lower part of the egg acquiring it after fertilization. Self -sterility of some hybrids is said to be due to the feeble growth of the pollen tube. — J. M. C. Histology of phloem. — There has been a tendency in recent years to assume that the doctrine of recapitulation is a law as valid and invariable as the laws of physics and chemistry, and to use it as a reliable short cut in the study of the evolution of plants. However, it is to be emphasized that a law is a statement of fact, not a theory or working hjrpothesis. If the doctrine of recapitulation and similar generalizations are to be accepted as true laws they must be capable of statistical or experimental proof. MacDaniels's points out that, although in a considerable number of woody dicotyls which he studied there is no funda- mental difference between the type of sieve tube found in seedhngs and first annual rings and that found in the mature condition, the remaining forms pos- sess a presumably less primitive type of structure in the earlier than the later stages of ontogeny. Furthermore, he shows that there is no close parallelism in the speciahzation oT sieve tubes, vessels, and floral structures. It has been a common morphological fallacy to assume that because the evolution of a selected structure progresses apparently in a given direction the sums of all structures (organisms) are moving in a similar direction. MacDaniels' comprehensive and painstaking piece of work is a valuable contribution to our knowledge of the histology of phloem. — I. W. Bailey. Enzyme secretion. — The influence of such inorganic salts as the nitrates, chloride sulphates, and monobasic phosphates of sodium and potassium, and the chlorides and sulphates of calcium and magnesium on the secretion of diastase by PeniciUium camemheriii has been investigated by Robbins.'^ '5 MacDaniels, L. H., The histology of the phloem in certain woody angiosperms. Am. Jour. Bot. 5:347-378. 1918. '^ RoBBiNS, W. J., Influence of certain salts and nutrient solutions on the secretion of diastase by PeniciUium camembertii. Amer. Jour. Bot. 3:234-260. 1916. iqiq] current literature 277 The general results show decrease in the amount of digestion of starch by the fungus in the presence of low concentrations (M/io,ooo and M/ioo,ooo) of the chlorides and sulphates. The view is taken that the decreased digestion is caused by decreased secretion of diastase rather than by inhibition of the activity of secreted diastase. Potassium salts decrease secretion more than corresponding sodium salts. Experiments with nutrient solutions instead of single salts showed the same general effect, decreased secretion. No evidence was found to support the idea that calcium or potassium is intimately related to diastase formation. On the other hand, nitrogen may possibly have some relation to enzyme formation. Nitrates added singly increase the actual amount of starch digestion, but since the mycelial growth is much increased, there is really less digestion per unit of dry weight of mycelium. — C. A. Shull. Reaction of the medium and nitrogen assimilating organisms. — Fred and Davenport'7 have studied the relation of the legume bacteria and Azobacter to low concentrations of acids and alkalies. When sulphuric acid was added to the nutrient solutions, the following hydrogen ion concentrations were found to be critical for the various legume organisms: alfalfa and sweet clover, Ph 4.9; garden pea, field pea, and vetch, Ph 4 . 7 ; red clover and common beans, Ph 4 . 2 ; soybeans and velvet beans, Ph 3 3; lupines, Ph 3 . 15. The authors believe a correlation exists between the acid resistance of the bacteria and the acid resistance of the higher plant with which they are associated. These organisms are not injured by normal alkali additions to the culture medium until the addi- tion is about 10 times that of sulphuric acid producing injury. There seems to be little difference in the several strains as to the alkali resistance. Azobacter is limited to a much narrower range of reaction than are the legume organisms, the critical limits being 6 . 5 Ph for acid and 8 . 6 Ph for alkali. It is to be regretted that the reaction was not determined by the gas chain as well as by the colorimetric method. — Wm. Crocker. 9 Transpiration. — Duggar and Bonns'* have issued a third paper from the Missouri Botanical Garden on the effect of a film of Bordeaux mixture and other films on the transpiration of leaves. In potted mesophytes such a film increases generally the transpiration at night, but has less or no effect during the day. Similar behavior is shown by excised leaves. In Cyperus esculentus, a plant of xerophytic surface modification, such films have no effect on transpiration rate. The writers offer as tentative the following explanation: the film of Bordeaux mixture on the surface of a plant in a state of guttation acts more or less as a bibulous surface, taking water directly from the interior of the plant, through at least some continuous water channels '7 Fred, E. B., and Davenport, Audrey, Influence of reaction on nitrogen- assimilating bacteria. Jour. Agric. Research 14:317-336. 1918. •8 Duggar, B. M., and Bonns, W. W., The effect of Bordeaux mixture on the rate of transpiration. Ann. Mo. Bot. Gard. 5:153-176. 1918. 278 BOTANICAL GAZETTE [march established by means of the open water-suffused stomata. This would account for the effectiveness of the film at night and for its lack of effectiveness with Cyperus with its very narrow stomata. The authors state that there are difficulties in the incipient guttation explanation as applied to excised leaves. — Wm. Crocker. • Turgor movements. — Blackman and Paine/' by use of a special con- ductivity cell, have studied the conductivity of the liquid extruded from the lower half of the excised pulvinus of Mimosa pudica due to the shock stimulus. The shock response gives an increase in conductivity, but not nearly enough to attribute the contraction to increased extrusion of solutes. They beUeve, therefore, that the contraction is due to a sudden condensation of solutes within the pulvinal cells of the lower half of the pulvinus. They consider the con- ductivity method far superior to the plasmolytic method used by previous authors, for it answers directly the amount of movement of solutes. Under certain conditions they get autonomic movements of this organ similar to those of the leaflets of Desmodium gyrans. A slow rise of temperature up to 50° C. shows little increase in exosmose of electrolytes from this organ. The increase of permeabiUty at higher temperatures seems to be due to lethal irreversible changes. — ^Wm. Crocker. Alternation of generations in Padina. — Padina variegata, one of the Dictyotaceae, is abundant at Beaufort, North Carolina, where it has been studied by Wolfe.^" Sperms, eggs, and tetraspores are borne on 3 separate plants which look aUke in the vegetative condition, but which are easily recognized during reproduction. Tetraspores give rise to only male and female plants in approximately equal numbers, so that sex is probably predetermined during the reduction division in the tetraspore mother cell. Fertilized eggs produce only tetrasporic plants, so that there is an alternation of sporophyte and gametophyte generations. Eggs often germinate without fertiUzation, but plants of such parthenogenetic origin do not mature. It would be interest- ing to know the chromosome numbers, especially in the parthenogenetic plants, and we hope that Wolfe, who is familiar with the cytological technique of the algae, will investigate this phase of the problem. — C. J. Chamberlain. The luminous moss. — Tod A" has made a physiological study of Schistostega osmundacea, the so-called luminous moss, his material having been obtained from a cave in Japan. He found the optimum intensity of light as well as the minimum and maximum intensities in terms of Bunsen's unit. In a dark place '9 Blackman, V. H., and Paine, S. G., Studies in the permeability of the pulvinus of Mimosa pudica. Ann. Botany 32:69-85. 1918. '" Wolfe, J. J., Alternation and parthenogenesis in Padina. Jour. Elisha Mitchell Scientific Soc. 34:78-109. 1918. '^ ToDA, Viscount Yasumochi, Physiological studies on Schistostega osmundacea (Dicks) Mohr. Jour. Coll. Sci. Tokyo 4o:no. 5. pp. 30. pis. 2. 1918. igig! CURRENT LITERATURE 279 the protonema can live for 7 months without producing a leafy shoot. He observed also the movement of "chomatophores," which became scattered in a day w-hen the protonema is placed in light, and when the direction of light is changed they all turn toward it in 7-10 days. Blue and violet light proved to be more favorable than any other of the visible rays, excepting of course white light. The optimum temperature for the development of the leafy shoot is 16-25° C.; the protonema does not die so long as the temperature is above — 20 . 5° C, but the leafy shoot dies at — 18° C. The spore at a temperature of 16-25° C. germinates in one month. — J. M. C. Angiosperm wood lacking vessels. — Bailey and Thompson,** in continuing their work on certain genera of angiosperms in which true vessels are absent from the normal wood of the stem, have obtained additional evidence. Their attention had been called to the occurrence of vessel-like structures in injured roots of a species of Dritnys, which might indicate that the ancestors of the 3 genera investigated possessed true vessels. An examination of these struc- tures has led to the conclusion that they are not vessel-Uke in structure, but are typical tracheids, which occur as well in uninjured stems of the 3 genera. They maintain, therefore, that true vessels do not occur in the xylem of these genera, and that there is no evidence that their ancestors possessed true vessels. — J. M. C. Permeability. — Paine and Saunders*^ find, that the testa of the pea is impervious to various reagents dissolved in water (copper ferrocyanid, sodium chloride, safranin) due to a waxy bloom deposited on the outer surface. This bloom is easily rubbed off so that the testa becomes pervious. In the wrinkled peas the bloom rubs off on the w-rinkles, leaving the depressions still impervious, while in the smooth pea the bloom rubs off uniformly on the whole surface. It is interesting to find such a superficial layer responsible for the peculiar per- meability characters of seed coats, for these characters are generally deter- mined by deeper layers. — ^Wm. Crocker. Agaricaceae of Michigan. — Kauefman,*^ in connection with his very full presentation of the Agaricaceae of Michigan, has monographed Russida (pp. 118-167), Pholiota (pp. 289-314), and Cortinarius (pp. 314-442), as represented in the state. In Russula he recognizes 53 species, 3 being new and 27 edible; in Pholiota 26 species, 4 of which are edible; in Cortinarius 154 species, 13 of which are new and 10 edible. As an illustration of the activity of Charles " Bailey, I. W., and Thompson, W. P., Additional notes upon the angiosperms Tetracentron, Trochodendron, and Drimys, in which vessels are absent from the wood. Ann. Botany 32:503-512. pi. 16. figs. p. 1918. 'i Paine, S. G., and Saunders, L. M., On a peculiarity exhibited by the testa of wrinkled peas. Ann. Botany 32:175. 1918. ^■'KAurFMAN, C. H., The Agaricaceae of Michigan. Mich. Geol. and Biol. Survey, Publ. 26. Biol. Series 5. December 1918. 28o BOTANICAL GAZETTE [march Peck in these groups it is interesting to note that he is credited with i6 species in Russula, ii in Pholiota, and 62 in Cortmarius, and this has to do only with Michigan species. — J. M. C. Seedling anatomy. — Holden and Bexon^s have begun a series of studies on the anatomy of teratological seedlings. The first paper deals with seedlings of Cheiranthus Cheiri, which showed " cotyledonary abnormality ranging from hemitricotyly to tetracotyly." The conclusion was reached that there are at least two methods of cotyledonary increase, cotyledonary fission and dichotomy of the growing point of the cotyledon. A third method is somewhat doubt- fully suggested, namely "the downward displacement of one or more epicotyle- donary leaves." — ^J. M. C. Apogamy in Camptosorus. — Mrs. Brown^* has described a case of apogamy in C. rhizophyllus that occurred in cultures to determine if apogamy covild be induced by the modification of external conditions. The apogamous outgrowth was in general a cylindrical process, with some interesting details as to shape and structure, in which a cluster of tracheids appeared. Previous experimental work had indicated that bright light and relative dryness were the factors involved; but in this case low nutrition seemed to be more important than either.— J. M. C. Tropical species of Eupatorium. — Robinson^^ has published the results of a study of Eupatorium as displayed in the American tropics. The wealth of species illustrates how much of the flora of the world remains to be discovered. There are 39 new species described, in addition to new varieties. He has in- cluded also a revision of the Colombian species, recognizing 93 species dis- tributed among 7 sections. "Keyed recensions" are given also of the species of Venezuela (35) and of Ecuador (50). — J. M. C. The orchids of Java. — Smith,^^ in a fifth paper on the orchids of Java, continues to bring to light the remarkably rich orchid flora of that island. He . discusses 61 species representing 27 genera, including 38 new species and 2 new genera (Chroniochilus and Saccolahiopsis). — J. M. C. A new genus of Compositae. — Pritzel^' has published a new genus {Base- dowia) of Compositae from Austraha. It resembles Helichrysuni, as the name (B. helichrysoides) suggests. The genus is named for Herbert Basedow, state geologist of South Austraha. — J. M. C. ^s HoLDEN, H. S., and Bexon, Dorothy, Observations on the anatomy of terato- logical seedlings. I. On the anatomy of some polycotylous seedlings of Cheiranthus Cheiri. Ann. Botany 32:513-530. figs. 17. 1918. =^ Brown, Elizabeth Dorothy Wuist, Apogamy in Camptosorus rhizophyllus. Bull. Torr. Bot. Club 46:27-30. pi. 2. 1919. ^^ Robinson, B. L., Contrib. Gray Herb. Proc. Amer. Acad. 54:235-367. 1918. ^* Smith, J. J., Die Orchideen von Java. Bull. Jard. Bot. Buitenzorg II. no. 26. pp. 135. 1918. 29 Pritzel, E., Basedowia, eine neue Gattung der Compositen aus Zentral- Australien. Ber. Deutsch. Bot. Gesell. 36:332-337. pi. 12. 1918. The Summer Vacation Opportunity at The University of Chicago The Summer Quarter 1919 will receive the added inspiration of profes- sors and instructors returning from war service in many lands. Stu- dents and teachers, interested in keeping abreast of the times or in completing work already begun, appreciate the opportunity of instruction in a regular season of study under members of the University staff. Scholars desiring to prosecute research in the libraries and laboratories will find facilities for work under the most favorable conditions. Courses are offered in all departments and include undergraduate and graduate instruction in Arts, Literature, Science, Commerce and Ad- ministration, Law, Medicine, Education, and Divinity. SUMMER QUARTER 1919 First Term June 16-July 23 Second Term July 24-Augu8t 29 Students may register for either term or both. For the complete ANNOUNCEMENT of courses address THE UNIVERSITY OF CHICAGO, CHICAGO, ILL. On Growth and Form. By D'arcy W. Thompson, C.B., D.Litt., F.R.S. With 408 illustrations. Demy 8vo. $6.50. "Professor Thompson has written an interesting and valuable book on a topic of enduring interest. ..... Why does the animal kingdom present to us a picture of such infinite variety? .... And why, running through the diversities, are there so many resemblances? .... The purpose of Prdfessor Thompsons book is to explain the 'must be' of structure." — Times. Fossil Plants. A Textbook for Students of Botany and Geology. By A. C. Seward, M.A., F.R.S. , Professor of Botany in the University of Cambridge. Volume III. Pteri- dospermeae, Cycadofilices, Cordaitales, C^xadophyta. \Vith frontispiece and 253 illus- trations. Demy Svo. $5.50. Cambridge Biological Series. British Grasses and Their Employment in Agriculture. By S.F.Armstrong, F.L.S., of the School of Agriculture, Cambridge. With 175 illustrations. Demy Svo. $2.00. Plants Poisonous to Live Stock. By Harold C. Long, B.Sc, (Edin.), of the Board of Agriculture and Fisheries. With frontispiece. Royal Svo. $2.00. Cambridge Agricultural Monographs. Science and the Nation. Essays by Cambridge Graduates, with an Introduction by the Right Hon. Lord Moxjlton, K.C.B., F.R.S. Edited by A. C. Seward, F.R.S., Master of Downing College, Cambridge. Second impression. Crown Svo. $1.50. Published by the Cambridge University Press, England Q. P. Putnam's Sons, Agents in the United States 2, 4 and 6 West 45th Street, New York City FS 2 Microscopes -ftice ♦4075 |{auscli"'|omb ^'ficroscopes are 'again in the market — backed by the prestige which attaches to successful war service in any significant capacity. Our release from this service is of course gradual. As the demands made upon us by the government are lessened from month to month, however, we are enabled to increase deliveries to our other patrons. Write ?or complete catalog of our revised microscope line and inform us regarding your requirements. gausch f:? ]pmb Optical ©. w^;^"}'*- 654 ST. PAUL STREET ROCHESTER, N.Y. San'^Francisco Leading American \fakeTi of Photographic Lenses, Microscopes. Projection Apparatus (Balopticons), Ophthalmic Lenses and Instruments, Photomicrographic Apparatus, Range Finders and Gun Sights for Army and Navy, Searchlight Reflectors, Stereo-Prism Binoculars, Magrtifiers and Other Hi^h-Grfide Optical Products. FINE INKS AND ADHESIVES For those who KNOW Higgin^' Drawing Inks Eternal Wrillns Ink Engrossing Ink Taurine Mucilage Photo Mounter Paste Drawing Board Paste Liquid Paste I Office Paste \ Vegetable Qlue, Etc. Are the Finest and Best Inks and Adhesives Emapcipate jroursclf from the use of corrosive and ill-smelling inks and adhesive? and adopt the niggins Inks and Adhesives. Tiieywillbea revelation to you, tljey are so sweet, clean, well put up, and withal sp efficient. At OftAtTS Qenerafly CHAS. M. ItlQQINS & CO., Mfrs. Branches; Chicago, London 271 Ninth street Brooklyn, N.Y. The Home Typewriter as well as The Business Typewriter is the Self-Starting Remington An the family can use it. Write for descriptive matter. REMINGTON TYPEWRITER COMPANY Incorporated 374 BROADWAY NEW YORK CITY r 'v'blume LXVII Number 4 THE Botanical Gazette Editor: JOHN M. COULTER APRIL 1919 After-Ripening and Germination of Seeds of Tilia, Sambucus, and Rubus R; C. Rose 281 Notes on American Willows. IV - - - - - Camillo Schneider 309 Respiration after Death - - - - - - A. R. C. Haas 347 (With three figures) Briefer Articles George Francis Atkinson ------ H. H. Whetzel 366 (With portrait) Current Literature Book Reviews ----------- 369 Manual of tree diseases Minor Notices ----------- 3"9 Notes for Students - - - - - - - - - "370 The University of Chicago Press CHICAGO, ILLINOIS, U.S.A. THE CAMBRIDGE UNIVERSITY PRESS, London and Edinburgh THE MARUZEN-KABUSHIKI-KAISHA, Tokyo, Osaka, Kyoto. Fukaoha, Sendai THE MISSION BOOK COMPANY. Shanghai Volume LXVII Number 4 THE Botanical Gazette A MONTHLY JOURNAL EMBRACING ALL DEPARTMENTS OF BOTANICAL SCIENCE EDITED BY JOHN M. COULTER With the assistance of other members of the botanical staff of the University of Chicago Issued April 18, 1919 The Botanical Gazette is published monthly by the University of Chicago at the University Press, 5750 Ellis Avenue, Chicago, HI. IfThe subscription price is $7 .00 per year; the price of ; single copies is 75 cents. Orders for service of less than a half-year will be charged at the single-copy rate. ^Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawaiian Islands, Philippine ' Islands, Guam, Samoan Islands, Shanghai. ^Postage is charged extra as follows: For Canada, ■ 35 cents on annual subscriptions (total $7-35)> on single copies, 3 cents (total 78 cents); for all other countries in the Postal Union, 84 cents on annual subscriptions (total $7.84), on single copies II cents (total 86 cents). ^Patrons are requested to make all remittances payable to The University of Chicago Press in postal or express money orders or bank drafts. The following are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.G. 4. Yearly subscriptions, including postage, £1 12s. 6d. each; single copies, including postage, 35. 6d. each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage. Yen 15.75 each; single copies, including postage. Yen 1.75 each. For China: The Mission Book Company, 18 Peking Road, Shanghai. Yearly sub- scriptions, $7.00; single copies, 75 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 84 cents, on single copies II cents. Claims for missing numbers should be made within the month following the regular month of publication. The publishers expect to supply missing numbers free only when losses have been sustained in transit and when the reserve stock will permit. Dehveries are not guaranteed within the war zones of Europe, nor in other countries where transportation is. disturbed by con- ditions incident to war. Business correspondence should be addressed to The University of Chicago Press, Chicago, 111. Conununications for the editor and manuscripts should be addressed to the Editor of the Botanical Gazette, The University of Chicago, Chicago, 111. Contributors are requested to write scientific and proper names with particular care, to use the metric system of weights and measures, and in citations to follow the form shown in the pages of the Botanical Gazette. Papers in excess of thirty-two printed pages are not accepted unless the author is wilhng to pay the cost of the additional pages, in which case the number of the pages in the^yolume is correspondingly increased. ■ Illustrations are furnished without cost to author only when suitable originals are supplied. ^ A copy of the suggestions made in the January number, 1907, will be sent on application. It is advisable to confer with the editors as to illustrations required in any article to be ofifered. Separates, if desired, must be ordered in advance of pubUcation. A table showing approxi-^ mate cost of separates is printed on an order blank which accompanies the proof; a copy will— be sent on request. Entered as second-class matter August 21, 1896, at the Post-Office at Chicago, Illinois, under the Act of March 3, 1879- Acceptance for mailing at special rate of postage provided for in Section 1103, Act of October 3, 1917, authorized oa July IS, 1918. VOLUME LXVII NUMBER 4 THE Botanical Gazette APRIL igig AFTER-RIPENING AND GERMINATION OF SEEDS OF TILIA, SA:\IBUCUS, AND RUBUS CONTRIBUTIONS FROM THE HULL BOTANICAL LABORATORY 247 R. C. Rose Introduction This paper gives the results of an attempt to determine the con- ditions favoring the after-ripening and germination of the seeds of Tilia americana, Sambucus canadensis, and Ruhiis Idaeus, and some of the chemical processes involved therein. Since layering of these seeds usually results in very low percentages of germination, it was thought possible to discover some other means of overcoming their dormancy. Literature The present state of our knowledge of the causes of delay in germination, and the means of overcoming it, is admirably sum- marized in a recent paper by Crocker (5). He divides seeds which show delay in germination into 7 classes. In 3 of these the seed coats play the important role, while in the fourth dormancy is occasioned by the embryo. Where dormancy or poor germination is due to the seed coat, the use of concentrated sulphuric acid as a carbonizing agent has become a common practice. Rose (23) mentions Rostrup (24) as the first to resort to this treatment, and HstS TODARO (25), HiLTNER (12), JaRZYMOWSKI (14), BOLLEY (2), and Lo\rE and Leighty (19) as investigators applying the same method. Ewart (9) found this treatment effective with several 281 282 BOTANICAL GAZETTE [april species of Acacia, as did Crocker (unpublished work) with Scirpus. The length of time required by this treatment varies from a few minues to several hours, depending upon the resistance of the coats. Boiling water or warm water, used as a forcing agent, has proved effective in a number of cases where hard-coatedness is the cause of the delay. Bruyning (3), working with the seeds of JJlex europaeus, found that a treatment of 1-5 seconds with boiling water raised the percentage of germination from 13 for untreated seeds to 53 • 5~75 • 5 for treated seeds. Honing (13) obtained his best results with Alhizzia seeds by soaking them in water at 60° C. for at least 3 hours, while with Mimosa 60-70° C. proved most effective, as did 70-75° C. for Pithecolohium. Soaking seeds of Crotalaria in warm water proved disadvantageous. Bolley (2) states that improvement in germination was obtained by this method if the exposure was not long enough to kill the embryo. NoBBE (21) mentions Alexander von Humboldt as the first investigator to use chemicals as forcing agents. From the time of Humboldt (1793) up to 1873, the date of publication of Nobbe's book, many investigators used as forcing agents a great variety of substances, both organic and inorganic. The range of substances used is more interesting than the results obtained. Moreover, quickly germinating seeds were used and in such cases the effect of forcing agents is not so striking as where dormancy is involved. Of the more recent workers in this field, Lehmann (16) was the first to emphasize the importance of chemical substances in connection with germination. He showed that the seeds of Ranunculus sceleratus are forced into germination by Knop's solution, by soil, by soil wet with weak solutions of hydrochloric acid, potassium hydroxide, ferric chloride, and hydrogen peroxide. Two years later Gassner (10) found Knop's solution effective on unthreshed seeds of Chloris ciliata, and more recently (11) has shown that for several other seeds various nitrogen compounds, especially nitrites and nitrates, are effective forcing agents. Chloris ciliata was found to have a membrane impermeable to potassium nitrate and magnesium nitrate, and from this Gassner concludes that the effect is upon the seed coat alone. Lehmann (17) and Lehmann and Ottenwalder (18), working with seeds representing jgig] ROSE— AFTER-RIPENING AND GERMINATION 283 a number of different families, showed that acids in low con- centrations, especially hydrochloric acid, are effective forcing agents. Crocker and Davis (6) obtained similar results for seeds of Amaranthiis (unpublished work) and Alisma. Bases are equally effective for Sagittaria and Alisma, but not for AmarantJms. According to Ottenwalder (22) bases exert an inhibitory effect on seeds of Epilobium hirsutum. In those cases where a state of dormancy exists in the erribryo itself (Crataegus and Mains), temperatures slightly above freezing have been found effective in hastening after-ripening (7). In Crataegus, as Eckerson (8) has shown, the hypocotyl becomes more acid as after-ripening progresses; hence dilute acids hasten after-ripening by acting upon the hypocotyl directly. Material The seeds used in these experiments were gathered in the summer or the fall of 1 916 and 191 7. Each year those of Sambucus were all collected on the same day from neighboring plants. Tilia seeds of the 191 6 crop were collected during October from trees growing on the dunes at the southern end of Lake Michigan. The 191 7 crop was gathered during September from trees in the parks of Washington, D.C. The seeds of Rubus were collected during late June 1916 from neighboring plants of several varieties, but no attempt was made to keep those of the different varieties separate. Among the seeds of all 3 species were found many without embryos or with defective embryos. In most cases this fact accounts for the varying number of seeds used in the cultures. Approximately , 60 per cent of Ruhus, 75 per cent of the 1916 Sambucus, and 80 per cent of the 191 6 crop of Tilia were viable. Not more than 5 per cent of the 191 7 crop of Tilia and Sambucus were defective. Histology and microchemistry of seed coats Sambucus: Endocarp. — The seed in cross-section shows in the lignified endocarp 3 regions: (i) the outermost, consisting of 3 or 4 layers of cells of irregular size and shape, with thin walls and large lumina; (2) a middle one of i or 2 layers of fibers in cross-section; and (3) an inner one of i or 2 layers of fibers in longitudinal section. Seed coat.— Thi?> consists of several layers of collapsed cells with 284 BOTANICAL GAZETTE [april lignified walls ; the cells contain a considerable quantity of reducing sugar. Tilia: Pericarp. — This is composed of two layers: (i) a surface region of loose fibers with cellulose walls, and (2) a thicker region of lignified fibers. Seed coat. — This consists of 3 regions: (i) cells with suberized or cutinized walls; (2) one layer of paHsade cells with (a) outer end walls of cellulose, (b) a lignified light zone, (c) a pectinized region, and (d) a Hgnified region; and (3) 3 or 4 layers of cells with walls which stain with ruthenium red and give the eerie acid test. RuBUS: Endocarp. — This consists of 2 layers: (i) an outer layer, variable in thickness, of lignified fibers longitudinally arranged in cross-section of the fruit; (2) an inner region of 4 or 5 layers of lignified fibers transversely arranged in cross-section of the fruit. Testa. — This consists of 4 regions: (i) i layer of cushion-shaped cells with lignified walls; (2) 4 layers of collapsed cells with cellulose walls; (3) I layer of collapsed cells with thick pectinized walls; and (4) I layer of cells with cellulose walls which appear as a thickened outer wall of the endosperm. Microchemistry In table I are given the results of the microchemical tests made upon the endosperm and embryo of each of the kinds of seeds used. Owing to the lack of a sufficient number of germinating seeds of Sa?nbucus several of the tests have not been completed. The storage materials in all the seeds are very similar, starch, fats, and protein being found in every case. In addition to these Sambucus contains amylodextrin. Tilia contains much more fat and phytos- terol than either Sambucus or Rubus. The phytosterol shows up as a bright red layer around the fat globules when sections of the seeds are placed in concentrated sulphuric acid. Oxidase is present in the dry seeds in very small quantities, and in the germinating seeds benzidine gives a positive test only after several hours. Peroxidase, while present in dry Tilia seeds, is much more abundant in the germinating seeds. Dry seeds of Sambucus and Rubus give no peroxidase reaction. Catalase is found in both dry and ger- minating seeds of all 3 species. I9I9] ROSE— AFTER-RIPENING AND GERMINATION 285 H < T3 0 g + + 1 + ++ + + + + + + + "2 < H 13 ■| 1 0 0 + 1 + ++++++++I in C 3 H » « T3 0 b g 1 1 1 +++++1 i+l § g a s e W 1 1 1 H^t-^n ' ' +1 0 -a g i- ■w d w .S g g .s C a 0 -a a W 't' T3 0 XI g vt II++ ' tM C g S 0 T3 l' + < C <: M 0 13 -r) XI g t" + ±±±1 1+ + ^'- 1 W l^; 0 O) t3 a. 0 + 1 1 -a c W W il U c Sd ,_( -4—' *j 0 <; •C =3 u >- ^ H -4-> C/3 03 a 0 QJ w rN wp ~P '-' rj 4-1 < Qj rt 0) a )<;piifiH fx. p. E- C p- c )p: 1 286 BOTANICAL GAZETTE [APRIL Conclusive determinations in regard to the reaction of fresh dry seeds of Tilia have not been made, but preliminary tests, where neutral red was used as an indicator, indicate that the endosperm and cotyledons are acid and the hypocotyl alkaline. Seeds kept in dry warm storage for 9 months show an acid reaction throughout. Hydrogen ion determinations, the data for which are given later, showed an acid reaction for the stored seed as well as for the germinating ones. As germination begins, the reaction of the embryo of Sambucus changes from alkaline to acid, but the endosperm remains alkaline. Both dry and germinating Rubus seeds are acid. A qualitative analysis of the ash of Tilia seeds showed iron, calcium, magnesium, potassium, and aluminium present. No tests were made for sodium. Experimental data Freshly harvested Tilia seeds with a moisture content of 10 per cent or less, or seeds kept in dry warm storage for several months, fail to germinate when placed on a moist substratum and kept at room temperature. This is true not only of seeds with coats intact, but for those with the coats chipped or entirely removed. Fungi and bacteria soon attack seeds with the coats broken and decay takes place in a few days. The percentage of water held by air-dry seeds is shown in table II. The seeds used for these deter- minations were dried in a partial vacuum at 80° C. until the weight was constant. TABLE II Water content of air-dry Tilia seeds Condition of seeds Weight of air-dry seeds in gm. Water loss in gm. Percentage of water loss Coats off 1-2754* 1-8559 2 . 1904 1.5024 0.0788 0. 1782 0.1574 0. II30 6.17 Coats on 9.60 Coats on 7.18 Coats on 7. =52 * Average of 4 duplicates. The variations in the percentage of water lost by the seeds with coats on is due to the presence of seed coats which contained no endosperm and embryo. That the failure of air-dry Tilia seeds, iQig] ROSE— AFTER-RIPENING AND GERMINATION 287 coats either on or off, to germinate is not due to an inability to absorb water is indicated by table III. The data given in this table were obtained by soaking seeds in distilled water at room tempera- ture until they had come to constant weight. Here again the TABLE III Water-holding capacity of air-dry Tilia seeds Condition of seeds Coats off Coats on Coats on Coats on Coats chipped . Coats chipped . Coats chipped . Weight of air-dry seeds in gm. 2848* 1540 ,7842 II98 4963 5040 I 9590 Water absorbed in gm. I .2071 0.7841 0.4146 O . 4804 I .5020 15717 1.918s Percentage of water absorbed 93 36 23 22 100 104 97 95 40 24 66 38 50 93 * Average of 4 duplicates. variations in the percentage of water absorbed are in part due to the presence of seed coats which contain no endosperm and embryo. Even with the coats chipped it is not always possible to eliminate all empty coats or defective seeds. The fact that the coats interfere with water absorption to a considerable extent is clearly shown in the table. The fact that seeds with coats removed or chipped, however, and with a moisture content approximately equal to their air-dry weight will not germinate when placed on a moist sub- stratum at room temperature, is sufficient proof that water absorp- tion is not the only limiting factor to growth. That seeds that have been stored in the air-dry condition when the seed coats are intact can be forced to germinate is shown by the following experiment. Approximately 7000 seeds (200 gm.) of the 19 1 6 crop, with pericarps removed and coats chipped, were placed on moist cotton in large Petri dishes and kept at 4-6° C. from March 24, 1917, to June 10, 1917, a total of 78 days. At the end of that time and before being transferred to a higher tempera- ture, several hundreds showed the hypocotyl protruding from the endosperm for i . 5-2 . 5 cm. Of these, 100 were planted in soil in the greenhouse and 71 per cent produced seedlings. A second lot of 100 seeds was planted in soil out of doors, and 64 per cent 288 ■ BOTANICAL GAZETTE [april produced seedlings. Two lots of 500 each were selected from the seeds in which the hypocotyl was still inclosed within the endosperm. These were planted in soil in the greenhouse and in the garden and gave 20 and 25 per cent germination respectively. All seeds not planted were again placed in cold storage. Twelve days later 400 with hypocotyls protruding from the endosperm were planted in soil in the greenhouse. Of these, 348, or 87 per cent, produced seed- lings within a week. By July 24, 1666 of these 7000 cold storage seeds had germinated at a low temperature. Of the ungerminated seeds 100 placed on moist cotton at room temperature gave 31 per cent germination in one week. The roots of these were short and thick and showed a great tendency to coil. At the same time air-dry seeds which had been stored at room temperature, when placed in soil or on moist cotton, decayed. Seeds kept in cold storage showed for the first few days a great tendency to mold, so that it was necessary to sterihze them with a 3 per cent solution of hydrogen peroxide for i hour on two separate occasions. With longer storage an immunity toward fungi is established, and although the coats may be covered with a thick layer of myceha the endosperm and embryo are not attacked. Sections of the seeds examined under the microscope failed to show any hyphae present within the living tissue. On November 6, 191 7, 6 cultures of 50 seeds each of both the 191 6 and 191 7 crops were placed in moist storage at 0-2° C, where they were allowed to remain for 140 days. At the end of that time no germination had taken place, which is in direct contrast with the result obtained in 191 6 with seeds stored at 4-6° C. The failure to obtain germination here is interpreted as being due to the use of too low a temperature. The assumption that the exposure to this temperature was too long will hardly explain the results obtained, since if the temperature were not too low germination should begin as soon as the after-ripening process is complete. The results given in table IV, showing the percentage of germination obtained when these seeds were transferred to a temperature of 10-12° C, indicate that the storage temperature and not the length of exposure to it is the limiting factor. This conclusion is strengthened further by the following experi- ment. Unfortunately no count of the number of seeds germinated iqiq] ROSE— AFTER-RIPENING AND GERMINATION 289 was made, as the experiment was used primarily for a different purpose. Approximately 1000 seeds of each of the 2 crops, stored under the same conditions as those indicated in table IV, showed no TABLE IV Seeds of Tilia stored at o-2°C. for 140 days; then at io-i2°C. Percentage of germination after Number of CULTURE • 12 days at 10-12° C. 19 days at 10-12° C. 1 91 6 seeds 1 91 7 seeds 1916 seeds 1 91 7 seeds I 66 68 70 74 70 19 24 20 12 34 26 22 74 72 80 82 76 S6 28 24 18 34 30 30 2 T. 4 c 6 germination after 140 days at a low temperature. When brought to the higher temperature the 191 6 seeds germinated vigorously and in large numbers for the first 12 days and until the hj-pocotyls were 2-3 cm. long. From this point on no development took place and the seedlings gradually died. Here a temperature of 10-12° C. seems to be too low for continued growth. The 191 7 seeds ger- minated much less vigorously, in fewer numbers, and only a few developed hj-pocotyls 2 cm. long. Comparing the results obtained in 1916 with those obtained in 191 7, it is seen that the seeds after- ripen and germinate at temperatures slightly above freezing. Davis and Rose (7) working with Crataegus found that after- ripening takes place most rapidly at 3-6° C, and that temperatures considerably higher are more favorable for germination and growth. At 0-2° C. Tilia seeds after-ripen but do not germinate. At 4-6° C. after-ripening and germination both take place, the latter taking considerable time. After-ripened seeds germinate poorly at room temperature. Once germination has begun at the low temperature, growth is best at temperatures above 12° C. The germination of Tilia seeds depends, therefore, upon the proper regulation of the temperature, and can be accomplished by a period of after-ripening in moist storage at 0-2° C, followed by a sojourn of 2 or 3 weeks at 10-12° C. until germination is well under way, 290 BOTANICAL GAZETTE [april and finally by a transfer to a still higher temperature in order to permit vigorous growth. These conclusions are drawn from the facts that (i) seeds after-ripened at 0-2° C. did not germinate until transferred to a temperature of 10-12° C; (2) although germination began at the higher temperature, growth soon ceased; and (3) seeds which had been after-ripened and which had begun to ger- minate at 4-6° C. grew well when transferred to soil in the green- house. Table IV suggests that one-year old seeds are better than fresh, but additional data upon this point are desirable. A nursery- man with many years' experience in the growing of trees and shrubs states that if Tilia seeds are allowed to become dry between the time of maturing and the time of layering a low percentage of germination results. On the other hand, if a high moisture content is maintained during this period no difficulty in germination is encountered. Up to the present time the author has been unable to obtain seeds which at the time of gathering had a moisture content of more than 10 per cent, and it seems probable that the water content of Tilia seeds is generally low at harvest time. While these seeds do not after-ripen to any considerable degree in air- dry storage, those that have been in the air-dry condition for a year after-ripen perfectly when put in a moist germinator at a low temperature. There seems to be no injury, therefore, even from protracted air-dry storage. No discussion is necessary to show that field conditions are not those most favorable for the obtaining of high percentages of. germination. Neither does the nurseryman, when layering seeds, control the temperatures to the extent necessary to secure maximum results. Hydrogen ion concentration. — The determinations of the hydrogen ion concentrations were made with the hydrogen electrode. Twenty seeds were pulverized in a mortar, and, except in instances to be noted later, 25 cc. of water added. The temperature varied from 27 to 33° C, but in every case the necessary correction was made. The determinations were made upon the seeds in the unaf ter- ripened condition, after-ripened but not germinated, with hypocotyl 2 mm. to 5 mm. long, and with hypocotyl 0.5 cm. to 2 cm. long. EcKERSGN (8) has already shown that the acidity of the hypo- cotyl of Crataegus increases as after-ripening progresses. Her igig] ROSE— AFTER-RIPENING AND GERMINATION 291 determinations were made by the titration method with pheno- phthalein as an indicator. The Ph of seeds with the hypocotyls 0.5 cm. to 2 cm. long is approximately 4 times as great as that of the unafter-ripened seeds. While this is not as great an increase as that found by Eckerson, it may be due to the fact that her determinations were made upon the dormant organ alone, while here the whole seed was used, or to differences between the two kinds of seeds. Determinations made by the titration method would also probably give values much higher than those obtained by the hydrogen elctrode. Table V shows that the weight of the seeds increases as after- ripening progresses. This is not due to an increase in dry weight, TABLE V Concentration of hydrogen ion of Tilia seeds in different stages OF after-ripening Condition of seeds 1 Air-dry 2 Air-dry 3 Air-dry 4 Air-dr\' 5 After-ripened 6 After-ripened 7 With hypocotyls 5 mm 8 With hj'pocotyls 5 mm 9 With h>'pocotyls 5 mm 10 With hypocotyls 0.5-2 cm 1 1 With h>-pocotyls o . 5-2 cm , i2*With hypocotyls 0.5-2 cm , i3*With h>^ocotyls o . 5-2 cm i4tWith hypocotyls o . 5-2 cm , i5tWith hj-pocotyls o . 5-2 cm 16 After-ripened at room temperature (10 days) 1 7 After- ripened at room temperature (10 days) 2. 24X10-7 2.00X10—7 2.58X10-7 2.40X10—7 3.24X10-7 3.02X10-7 6.76X10-7 5.75X10-7 7.59X10-7 1.18X10-6 1. 10X10-^ 9-33X10-7 9.33X10-7 1. 05X10-6 1. 05X10-6 2.51X10-7 2. 51X10— 7 * 50 cc. of water used. t loo cc. of water used; 25 cc. of water used for all others. since no photosynthesis has taken place, but to the large amount of water absorbed. Eckerson (8) likewise observed an increased water-holding capacity for the hypocotyl of Crataegus as after- ripening progressed. Of greater significance in this connection is the fact that, at least for the most advanced stage of after-ripening, variations in the amount of water used with the sample had Uttle 2g2 BOTANICAL GAZETTE [APRIL effect upon the hydrogen ion concentration. With samples lo and II, 25 cc. of water were used, with samples 12 and 13, 50 cc, and with samples 14 and 15, 100 cc. Although the variation of Ph is considerable, it is by no means as great as that of the amount of water used, nor is it in the same direction. That the degree of dilution has no effect upon the Ph suggests the presence of buffer salts, formed by the action of fatty acids produced during germina- tion and the constituents of the ash already mentioned. After-ripened seeds similar to those used in samples 5 and 6, which had failed to germinate when kept at room temperature for 10 days, gave a Ph corresponding very closely to that shown by unafter-ripened seeds. This suggests that after-ripening is a reversible process, a fact to which Crocker (5) has called attention, and that a decrease in acidity may lead to secondary dormancy. TiTRATABLE ACID. — Determinations of the titratable acid were made upon dry, after-ripened, and germinating seeds. For each determination the seeds were ground in a mortar with 10 cc. of water and titrated with N/io sodium hydroxide with pheno- phthalein as an indicator. Titrations were made with freshly pre- pared samples and with others which had been allowed to stand for 48 hours. To the latter were added 10 drops of toluol and o. 5 cc. of N/io hydrochloric acid. Table VI shows the number of cubic centimeters of N/io sodium hydroxide necessary to neutraUze the free acid in each sample. The figures are the average of dupHcate determinations. Corrections have been made for the acid added. TABLE VI Condition of seeds Fresh samples After 48 hours Percentage of increase Dry 0.41 0.45 1. 18 0.87 1.87 2.80 112. 2 After-rinened 3IS-S Germinating 137.2 While the amount of acid present is greatest in germinating seeds, it is seen that after autodigesting 48 hours the greatest percentage of increase over the freshly prepared samples is in seeds well after-ripened. Here is shown the fact that the after-ripened igig] ROSE— AFTER-RIPENING AND GERMINATION 293 seeds have a great power of increasing their alkali absorption, which may be due to lipase activity. Catalase.— Determinations of catalase activity of dry, after- ripened, and germinating seeds were made by means of Appleman's apparatus (i). The samples, ground in a mortar, were all reduced to the same degree of fineness by rubbing them through bolting cloth. The catalase determinations were made at 25° C. To 5 cc. of water containing o . 02 gm. of pulverized seed material was added 5 cc. of Oakland dioxygen and the amount of oxygen released was measured after i, 2, 3, and 5 minutes of activity. Appleman has pointed out that small amounts of acid greatly reduce or entirely destroy catalase activity. In order to remove this possible source of error the Oakland dioxygen used was neutrahzed by the addition of N/io NaOH, or an excess of CaCOj was added to the meal. The data given in table VII are the averages of duphcate determinations. They show that dry, after-ripened, and germinating seeds, in the order named, exhibit increasing catalase activity. Eckerson (8), employing microchemical methods, arrived at similar conclusions for seeds of Crataegus. TABLE VII Condition of seeds 1. Dry seeds 2. After-ripened (dried 2 daj-s) . 3. After- ripened (dried 2 days) . 4. After-ripened (not dried) .... 5. Germinating Re.action of REAGENT Oxygen in cc. liberated after Neutralized* Neutralized* With CaCOj With CaCOj With CaCO, I minute 2.5 7-1 6.8 6.75 19.4 2 minutes 3 minutes 4.2 11. 8 11. 9 II-3 27.86 5-25 15-4 14.8 15 OS 315 S minutes 7.0 21.5 21 .05 20.75 37-03 '0.80 CC. N/io NaOH to neutralize 25 cc. dioxygen. Drying after-ripened seeds for 2 days at room temperature has no effect on the amount of oxygen hberated, as is shown by com- parison of samples 3 and 4. Further evidence for the effect of the acid of the dioxygen upon catalase activity is shown in table VIII. Determinations made with after-ripened seeds not dried and with germinating seeds gave similar results. A comparison of the last 2 determinations show 294 BOTANICAL GAZETTE [APRIL that the neutrahzation of dioxygen or the addition of CaCOj is sufficient to eliminate any error due to the acidity of the reagent or the meal. TABLE VIII Effect of reaction of solution upon amount of oxygen liberated FROM DIOXYGEN BY Tilia SEEDS Reaction of REAGENT Oxygen in cc. liberated after Condition of seeds I minute 2 minutes 3 minutes S mmutes After-ripened (dried 2 days) .... After-ripened (dried 2 days) .... After-ripened (dried 2 days) .... Not neutralized Neutralized With CaCO^ 2. 1 7-1 6.8 31 11. 8 11. 9 3-6 iS-4 14.8 4-3 21-5 21.05 Oxidase activity. — The determinations of oxidase activity were made on dry, after-ripened, and germinating seeds in Bunzell's (4) simplified apparatus with pyrogallol as the reagent. The material used, except in the case of the dried seeds, had been after-ripened at 0-2° C. for 140 days and then kept at 10-12° C. until a large percentage of the seeds had begun to germinate. After being dried in a vacuum over lime for 3 days at room tempera- ture the seeds were ground in a mortar and the determinations made on 0.02 gm. of meal. Table IX shows the readings in centi- meters of mercury after 3 hours and after 20.5 hours. TABLE IX Oxidase activity of dry, after-ripened, and germinating SEEDS of Tilia Time • Dry seeds After- ripened Hypocotyls i-S mm. long Hypocotyls 0.5-2 cm. long After 3 hours 0.52 0.53 0.67 0.68 1.03 1 . 10 1-52 1.67 2.01 1.92 2-57 2.52 1-39 After 3 hours 152 After 20 . 5 hours After 20 . 5 hours 2.42 2.07 During the experiment the temperature averaged 31.3° C. with a variation of =^=0.1 of a degree. Variations in the volume of air in the tubes due to this slight variation in temperature have been corrected by means of check tubes containing water only. The iqiq] rose— AFTER-RIPENING AND GERMINATION 295 results show that the oxidase activity rises with after-ripening and germination. Once germination has begun, no increase is to be noted. Discussion.— The results obtained show that the dormancy exhibited by the seeds of Tilia is not due to any property of the seed coat, although that structure may serve to lengthen the dormant period, but is to be ascribed to conditions obtaining within the endosperm or the embryo or both. In this respect Tilia resembles Crataegus, and the conditions necessary for after-ripening and germinating of the former are very similar to those required by the latter. Even with these conditions well known and various dif- ferences between dormant and after-ripened seeds clearly shown, it is still impossible to define the term after-ripening in anything more than general terms. The similarity of Tilia and Crataegus, with respect to the conditions necessary for after-ripening, does not permit one to conclude that the process in the two is the same. In any case after-ripening is not to be attributed to a change in any one condition, but to a series of changes which may vary for each individual case. Dormancy is to be looked upon, perhaps, as a condition of equilibrium in a series of chemical reactions; after- ripening as a displacement of this condition. Why low tempera- tures are effective in causing these changes and why the range of effective temperatures is so narrow are questions still to be answered. Sambucus KiNZEL (15) states that for Sambucus nigra freezing for 2 winters is sufficient to bring only 39 per cent of the seeds to germina- tion. Even longer freezing is necessary for the seeds 'of S. racemosus. Results obtained by the writer in experiments to be described are very similar to those given by Kinzel, and show that in neither case have the conditions necessary for germination been even approximately determined. Nurserymen claim that layering results in almost perfect ger- mination if the seeds are not allowed to become dry between the time of maturing and the time of layering. Air-dry seeds are considered worthless. These statements are in a large measure 296 BOTANICAL GAZETTE [april confirmed by the following experiments, although sufiicient data are not yet available to warrant a final statement. Seeds removed from berries and allowed to dry at room tempera- ture for 2 days failed to germinate within 2 weeks when placed on moist cotton, although they never contained less than 22 per cent of moisture. Fresh seeds on moist cotton kept at 4-6°, 0-2°, or 8° C. have never given more than 20 per cent germination when placed at room temperature or above. Air-dry seeds have given no better results. Although these seeds were kept at the low temperature for not less than 2 months, a longer period may be necessary. The experiments show that failure to germinate is not entirely due to injury resulting from drying, although that may be one of the determining factors. Neither is it to be attributed to inability of air-dry seeds to absorb water, since the quantity taken up in 48 hours by seeds with coats intact is equal to 38. 55 per cent of their air-dry weight, while seeds with coats punctured absorb 39.16 per cent. Air-dry seeds contain approximately 6 per cent of water. The effect of layering is shown by the following experiments, in which the number of seeds used for the 19 16 crop was 1000 and for the 191 7 crop 5000. Two lots of air-dry seeds of the 1916 crop were mixed with soil. One lot was kept at 15-20° C, the other out of doors over winter. In spring the percentages of germination were 8 and 44 respectively. Fresh seed of the 191 7 crop, which had not been permitted to become dry when treated in the same way, gave 51 per cent and 77 per cent respectively. Air-dry seeds of the 1 9 16 crop one year old failed to show any germination. Loss of water seems to be accompanied by a reduction in vitality. Air-dry seeds gathered on October 14, 1916, were treated within 30 days with weak solutions of a large number of acids, bases, and salts. The acids used were mahc, citric, tartaric, acetic, and butyric; the bases, potassium hydroxide, ammonium hydroxide, and sodium hydroxide; and the salts, sodium sulphate, nickel sulphate, ammonium sulphate, zinc sulphate, potassium sulphate, potassium nitrate, sodium nitrate, cobalt nitrate, ammonium nitrate, calcium chloride, sodium chloride, barium chloride, and potassium thiocyanate. The dilutions of the acids were N/200 iqiq] rose— AFTER-RIPENING AND GERMINATION 297 and N/400; of the bases, N/iooo, N/2500, N/5000, and N/io,ooo; and of the salts, N/20 and N/2do. The number of perfect seeds in the cultures varied from 43 to 96. In only 4 cases was the number below 60, and the average was 75. This variation is due to the presence of empty seed coats which could not be distinguished from the perfect seeds until they had taken up a considerable quantity of water. It was later found possible to candle the seeds and thus eliminate the majority of the empty coats. The candling was done by means of an incandescent light supported below a glass plate upon which the seeds were placed. Between the light and the plate was placed a vessel of water to prevent undue heating. The seeds were placed in 20 cc. test tubes containing the solutions and allowed to soak for 24 hours. At the end of that time the solutions were drawn off and the seeds distributed over the moist walls of the test tubes, which were then plugged with cotton and kept at a tempera- ture varying from 4 to 23° C. As soon as the seeds began to show signs of germination, they were removed from the tubes and placed in Petri dishes on moist cotton and kept at room temperature. Germination was slow, in the majority of cases extending over a period of 3 months. In the case of acetic acid, N/400, 58 per cent of the seeds germinated at the end of 176 days. The acids other than acetic showed Httle effect. The length of time over which bases can have any effect must be short, since in dilute solutions they are sooii neutralized by the carbon dioxide of the air and that produced by the seeds. The cultures which showed germinations equal to or better than the checks are Hsted in table X. In order to test the effect of constant low temperature upon seeds soaked in solution of various chemicals, a second set of cultures was prepared in the manner already described and kept at 4-6° C. for 63 days. At the end of that time the tubes were placed at room temperature. To the list of substances used in the pre- ceding experiment were added potassium citrate, potassium tar- trate, potassium acetate, potassium chlorate, ammonium nitrate, potassium iodide, lithium chloride, ammonium chloride, magnesium chloride, sodium nitrite, and dipotassium phosphate, and also hydrochloric acid and sulphuric acid. The concentrations of the mineral acids were N/iooo, N/2500, N/5000, N/io,ooo, and of the 298 BOTANICAL GAZETTE [APRIL salts N/20, N/200, and N/iooo. Germination began 4 days after the cultures were placed at room temperature and continued for 18 days. At the end of that time in practically all of the cultures, in addition to the seeds which had germinated, others were found TABLE X Sambucus seeds in dilutions of acids, bases, and salts; TEMPERATURE 4-23° C. Substance Distilled water. Distilled water. Distilled water. Distilled water. Acetic acid. . . . Acetic acid. . . . Malic acid .... NH4OH NaOH NaOH (NH4)2S04.... (NHJ.SO^.... ZnSO^ KNO,, NaNOi NaNOj C0NO3 KCNS Normality of solution N/200 N/400 N/400 N/2500 N/iooo N/2500 N/20 N/200 N/20 N/200 N/20 N/200 N/200 N/200 Number of seeds 78 8S 68 83 79 72 75 77 88 70 75 67 50 80 46 72 59 66 Percentage of germina- tion 12 10 10 13 18 28 15 18 19 17 28 15 22 30 19 30 71 31 with the seed coat ruptured, but showing no sign of growth. All cultures in which a forcing effect of the solution is indicated by the germination of 20 per cent or more of the seeds are listed in table XI. Out of 13 other substances not given in the table, 5 showed results equal to or better than the average of the checks in at least one dilution. The nitrates and sulphates are again found among the more effective substances. So far as the nitrogen compounds are concerned, these results agree with those of Gassner (10) for seeds of Chloris ciliata. Potassium nitrate, mercuric chloride, and potassium iodide used in connection with alternating temperatures had even less forcing effect than the substances given in table XI. The concentrations used were for potassium nitrate N/20, N/ioo, N/200, N/500, I9I9] ROSE— AFTER-RIPENING AND GERMINATION 299 N/iooo, N/2000; for mercuric chloride N/400, N/iooo, N/2000, N/4000, N/io,ooo; and for potassium iodide N/20, N/ioo, N/500, N/iooo, N/2000. Three sets of cultures were set up in duplicate. TABLE XI Samhucus seeds in acids, bases, and salts Substance HCl H.SO4 H.SO4 NH4OH NH4OH C6H8O9 KNO3 KNOj KNO3 C0NO3 C0NO3 NH4NO3 NH4NO3 NaNOj NaNOj NaNO, NaNO, Na.S04 NiSO^ NiS04 NiS04 (NH4).S04 ZnS04 KCL LiCl NaCl NH4CI NH4CI _. Potassium citrate . Potassium citrate . KCIO3 KI KCNS K.HPO4 KaHP04 Distilled water. . . Distilled water. . . Distilled water . . . Normality of solution N/5000 N/2500 N/ 1 0,000 N/iooo N/5000 N/400 N/20 N/200 N/iooo N/200 N/iooo N/200 N/iooo N/20 N/200 N/200 N/iooo N/200 N/20 N/200 N/iooo N/200 N/20 N/20 N/iooo N/200 N/20 N/200 N/20 N/iooo N/iooo N/20 N/200 N/20 N/iooo Number of seeds 65 66 63 89 96 lOI 78 77 86 92 87 108 56 96 83 72 65 94 84 86 76 85 86 80 80 89 97 84 84 95 81 89 90 83 55 70 85 109 Percentage of germination 17 18 15 23 13 21 28 13 24 4 24 14 14 20 'Z3> 25 9 10 10 7 15 5 41 o o I 2 13 o 7 9 10 9 I o o 8 I Percentage of seeds with ruptured coats 6 10 6 16 8 4 25 18 8 44 23 39 12 18 20 26 18 10 10 34 10 37 I 22 22 37 24 15 21 37 18 2>2, 12 24 34 7 7 14 Total percentage of seeds affected 23 28 21 39 21 25 53 31 32 48 47 53 26 38 53 51 27 20 20 41 25 42 42 22 22 38 26 28 21 44 27 43 21 25 34 7 17 15 One set was kept at 20° C. and a second at 30° C. The third set was kept at 20° C. for 18 hours and at 30° C. for 6 out of every 24 hours. The air in the tubes was changed every second day. The 300 BOTANICAL GAZETTE [april duration of the experiment was 38 days. At the end of that time the only germinations obtained were those in the potassium nitrate, and in no case did these exceed 4 per cent. The seeds in the stronger mercuric chloride solutions were killed. The role played by the coat in the behavior of the seeds has not been determined. Of naked embryos placed on moist cotton 32 per cent developed chlorophyll within a week, formed the hypocotyl arch, and attained a length of 5-10 mm. Naked embryos pre- viously soaked in dilutions of hydrochloric acid and butyric acid and then placed on moist cotton gave no better results. Seeds treated with concentrated sulphuric acid for 4-60 min- utes and then kept under various conditions in regard to Hght, temperature, and oxygen pressure have never given over 20 per cent germination. A slight forcing effect by low concentrations of sulphuric acid was observed on seeds previously treated with con- centrated sulphuric acid for 2-14 minutes and kept in the light at room temperature. Seeds immersed for 5 minutes in water at 40° C. in 55 days gave 25 per cent germination. Reheated at the same temperature for 3 minutes, 33 per cent germinated after 40 days. Longer heating at 40° C. or up to 70° C. gave lower per- centages of germination. Untreated seeds gave no germination in the same length of time. These results emphasize the following facts concerning the conditions necessary for the germination of Samhucus seeds: (i) air-dry seeds with a moisture content of 6 per cent or fresh seeds with a moisture content of 22 per cent will not germinate when placed on a moist substratum at room temperature; (2) this is not due entirely to injury resulting from drying, although that may be one of the determining factors; (3) air-dry seeds are able to absorb water to the extent of approximately 40 per cent of their air-dry weight, indicating that failure to germinate is not due to lack of water ; (4) the effect of chemicals upon air-dry seeds is not marked, a slight forcing effect of several acids, bases, and salts has been observed, among which substances are found nitrates and sulphates ; (5) the role played by the coat in the behavior of the seed has not been fully determined; (6) a sHght forcing effect by low concen- trations of sulphuric acid and by water at 40° C. has been observed; iqiq] rose— AFTER-RIPENING AND GERMINATION 301 (7) seeds remaining in contact with moist soil out of doors over winter gave 77 per cent of germination the next spring; whether this result is due to the low temperature, to certain constituents of the soil, or to a combination of these or other factors one cannot say. The results obtained by Kinzel (15), together with those just summarized, show that as yet the conditions necessary for the germination of Samhucus seeds are not fully determined. To permit the water content of the seeds to fall below an undetermined critical point may lessen their viabihty. However, that some other condition or combination of conditions is responsible for the low percentages of germination must not be overlooked. Kinzel's suggestion that prolonged freezing is necessary should be given due consideration. RUBUS Seed fruits of Ruhiis Idaeus, like the seeds of the 2 species already discussed, fail to germinate when placed on a moist substratum. It was determined that this is not due to an immature condition of the embryo. If the pericarp is left intact all treatments with low concentrations of acids, bases, and salts, immersion in warm water, cold storage, exposure to increased oxygen pressure, or to ether vapor, freezing and thawing, and injection with water under pressure are ineffective. When buried in the soil at 15-20° C. or out of doors over winter, a low percentage of germination takes place if the seeds are kept moist. Two lots of 720 viable seeds buried for 140 days under these conditions gave respectively 40 per cent and 20 per cent germination. Of 2 similar lots of seeds buried in tightly stoppered bottles, one at constant, the other at varying temperatures, none germinated when planted in the soil in the greenhouse. That these results are not due to injury resulting from drying or to inability to absorb water is indicated in table XII. The removal of the endocarp was accomplished by soaking the seeds in con- centrated sulphuric acid for approximately 2 hours. Following this treatment the seeds were washed quickly in a large amount of running water to prevent heating, then immersed in a 5 per cent solution of sodium bicarbonate to neutralize the remaining acid, 302 BOTANICAL GAZETTE [APRIL and finally rinsed in running water for 15 or 20 minutes. The carbonized endocarp was removed by rubbing the treated seed on filter paper. The selection of perfect seeds was now an easy matter. Table XII shows that the water-absorbing power for the seeds with the endocarp removed is 36-37 per cent of their air-dry weight, while that for the seeds with the endocarp intact reaches TABLE XII Water content and water holding capacity of Rubus seeds Condition of seeds Endocarp removed Endocarp removed Endocarp intact. . . Endocarp intact. . . Weight of air-dry seeds in gm. 0.6686 0.6864 I . 0464 2 . 0960 Weight of seeds dried in vacuum at7S°C. 0.5842 o . 600Q 0.9328 1.8680 Percentage of water in air-dried seeds 12.62* 12.45 10.85 10.87 Weight of soaked seeds in gm. 0.9120 0.9414 I 5053 3 .0080 Water Percentage absorbed byi of water air-dry seeds absorbed by in gm. air-dry seeds 0.2434 0-2550 0.4589 0.9120 36.40* 37-15 43-85 43-51 * On basis of air-dry weight. almost 44 per cent of their air-dry weight. From this it follows that the water absorbing power of the endocarp is greater than that of the seed with the endocarp removed. There is no evidence to show that the endocarp possesses any structure which would prevent the water absorbed by it from being passed on to the seed. Although seeds with the 'endocarp intact will not germinate, when that structure is removed by means of the sulphuric acid treat- ment germination takes place within a few days, as is shown in table XIII. The greater amount of the germination takes place between the fourth and tenth days. In seeds germinating after the tenth or twelfth day, growth is usually slow and the seedlings are weak. Failure to secure 100 per cent germination is due to the fact that during the removal of the carbonized endocarp in almost every case the seed coat is ruptured and the endosperm exposed to the attacks of bacteria and fungi. With the carbonized endocarp intact, uncertainty as to the extent to which the acid had penetrated and the inabihty to determine the number of fruits containing viable embryos lead to greater error than that occasioned by the attacks of the bacteria and fungi. I9I9] ROSE— AFTER-RIPENING AND GERMINATION 303 MuLLER (20) has recently pointed out that in various seeds that germinate readily the outward pressure of the contents at the time of rupture was but slightly greater than the breaking strength of the water-saturated coat, and Crocker and Davis (6) have found that seeds of Alisma are held in a dormant condition because the force of the expanding contents is not sufficient to rupture the coats. TABLE XIII Seeds of Rub us Idaens with endocarp removed; 100 SEEDS per culture; temperature 18-23° C. Treated with acid May 4. May 13. May 13. May 13. May 24. May 24. May 24. May 24. May 24'' May 24* June 9. Percentage of germination after 4 days 2 24 6 days 8 days Todays 45 48 46 50 20 22 40 46 57 50 70 52 84 63 61 73 03 83 63 77 61 78 86 •78 86 77 82 86 20 days 96 70 61 55 88 84 80 88 93 89 95 * In darkness. Failure to absorb water is not the hmiting factor, since both reach saturation after about 5 hours' soaking. Two facts indicate that Ruhiis seeds belong in the same class with Alisma. In the first place they germinate readily once the endocarp is removed, and in the second place even with the endocarp intact they absorb water readily. Occasionally ungerminated seeds with the endocarp removed have been found which when examined closely show no break in the coat. This suggests that the inner pectinized layer of the coat may play a part in the delay, either by limiting water or oxygen absorption, or both. As already indicated, the removal of the carbonized endocarp resulted in the rupture of the coat in practically 100 per cent of the seeds. This renders extremely difficult the determination of the part played by that structure. Table XIV shows that the substrata most favorable for germina- tion of naked seed are cotton, filter paper, and quartz sand. An 304 BOTANICAL GAZETTE [APRIL inhibitory effect is shown by garden soil, clay, and greenhouse soil, the effect of the last named being greatest. These soils acidified gave no better results. Calcium carbonate used on filter paper or in sand to neutrahze any acid present in the medium or remaining on the seeds after the sulphuric acid treatment had no inhibitory effect. Glass wool moistened with a boihng water extract or a cold water extract of greenhouse soil cut down the percentage of germi- nation to less than 50. Moreover, the seedHngs were weak, with enlarged and discolored roots. In many cases germination started, but the roots were killed as soon as they came in contact with the substratum. Bone meal had been added to the greenhouse soil and this probably accounts for the injurious effect of the soil and the extracts. As is shown in table XIV, soaking in water for 24 hours TABLE XIV Effect of substratum upon germination of naked seeds of Riibiis Idaeiis; 100 seeds per culture; temperature 18-23° C. Substratum Filter paper Filter paper with CaC03 Quartz sand Quartz sand Quartz sand with 5 per cent CaCOs . Greenhouse soil Greenhouse soil , Greenhouse soil, acid Hot water extract greenhouse soil. . , Cold water extract greenhouse soil . , Garden soil Garden soil, acid Clay Percentage of germination after 6 days SO 31 48 42 43 4 5 5 I 8 days 77 84 80 83 78 22 25 25 9 3 I o days 1 2 days 86 87 83 88 80 10 41 48 90 90 85 89 13 I I 29 10 II 25 days 95 92 92 93 88 25 I I 43 48 29 10 19 previous to planting in garden soil raises the percentage of germina- tion to 55. On the other hand, soaked seeds planted on moist cotton gave 71 per cent as against 72 per cent for unsoaked seeds. Germination was at practically the same rate in the two cases. Seeds planted on 5 per cent agar gave almost as high percentages as those on i per cent. The results given in tables XIV and XV show that the water supply is not the limiting factor. iQig] ROSE— AFTER-RIPENING AND GERMINATION 305 Seeds in the soil are more exposed to attacks by fungi than those on agar or cotton. Previous soaking shortens the time the seeds must He in the soil before germination begins, and hence lessens the chance for infection. Unsoaked seeds placed on moist cotton TABLE XV Effect of water supply xjpon germination of seeds of Ruhus Idaeiis; 100 SEEDS PER culture; TEMPERATURE 20-25° C. Substratum Percentage of germination after 6 days 8 days Todays 1 2 days 16 days per cent agar* . per cent agar*, per cent agar*, per cent agar* . per cent agar*, per cent agar* Soaked 24 hours, then in garden soil. . . Soaked 24 hours, then on moist cotton. Not soaked, on moist cotton 37 22 30 40 35 32 12 43 32 49 5° 37 64 42 52 36 64 66 70 70 62 50 50 "60" 73 6i 47 50 70 71 72 70 62 50 73 60 61 55 71 72 * .Average of 2 duplicate determinations. absorb water easily, hence swell more rapidly than in the soil, and moreover are less liable to infection. Under these conditions soak- ing offers no advantage. Summary General. — Air-dry seeds of Tilia americana, Samhucus canaden- sis, and Ruhus Idaeus do not germinate when placed on a moist substratum at room temperature. In no case does water absorption seem to be the limiting factor. Air-dry seeds planted in the soil over winter give low percentages of germination. Tilia. — Seed coats are not the cause of dormancy, although they may serve to lengthen the dormant period. A state of dormancy exists in the endosperm or embryo, or both. Seeds with coats removed after-ripen at temperatures slightly above freezing. At 0-2° C. seeds after-ripen, but do not germinate. At 4-6° C. both after-ripening and germination take place. Seeds after-ripened at 0-2° C. germinate readily at 10-12° C, but very poorly at room temperature. Once germination has begun growth proceeds best at temperatures above 12° C. 3o6 BOTANICAL GAZETTE [april As after-ripening progresses the hydrogen ion concentration increases, as do also the water holding capacity and the oxidase and catalase activities. The greatest amount of free acid is present in the germinating seeds. Autodigestion of pulverized seeds shows the greatest acid increase in the after-ripened ungerminated seeds. This is probably due to their high lipase activity. Sambucus. — As high as 77 per cent of germination was obtained by layering fresh seeds out of doors over winter. No satisfactory forcing agent has yet been found. A slight forcing effect of several acids, bases, and salts has been observed. The best of these forcing agents are nitrates and sulphates. Although Sambucus seeds are probably injured by drying, that is not the only factor to be considered, since freshly gathered seeds with a moisture content of 22 per cent will not germinate when placed on a moist substratum. As yet it has been impossible to approximate perfect germina- tion, and much still remains to be learned concerning the conditions necessary to reach it. RuBUS. — Dormancy is probably due to the high breaking strength of the endocarp. Seeds treated with concentrated sul- phuric acid for 2 hours, then thoroughly washed, germinate readily on cotton, filter paper, or quartz sand. The optimum temperature for germination lies between 20° and 25° C. Seeds germinate equally well in light or darkness. Naked seeds germinate poorly in soil. This may be due to the action of fungi, bacteria, or to other causes as yet unknown. As a practical method for the germination of Ruhus seeds, if one is not to resort to layering, the writer suggests the following: The seeds should be removed from the pulp as completely as possible. If the berries are crushed and then thrown into water most of the pulp can be floated off. The pulp still clinging to the seeds may be removed by allowing fermentation in water to take place or by treating the seeds with a 5 per cent solution of sodium hydroxide for 15-20 minutes, after which they should be thoroughly washed in running water. It is essential to dry the seeds for at least 24 hours, or the treatment with concentrated sulphuric acid which follows iQig] ROSE— AFTER-RIPENING AND GERMINATION 307 will result in heating. The seeds should be left in the acid for approximately 2 hours. In order to obtain uniform results it is advisable to use a large excess of acid and to prevent the seeds from gathering in clumps or ■layers. Frequent stirring is essential. By rubbing a few of the seeds in the palm of the hand from time to time it is possible to determine when the entire endocarp on a majority of the seeds has been carbonized. When this point is reached the acid should be drained away and the seeds thrown into an excess of cold water. It is advisable to change the water frequently or to put the seeds in running water, where they should be left for at least 15 minutes. When they are removed from the water they should be treated with an excess of a 5 per cent solution of sodium bicarbonate until bubbles cease to rise, after which they may be washed in running water for 15 minutes. In order to remove the carbonized endocarp the seeds may be placed on filter paper and rubbed under the fingers. It is impossible to remove the endocarp if it has been allowed to become dry follow- ing the last washing. The writer is indebted to Dr. William Crocker and Dr. Sophia H. Eckerson for many helpful criticisms and suggestions during the progress of the work. Missouri State Fruit Experiment Station Mountain Grove, Mo. LITERATURE CITED 1. Appleman, C. O., Some observations on catalase. BoT. Gaz. 50: 182-192. 1910. 2. BoLLEY, H. L., The agricultural value of hard coats in alfalfa and clover seed. Paper read before the Association of Seed Analysts, 1910. 3. Bruyning, J. F. On the use of hot water for forcing germination in hard coated seeds. Jour. Landw. 41:86. 1896. 4. BuNZELL H. H., A simplified and inexpensive oxidase apparatus. Jour. Biol. Chem. 17:409-411. 1914. 5. Crocker, W., Mechanics of dormancy. Amer. Jour. Bot. 3:99-120. 1916. 6. Crocker, W., and D.avis, W. E., Delayed germination in the seeds of Alisma Plantago. Bot. Gaz. 58: 285-321. 1914. 7. Davis, W. E., and Rose, R. C, The effects of external conditions upon the after-ripening of the seeds of Crataegus mollis. Box. Gaz. 54:49-62. 191 2. 3o8 BOTANICAL GAZETTE [april 8. EcKERSON, S. H., A physiological and chemical study of after-ripening. Box. Gaz. 55:286-299. 1913. 9. EwART, A. J., On the longevity of seeds. Victoria, Australia. 1908. 10. Gassner, G., Untersuchungen iiber die Wirkungen des Lichtes und des Temperaturwechsels auf die Keimung von Chloris ciliata. Jahr. Hamb. Wiss. Anst. Beih. 3. 29:1-121. 1911. 11. , Einige neue Falle von Keimungsauslosender Wirkung des Stick- stoffverbindungen auf Lichtempfindlicher Samen. Ber. Deutsch. Bot. Gesells. 33: 217. 1917. 12. HiLTNER, L., Die Keimungsverhaltniss der Leguminosen und ihre Be- einfliissung durch Organismenwirkung. Arbeiten. Biol. Abt. Land. Forstw. 3:30. 1902. 13. Honing, J. A., The warm water treatment of the seeds of certain herbaceous and green manure plants that are difficult to germinate. Meded. Deli- Proefstat. Medan. 10: 16-23. 1916; review E.S.R. 36:430. 1917. 14. Jarzymowski, a. von., HartschaUigkeit von Leguminosensamen und ihre Beiseitigung. Inaug. Diss. HaUe. 1905. 15. KiNZEL, W., Frost und Licht als beeinflussende Krafte bei der Samen- keimung. Stuttgart. 1913. 16. Lehmann, E., Zur Keimungs physiologie und biologie von Ranunculus sceleratus und einigen anderen Samen. Ber. Deutsch. Bot. Gesells. 27: 476-494. 1909. 17. , tjber katalytische Lichtwirkung bei der Samenkeimung. Biochem. Zeitschr. 50:388-392. 1913. . 18. Lehmann, E., und Ottenwalder, A., Uber katalytische Wirkung des Lichtes bei der Keimung lichtempfindlicher Samen. Zeitschr. Bot. 5:337-364. 1913. 19. Love, H. H., and Leighty, C. E., Germination of seeds as affected by sulphuric acid treatment. N.Y. (Cornell) Exp. Sta. Bull. 312. 293-336. 1912. 20. MtJLLER, G., Beitrage zur Keimungsphysiologie. Untersuchungen iiber die Sprengung der Samen und Fruchthiillen bei Keimung. Jahrb. Wiss. Bot. 54:529-644. 1914. 21. NoBBE, F., Handbuch der Samenkiinde. 22. Ottenwalder, A. von., Lichtintensitat und Substrat bei der Licht- keimung. Zeitschr. Bot. 6:785-848. 1914. 23. Rose, D. H., A study of delayed germination in economic seeds. Box. Gaz. 59:425-444. 1915. 24. RosTRUP, O., Rept. Danish seed control for 1896-1897. pp. 37. 1898; review E.S.R. 10:53-54. 1898. 25. ToDARO, F., Azione dell acido solforico concentrato su alcuni semi e in particulare sopra i semi duri dell Leguminosae. Staz. Sper. Agric. Ital. 34:613-689. 1901. NOTES ON AMERICAN WILLOWS. IV SPECIES AND VARIETIES OF SECTION LONGIFOLIAE Camillo Schneider In my paper on Mexican willows (Bot. Gaz. 65:22. 1918) I have already dealt with some species of this well marked and entirely American section. In this article I intend to discuss all the members of this interesting group, which is, as M. S. Bebb (1891) and W. W. Rowlee (1900) rightly stated, clearly defined from the other sections of the genus in both the New and the Old World. Andersson (1858) was the first to recognize the close relationship of species like S. sessilijolia Nutt., S. Hindsiana Benth., and S. taxifolia Kth. to S. longifolia Muhl. Unfortunately he misunder- stood most of the species described by Nuttall, and therefore he did not give, even in 1868, a proper analysis of the forms of this section. In 1900 W. W. Rowlee (Bull. Torr. Bot. Club 27:247) made an attempt to rehabilitate all of Nuttall's species, and described several new species and varieties from the southwest, especially from California. His interpretation of Nuttall's species, however, is not free from grave errors owing to the lack of sufficient t^'pe material. Later C. V. Peper studied those types of Nuttall which are preserved in the British Museum, and com- municated his notes to C. R. Ball, who in 191 5 (Bot. Gaz. 60:49) was able to identify S. sessilifolia and S. fluviatilis Nutt. I have not seen the t>'pes in the British Museum, but I have photographs of Nuttall's specimens of S. exigua, S. macrostachya, and S. melanopsis from the Herbarium of the Academy of Science at Philadelphia. Besides this I have also examined a few of Nuttall's willows at the Gray Herbarium, which also contains some cotypes of forms described by Andersson. Photographs and fragments of Andersson's types from the Hookerian Herbarium at Kew are now in possession of the Arnold Arboretum, and Pro- fessor W. W. Rowlee kindly sent me the types of his new species and forms so far as they are preserved in the Herbarium of Cornell 309] [Botanical Gazette, vol. 67 3IO BOTANICAL GAZETTE [april University. I wish to acknowledge here his courteous assistance, and to give the same acknowledgment to the curators of the Herbarium of the Geological Survey of Canada at Ottawa, of the Gray Herbarium, of the Herbarium of the Royal Gardens at Kew, of the Missouri Botanical Garden, of the New York Botanical Garden, of Stanford University, and of the U.S. National Her- barium for the loan of material representing the forms under discussion. For further material I am indebted to Miss Alice East-wood, San Francisco, California, Professor J. K. Henry, Van- couver, B.C., Professor W. L. Jepson, Berkeley, California, Mr. I. M. Johnston, Upland, California, and Mr. J. C. Nelson, Salem, Oregon. I have also been able to go over the material of the Bebb Herbarium at the Field Museum, and am under obligation to Dr. C. F. MiLLSPAUGH for what he has done to further my studies. Sect. LoNGiFOLiAE Audcrssou in Ofv. K. Vet.-Akad. Forh. 15: 116. 1858; for further Kterature see Schneider in BoT. Gaz. 65: 22. 1 91 8. — Frutices mediocres (rariter parvi) vel alti arboresque, ramis densis caespitosis, cortice cinereo vel pl.m. brunnescente, ramulis elongatis virgatis brunneis vel purpureo-brunneis interdum niti- dulis. Folia linearia, lanceolata, vel elliptico-oblonga, denticulata vel integerrima, nervis lateralibus satis distantibus, petiolis vulgo satis brevibus, stipulis saepe deficientibus vulgo parvis lanceolatis denticulatis. Amenta serotina vel primaria coetanea, pl.m. pedunculata vel ramos laterales normaliter foliatos saepe satis longos terminantia, singula vel ad 2-3 aggregata, pl.m. cylindrica, rarius ovalia; bracteae conco lores, flavescentes, deciduae; flores masculi vulgo biglandulosi, diandri, filamentis liberis pilosis; feminei fere semper uniglandulosi, stylis nullis vel brevibus, stig- matibus bifidis laciniis linearibus vel brevibus; ovaria fructusve pilosi vel glabri, subsessiles vel pedicello glandulam usque duplo (rarius magis) superante instructi. As already stated, the Longifoliae is an entirely American group, of which S. taxifolia var. microphylla ranges as far south as Guatemala, while a form of S. longifolia almost reaches the Arctic Circle in the Yukon Territory. From west to east the range of the group extends from the shores of the Pacific to those of the Atlantic, but it is not represented in southeastern United States from central 1919] SCHNEIDER— AMERICAN WILLOWS 311 Virginia to Alabama and Florida. The center of its development is from California to Washington, Montana, and Texas. Among the American willows the Longifoliae occupy an isolated position, and of the willows of the Old World it is difficult to say which can be taken for the nearest relatives of this group. I shall discuss this point later, and I can now only repeat that probably the forms of the sect. Albae Borr. might be regarded as rather closely related genetically to the Longifoliae. In the following key it is recognized' that there are two rather well marked types in the group based on the form of the stigmas. In one, represented by 6". taxifolia and S. sessilifolia, the lobes of the stigma are narrow and elongated, and in the older flowers mostly more or less revolute; while in the other group, the types of which are S. exigiia and S. longifolia, the lobes are shorter and broader, not linear-lanceolate, the whole stigma often being quasi capitate. In some forms of S. longifolia, especially of var. Wheeleri from the northeast, the shape of the stigmas is rather intermediate. In the first group S. taxifolia is well distinguished from 5. sessilifolia and its relatives by the short small aments, the small more or less globose anthers, and the small hnear leaves; while S. sessilifolia and its varieties and 5. fltiviatilis have long cyhndric aments, oblong-ellipsoid anthers, and longer, broader leaves. In the second group it is more difficult to separate the species because the main characters, glabrousness or pubescence of the ovaries and leaves, are more liable to variation. S. melanopsis .with var. Bolanderiana represents a rather well marked t^-pe with glabrous ovaries, but in S. exigua as well as in S. longifolia we meet with forms of which the ovaries vary from densely pubescent to entirely glabrous. The ' It seems to be of interest to quote Bebb's opinion as to the possibility of a taxo- nomic arrangement of the forms of this section (Box. Gaz. 16:104. 1891): "Clearly marked as are the outer limits of the group it presents no Imes of cleavage within by which it can be satisfactorily divided. No natural characters are found to coincide with such assumed distinctions, for instance, the 'Hnear lobes of the stigma,' made promment in the attempt to separate S. sessilifolia. Each portion after subdivision remains as heterogeneous as was before the aggregate group. It may be possible, by emphasizing first one character and then another, as these are found to predominate in the different forms, to designate a number of subspecies and varieties; but so bewildering and intangible is the reticulated intergradmg that the difficulty of segre- gation seems only to be heightened by every fresh acquisition of the material." 312 BOTANICAL GAZETTE [april pubescence of the leaves too is very changeable, and only in con- nection with other characters can it be used to separate certain species and varieties. Clavis specierum Amenta brevia, mascula 5-13 mm. longa et circ. 8 mm. crassa, feminea satis pauciflora, fructifera haud ultra 2:1.2 cm. magna; antherae minimae pl.m. globosae vel subglobosae, haud vel pauUo longiores quam latae; stigmatum lobi lineares vel lineari-lanceolati, vulgo 4-6plo longiores quam lati; stylus nullus vel subnullus; ovaria sessiHa vel brevissime pedicellata; bracteae vulgo satis late obovato-rhombicae, pl.m. acutae, praesertim extus satis dense villosae; folia minima vel parva, linearia vel lineari-lanceolata, 10-30:1.5-3.5 mm. magna, subtus semper pl.m. sericea, margine breviter denticulata vel subintegerrima , . . . i. S. taxifolia Amenta longiora vel antherae ellipticae, circ. i|-2plo longiores quam latae vel foHa majora. Stigmatum lobi lineares vel lineari-lanceolati, elongati, vulgo 4-5plo longiores quam lati, adulti pl.m. revoluti, stylo satis distincto iis breviore vel brevissimo fere semper bifido suffulti vel pl.m. sessiles; ovaria (saltem juniora) distincte sericea vel sericeo-villosula ; folia novella semper utrinque pl.m. dense sericea vel sericeo-villosa. Ramuli hornotini dense, etiam annotini pl.m. sericeo-villosi vel tomentelli; folia etiam adulta utrinque concoloria, can- escentia, canoviridia vel viridescentia, semper pl.m. sericea vel sericeo-villosa, nervis primariss vix vel non visibilibus; ovaria semper satis dense sericeo-pilosa, sessilia vel subsessilia, pedicello fructuum quam glandula plus quam 2plo breviore; bracteae rarius extus versus apicem glabrescentes (confer etiam 4. S. Parishianam). Folia ramulorum fertilium lineari- vel anguste lanceolata, fere semper distincte integerrima, etiam majora vix ad 8 mm. lata, apice pl.m. sensim acuminata, basi acuta, in petiolum satis distinctum attenuata, stipulis fere semper nullis, vel folia maxima majora, 6-8 cm. longa et ultra 8 mm. lata; amenta mascula i . 5-3 cm. longa et 5-6 (rarius 8) mm. iqiq] SCHNEIDER— AMERICAN WILLOWS 313 crassa, feminea fructifera 2-4 (-6): 0.8-1 cm. magna, ovaria (fructusque) sessilia vel subsessilia. Folia fere semper lineari- vel anguste lanceolata et vulgo integerrima, stigmata semper satis elongata et pl.m. revo- luta 2b. S. sessilifolia var. Hindsiana Folia fere semper remote denticulata, interdum late lanceolata; stigmata breviora, paullo curvata et magis sessilia 2c. S. sessilifolia var. leucodendroides Folia ramulorum fertilium (anguste vel) late lanceolata vel elliptico-lanceolata, majora 8-1 5 (-17) mm. lata, saepe (saltem ad apicem) pl.m. distincte subspinuloso-denticulata, interdum paene sessilia, stipulis saepe pl.m. evolutis; amenta mascula 3-4.5 cm. longa et circ. 7 mm. crassa, feminea fructifera 4-6(-io) cm.: 8-10 mm. magna, ovaria (fructusque) subsessilia vel brevissime pedicellata 2. S. sessilifolia Ramuli tantum novelli satis dense sericeo-tomentelli, jam hornotini glabrescentes vel glabriusculi vel folia adultiora satis glabra subdiscoloria, vel ovaria fructusque glabri vel subglabri (confer etiam var. Wheeleri sub 8. 5. longifolia). Folia anguste lanceolata ellipticave, interdum oblanceolata, apice pl.m. acuminata, basi acuta, distincte petiolata, stipulis saepe evolutis, adultiora superne intense viridia, subtus interdum subglaucescentia, satis glabrescentia vel tenuissime sericeo-pilosa, nervis etiam secundariis utrinque pl.m. visibilibus, ramulorum fertilium 7-14 mm. lata; ovaria initio pl.m. sericea vel sericeo-villosa, matura vulgo fere tota glabrescentia, subsessilia, pedicello fructuum glandula sicca interdum subaequilongo, bracteae fere sem- per extus versus apicem glabrescentes interdum basi excepta glabra 3. S. fluviatilis Folia anguste linearia ad lineari-lanceolata, i . 5-5 (-8) mm. lata, utrinque pl.m. dense adpresse sericea; ovaria pl.m. sericea vel fere glabra, fructus partim pilosi vel glabri sed pedicello brevissimo vulgo piloso 4. 5. Parishiana Stigmatum lobi lanceolati vel elliptici, satis breves, saepissime 2-3plo longiores quam lati, adulti ut videtur nunquam distincte 314 BOTANICAL GAZETTE [april revoluti, stylo nullo vel brevissimo non bifido suffulti, ovaria sericea vel glabra, subsessilia vel fructus pedicello glandulam interdum duplo superante instruct!; folia ramulorum fertilium pl.m. dense sericea vel glabra. Flores feminei glandulis 2 (dorsali interdum minima) instruct! 6c. S. exigua var. nevadensis Flores feminei glandula tantum ventrali instruct!. Glandulae florum masculorum 2 (ventralis et dorsalis). Ovaria etiam juvenilia glaberrima. Folia tantum valde juvenilia pl.m. distincte sericea vel ab initio pl.m. glabra vel tenuiter pilosa pilis saepe tantum sub lente visibilibus, utrinque concoloria vel superne viridia, subtus pallidiora, saepe pl.m. glauces- centia, nervis lateralibus secundariisquepl.m.prominulis. Amenta fructifera valde densa, fructibus condensis breviter conicis pedicello subnullo vel satis brevi glandulam vix superante instructis, bracteae florum vulgo satis obovatae et truncatae; folia subtus fere semper pl.m. pallidiora vel glaucescentia, ramulorum sterilium satis late vel elliptico-lanceolata vel ob- lanceolata, rarius lineari-lanceolata. Fructus 4.5-5.5 mm. longi (pedicello brevi ex- cluso), amenta fructifera circ. 8-9 mm. crassa; folia ramulorum fertilium 3 : o . 4 ad 8 : i . 2, interdum ad 6.5:1.5 cm. magna, citissime glabrescentia vel pilis difficile visibilibus praedita (rarius initio satis dense adpresse argyraceo-sericea) , satis distanter et breviter denticulata vel pl.m. integerrima; ramuli hornotini vulgo cito glabrescentes .7.5. melanopsis Fructus ad 6. 5 mm. longi, amenta fructifera ad i . 2 cm. crassa; folia ramulorum fertilium ad 9: i . 5 vel 17:1.7 cm. longa vel distinctius pilosa et denticu- lata vel ramuK hornotini magis pilosi 7b. S. melanopsis var. Bolanderiana Amenta fructifera satis laxiflora fructibus separatis vel ovariis fructibusque longius conico-rostratis et pedicello distincto glandulam saepe duplo superante 1919] SCHNEIDER—AMERICAN WILLOWS 315 instructis; bracteae florum vulgo oblongiores acu- tioresque; folia utrinque concoloria, pl.m. lineari- lanceolata vel linearia vel anguste lanceolata et satis distincte subdensius denticulata. Fructus vix ultra 6 mm. (pedicello excluso) longi; folia anguste linearia, 2-4 mm. lata, venis laterali- bus vix visibilibus magis impressis quam prominu- lis 6d. S. exigua var. tenenima Fructus (5-) 7-9 mm. longi; folia interdum paullo latiora, venis lateralibus pl.m. distincte prominulis 8b. S. longifolia var. pedicellata Folia etiam adulta pl.m. sericea, utrinque (praecipue subtus) canescentia, venis lateralibus haud vel vix prominulis, ramulorum fertilium pl.m. lineari-lanceolata, integerrima vel satis distincte remote breviter den- ticulata; amenta fructifera pl.m. densiflora, fructibus pedicello glandulam saepe duplo superante instructis 6b. S. exigua v^ar. stenophylla Ovaria semper distincte sed interdum tantum pro parte sericeo-villosa vel sericea, fructus interdum fere vel omnino glabrescentes, subsessiles vel pedicello quam glandula pl.m. breviore suffulti, rarius distincte sessiles. Folia ramulorum fertilium pl.m. integerrima vel tantum ad apicem parce et saepe indistincte denticulata, utrin- que pl.m. canescentia, satis dense sericea vel etiam adulta non distincte glabrescentia et viridia venis etiam primariis vix vel paullo prominulis; fructus satis breviter conico-rostrati, amenta fructifera densa. Folia etiam semiadulta utrinque (praesertim subtus) dense argenteo-sericeo-villosula, ramulorum fertilium saepe satis lanceolata, \'ulgo ad 8-10 mm. lata; ovaria juvenilia dense et longe sericea vel sericeo-villosula; fructus ellipsoideo-conici, 5-6.5 mm. longi (confer etiam 6c. 5. exiguam var. luteo-sericeam) 5. 5. argophylla Folia minus dense, saepe tenuiter breviter adpresse sericea, ramulorum fertilium linearia vel lineari- 3i6 BOTANICAL GAZETTE [april lanceolata, vulgo vix ultra 8 mm. lata vel ovaria angustiora apice magis capitata (incrassata) et fructus magis elongati 6. S. exigiia Folia ramulorum fertilium pl.m. distincte denticulata, vulgo cito utrinque viridescentia et glabrescentia, adulta intense laete viridia et glabra (vel in var. W heeler i utrinque pl.m. sericea), nervis etiam secundariis utrin- que pl.m. prominulis; fructus magis elongati et rostrati; amenta fructifera pl.m. laxiflora (si amenta sunt valde densiflora et ovaria parce vel partim pilosa conf. etiam S. melanopsidem var. Bolanderianam) . . .8. S. longifolia Glandula florum masculorum tantum una ventralis (rarius dorsalis minima adest); amenta feminea saltem novella ovariis dense albo-sericeo-villosis subsessilibus pl.m, mi- cantia; glandula satis lata; folia ramulorum fertilium pl.m. linearia, 4-8 cm. longa et 1-5 mm. lata, ut in 5. longifolia dentata et nervata 8c. S. longifolia var. angustissima Enumeratio specierum 1. S. TAXiFOLiA Kunth in Humb. and Bonpl., Nov. Gen. PL 2:18. 1817; Sargent, Silva N. Am. 9:129. pi. 476. 1896; Man. Trees N. Am. 175. fig. 147. 1905; Sudworth, Nomencl. Arb. Fl. U.S. 123. 1897, pro parte; Britton and Shafer, N. Am. Trees 202. fig. 164. 1908; for further literature and synonymy see Schneider in BoT. Gaz. 65:23. 1918. — At present I have nothing to add to what I have already stated {I.e.) with regard to this species and its var. microphylla (Schl. and Cham.) Schn. There are several forms which look rather similar to S. taxifolia, but differ in the shape of the anthers and some other respects. I shall discuss them under S. exigua. 2. S. SESSiLiFOLiA Nutt. N. Am. Sylva 1:68. 1843,^ reprint 1852; Anders, in Ofv. K. Vet.-Akad. Forh. 15:116. 1858; in Proc. Amer. Acad. 4:56 (Sal. Bor.-Am. 10). 1858; in Walp., Ann. Bot. 5:746. 1858, incl. var. villosa; in K. Sv. Vet.-Akad. Handl. 6:55. pi. 4. fig. 36 (Monogr. Salic). 1867; in DC. Prodr. 16^:214. 1868; ^ Nuttall's vol. I was issued in 2 parts; part i in 1842, containing pp. 1-54; while part 2, pp. 57-136, including the Salices, appeared in 1843. 1919] SCHNEIDER— AMERICAN WILLOWS 317 Bebb in Watson, Bot. Calif. 2:85. 1879,^ pro parte et exclud. synon.; Sargent, Rep. For. N. Am. loth Census U.S. 9:168. 1884, pro parte et excl. var.; Silva N. Am. 9: 127. 1896, pro parte; Sud- worth in Bull. U.S. Dept. Agric. Div. For. 14:122 (Nomencl. Arb. Fl.). 1897, pro parte; For. Trees Pac. Slope 223. 1908, pro parte; Eastwood, Handb. Trees Calif. 37. 1905, pro parte; Britton and Shafer, N. Am. Trees 196. 1908, pro parte minima; Howell, Fl. Northw. Am. 1:618. 1902; Piper in Contr. U.S. Nat. Herb. 11: 213 (Fl. State Wash.). 1906; Ball in Box. Gaz. 60:49. ^g. 2. 1915; in Piper and Beattie, Fl. Northw. Coast 115. 191 5; Henry, Fl. S. Br. Col. 96. 1915; Rydberg, Fl. Rocky Mts. 192. 1917, pro parte. —S. sessilifolia var. villosa And. in K. Sv. I.e. 56 et Prodr. I.e. 214. — S. macrostachya Nutt., N. Am. Sylva 1:72. 1843; Howell, Fl. I.e. 619, pro parte; Rowlee in Bull. Torr. Bot. Club 27:250. 1900, pro parte et excl. var.; Rydberg, Fl. I.e. 192. — S. macrostachya var. Cusickii Rowlee, in Bull. I.e. pi. 9, fig. 5, sine descr. — S. longijolia var. sessilifolia Jones, Willow Fam. 24. 1908. Type locality. — Oregon, "on the rocky borders of the Oregon [Columbia] at the confluence of the Wahlamet" [Willamette]. Range: from western Oregon, Douglas County, along the Umpqua and Willamette River to the Columbia and Lewis rivers in Washington, thence again in northern Washing- ton, Whatcom County, and southwestern British Columbia. S. sessilifolia was the only one of Nuttall's species which has been correctly interpreted by Andersson, who cites for the type LyalVs specimens from the Sumass Prairie, of which the male is no. 78 and the female no. 31 in Herb. K.-* They were collected in 1858 "near the 49th parallel of lat." In the herbarium Andersson first had named the specimen 5. Grayi, but this name has never been published. For his var. villosa the type was collected by Lohh in 1852 in Oregon, bearing the no. 218 in Herb. K. LobVs and LyalVs 3 Bebb's treatment of the Californian Salices in Watson's Flora was published separately in 1879. ^ Besides the abbreviations mentioned in Box. Gaz. 65 : 9 and 66: 121, the following will be used: Cal., Herbarium of the CaHfornia Academy of Science; K., Kew Her- barium; Jeps., Herbarium of Professor W. L. Jepson, Berkeley, Cal.; N.E., Herbarium of the New England Botanical Club; P., Herbarium of the Academy of Science at Philadelphia, Pa.; Reno, Herbarium of the Nevada Agric. Exper. Station, Reno, Nev.; St., Herbarium of the Leland Stanford University. 3i8 BOTANICAL GAZETTE [april plants are exactly alike. Nuttall himself gave an excellent description, and it is rather astonishing that the species could have ever been misunderstood. He described another species, however, S. macrostachya, "from the banks of the Oregon" [Columbia], of which there is a sterile(!) cotype in Herb. P. and a branchlet with an old fruiting ament in Herb. G. According to some remnants the style and the stigmas are exactly as in S. sessilifolia, and there is no difference in the shape and pubescence of the leaves. Judging by Nuttall's statement "amentis longissimis praecocibus," he had before him a very early flowering state, and the fragment in G. shows an old, almost sessile, long, fruiting ament which naturally looks very different from the normal late flowering form with the aments at the top of rather long leafy branchlets. The sheet in P. also contains a female branch of which I do not know the origin, because it is only partly represented in the photograph. It seems to me that this branchlet belongs to the true 5. argophylla Nutt., which has a similar foliage and pubescence but shorter stigmas, looking more or less intermediate between 5'. sessilifolia and S. exigua. I shall deal with it later. It has been mostly taken hitherto for S. macrostachya. Ball (191 5) also referred specimens from Cahfornia to 5. sessilifolia, but those forms I take for var. Hindsiana. Sterile specimens collected by /. G. Jack in Oregon, Josephine County, Grant's Pass, August 23, 1904, and at the same locality and time by A. Rehder, seem to me to belong rather to S. argophylla than to S. sessilifolia. In British Columbia, West- minster County, New Westminster, banks of Fraser River, /. K. Henry collected good material on June 24, 191 2, and May 9 and September 25, 1914 (m., fr., st.; Cal.). The largest leaves I have seen measure up to 9:2 cm. 2b. S. SESSILIFOLIA var. Hindsiana And. in Ofv. K. Vet.-Akad. Forh. 15:117. 1858; in Proc. Am. Acad. 4:56 (Sal. Bor.-Am. 11). 1858; in Walp., Rep. Bot. 5:746. 1858; Bebb in Watson, Bot. Calif. 2:85. 1879; Sargent, Rep. For. N. Am. loth Census U.S. 9:169. 1884, excl. synon. var. tenuifolia; Eastwood, Handb. Trees Calif. 38. 1905.— 5. Hindsiana Benth. PI. Hartw. 335. 1857; Torrey in Pacif. R.R. Rep. 4^:138. 1857; Newberry in Pacif. iqiq] SCHNEIDER— AMERICAN WILLOWS 319 R.R. Rep. 6^:89. 1857; And. in K. Vet.-Acad. Handl. 6:56 (Mon. Salic). 1867, excl. pi. 4, fig. j/ et var. — S. longijolia var. argyro- phylla f. angiistissima And. I.e. 55; in DC. Prodr. 16^: 214. 1868. sec. specim. Fremontii. — S. longijolia Greene, Man. Bot. S. Fran. Bay 299. 1894, pro parte max. — 6*. sessilijolia Sarg., Silva N. Am. 9:127. 1896, pro parte; Jepson, Fl. Calif. 339. 1909, pro parte max.; in Mem. Univ. Calif. 2:178 (Silva Calif.). 1910, prop arte max.; Ball in Box. Gaz. 60:51. 1915, pro parte. Type locality. — California, "ad ripas fluvii Sacramento." Range: central California to southwestern Oregon. Of S. Hindsiana I have seen a photograph of the type (K.) and cot>^es (G., N.) collected by Hartweg, which are all perfectly identical. It is closely related to t>^ical S. sessilijolia, from which it differs chiefly by its more Unear or narrowly lanceolate and almost always entire leaves, which are more or less distinctly petioled, and by its usually smaller and thinner aments. If it were not for some specimens which seem to combine var. Hindsiana with the northern S. sessilijolia, and others that I can hardly distinguish from the southern var. leucodendroides (for instance a vigorous sterile speci- men from Yolo County, mouth of Buckeye Creek, Ig. R. Stinchfield, no. 334; St.), I should take it for a distinct species. A closer study of those forms in the field is certainly needed. There seems to occur, a form with almost glabrate ovaries, judging by a specimen collected by R. S. Ferris in Colusa County, Sycamore Slough, April 17, 191 7 (no. 619, m., f.; St.). It is other- wise rather topical var. Hindsiana and needs further study. The range of this variety extends to Jackson County in south- ern Oregon iWalpole, no. 255; Applegate, nos. 624 and 2198) in the north, and to Monterey^ and Kern counties in California in the south, but the southern forms like Piper's no. 6406 from Bakersfield come very near var. leucodendroides. 2C. S. SESSiLiFOLiA var. LEUCODENDROIDES Schneider in Box. Gaz. 65 : 26. 1918. — S. macrostachya leucodendroides Rowlee in Bull. 5 From this county is Brewer's no. 544, which came from the Nacimiento or Naci- mento River or Creek, not " Narsismente " or "Nasismento" River, as the name is spelled by Rowlee and Ball according to the label in C. It is a male specimen with leaves much like var. leucodendroides , to which it may belong after all. 320 BOTANICAL GAZETTE [april Torr. Bot. Club 27:250. pi. g.fig. 6. 1900; Abrams, Fl. Los Angeles suppl. ed. 102. 191 1. — S. integrifolia var. leucodendroides Rowl., I.e. sphalmate in textu. — S. argophylla RowL, I.e., pro parte. — S. exigua var. virens Row!., I.e. 256, pro parte. — S. sessilijolia East- wood, Handb. Trees Calif. 37. 1905, pro parte; Britton and Shafer, N. ' Am. Trees 196. 1908, pro parte; Jepson in Mem. Univ. Calif. 2 : 178 (Silva Calif.). 1910, pro parte.— 5. macrostachya Abrams, Fl. I.e. loi, non Nutt. — Rowlee cites 3 specimens from southern California under his variety, namely Parish's nos. 2134, 2040, and 640. The last number is quoted by him also under his S. argo- phylla. It belongs to var. leucodendroides. No. 2134 represents an early flowering state of the male plant with small leaves and short peduncles of the catkins which measure up to 2 : o . 9 cm. The bracts are almost glabrate and often somewhat denticulate at apex, a fact we may also observe in other forms of S. sessilijolia. No. 2040, in my opinion, can be regarded as the typical var. leucoden- droides, which seems to differ from var. Hindsiana chiefly in its comparatively longer and broader, very often distinctly denticulate leaves (with fine distant teeth), measuring usually from 7:1.2 to 13:1.8-2 cm. (in var. Hindsiana the corresponding entire leaves are about 3-10 cm. long and 3-10 mm. wide, while in the typical sessilijolia they measure from 5:0.8-1 to 8:3 cm., being distinctly denticulate with fine linear teeth), and by its stigmas, which usually are almost sessile and somewhat shorter and broader than in var. typica or var. Hindsiana. Some plants look almost like hybrids with 5. Parishiana or the form of S. exigua from southern California. I can but repeat that a proper understanding of all these forms can only be gained by a careful study of them in the field. See also my remarks under 5. Parishiana and S. argophylla. I give an enumeration of the specimens I am inclined to refer to var. leucodendroides, and I should be glad to receive some informa- tion by collectors who visit these locahties as to the different forms of willows growing together there. Specimens examined. — San Diego County: Santa Ysabel Creek, May 1893, R. D. Alderson (no. 700, f. ; Cor. ; ovariis parce sericeis; cited by Rowlee under S. exigua virens); Mountain Spring, May 10, 1894, E. A. Mearns (no. iQig] SCHNEIDER— AMERICAN WILLOWS 321 3040, m.; W.). — Riverside County: Santa Ana River, N.W. of Corona, very common, 150m., May 26, 1918, /. M. Johnston (no. 1994, m., f.; A.); same place, 180 m., very common along river banks, June 9, 1917, Crawford and 7o//«5/o» (no. 1244, m.; A.; St.); Temescal Canyon, along a dry wash, 400 m., May 30, 1918, I. M. Johnston (no. 2017, fr. ; A.); same river, near Riverside, May 1888, 5. B. Parish (no. 2040, f., type; C, Cor.); San Jacinto, along San Jacinto River, March 31, 1896, A. J. McClatchie (m., f.; N.; early flowering form, somewhat uncertain); eastern base of San Jacinto Mts., along the borders of the Colorado Desert, June 1901, H. M. Hall (no. 2105, m., f.; M.; ovariis laxe sericeis, stigmatibus mediocribus) ; San Jacinto River Canyon, gravelly ground along the river, common. May 12, 1918, Durand and Street (no. 23, f.; A.). — Orange County: Santa Ana River, June 1880, 5'. B. Parish (m.; A., M.; "12 ft. high"). — Los Angeles County: Los Angeles, 1879, /. C. Nevin (m.; G.; fragment); San Gabriel River at El Monte, common along river, 90 m., ]\Iay 13, 1917, /. M. Johnston (no. 1242, m., f.) ; same place, July 7, 1887, Tracy and Evans (no. 383, m.; N.); San Gabriel Mts., San Antonio Canyon, 1450 m., July 9, 1918, F. G. Peirson (no. 14, m.; Jeps.); canyon near San Rafael, March 31, 1888, H. E. Hasse (no. 3801, f.; N.; var. Hindsianae valde simihs); sandy flat along the Los Angeles River, May 30, 1888, H. E. Hassc (no. 4092, m., f. ; N.; stigmata iis S. cxigiiae satis similia); Los Angeles River bottom, near Los Angeles, September 9, igiy, F. Grimmel (fr.; St.). — San Bernardino County: San Bernardino Valley, dry sandy banks of Lyth Creek, in a large thicket, April 4, 1891, S. B. Parish (no. 2134, m. syn- type; Cor. and C, both named S. macrostachya by Rowlee; "about 4 ft. high"); Lyth Creek Wash, damp land, alt. circ. 300 m.. May 2, 1917, 5. B. Parish (no. 11134, f., fr.; A.; fructibus satis glabris); vicinity of San Ber- nardino, alt. 300-750 m., April 8, 1899, 5. B. Parish (no. 4591, m.; St.; 4592, f., fr.; St.; the last number represents a small-leaved form much resem- bling 5. taxi folia as well as var. Hindsiana; needs further observation) ; April 13, 1903, 5. B. Parish (no. 5197, f.; St.; same small-leaved form); May 15, 1901, S. B. Parish (nos. 4786, m., 4787, f., fr.; N., St.; structura florum paullo ad S. exiguam vergens); March 1881, S. B. and W. F. Parish (no. 640, m., fr.; A.; var. Hindsianae satis similis, sed stigmatibus subbrevioribus, in C. magis typica); February 20, 1881, W. G. Wright (nos. 10, 11, m., 12, f.; C; "small bush 6-10 ft.") ; March i and 14, 1881, W. G. Wright (nos. 6, m., 7, f.; C; early flowering specimens with short aments which look rather different); Colton, April 28, 1882, M. E. Jones (m., fr.; A.); Waterman Canyon, August 1900, Shaw and Illingworth (no. 4, m.; St.; amentis brevibus, antheris parvis, sed foHis normalibus); Keenbrook, Kajon Pass, May 30, 1901, S. B. Parish (no. 4930, f., m.; St.; very much like 5. exigna, but the female flowers more like those of var. leucodendroides) ; same Pass, July 6, 1908, LeRoy Abrams and L. E. McGregor (no. 694, f.; St.); Cucamonga Canyon, small colony on bed of a small side canyon, alt. 900 m., May 27, 191 7, /. M. Johnston (no. 32 2 BOTANICAL GAZETTE [april 1241,'^ m.; St.)- — Ventura County: Ventura, along beach, April 17, 1916, A. Eastwood (no. 5034, m., 5035, f. ; Cal.). — Santa Barbara County: Santa Ynez River, alt. 600 m., May 1894, C. Franccschi (m.; A.; quasi ad var. Hindsianam transiens). — Tulare County: shores of Kern River, Peppermint Valley, alt. 1440m., July 16, 1895, W. R. Dudley (no. 779, m.; St.); gravelly bars of Kaweah River at Three Rivers, July 20, 1900, W. R. Dudley (no. 2703, St.; St.); Three Rivers, near Brittons, June 15, 1902, W. R. Dudley (m., fr.; St.; all these forms of Tulare County come near var. Hindsiana; the fruiting aments of the last specimen measure up to 6:1 cm.). See also Brewer's no. 544 mentioned in the preceding note. Specimens from Kern County, Bakersfield, September 28, 1910, E. M. McGregor (no. 13, m.; St.), look much like S. exigua and need further observa- tion. There is a specimen from Santa Barbara County, Ojai, Cliff Glen, March 15, f., April 3, 1896, m., F. W. Hubby (no. 56; Cor.), of which the leaves much resemble 5. taxifolia, but those of the more vigorous shoots seem to become larger. The female flowers have 2 glands, and the stigmas are rather short but agree with those of some forms I have referred to var. leucodendroides. I am not quite sure about this specimen, but I strongly suspect that it is a form of var. leucodendroides grown in a very arid position. It is similar to Parish's nos. 4591, 4592 already mentioned. 3. S. FLUViATiLis Nuttall, N. Am. Sylva 1:73. 1^43; B^H in Box. Gaz. 60:52. ^g. J. 1915; in Piper and Beattie, Fl. Northwest Coast 114. 1915. — S. sessilijolia Sargent, Silva N. Am. 9:127. pi. 475. 1896, pro parte, non Nuttall; Rowlee in Bull. Torr. Bot. Club. 27:250. pi. g. fig. 8. 1900; Howell, Fl. Northwest. Am. 1:618. 1902, pro parte; Sudworth, For. Trees Pacif. 223. figs, gi, gz. 1908, pro parte; Rydberg, Fl. Rocky Mts. 192. 1917, pro parte. — Nuttall says: ''This species lines the immediate border of the Oregon [Columbia] a little below its confluence with the Wahlamet" 'No. 1243 of the same collector from Red Hill, near Upland, April 28, 1917, apparently represents the female form of the same willow. Mr. Johnston kindly sent me the following note regarding this number: "1243 from Cucamonga Canyon. Small colonies of this willow occur in scattered localities in the lower canyons of the San Antonio Mountains; although common in the valley it is uncommon in the mountains. 1 243 came from one of these isolated colonies, and from absolute knowl- edge I know that no other colony of this or any other Longifoliae occurs within 3 miles. The associated Salix spp. were S. laevigata and S. lasiolepis. Nothing like 5. exigua occurs for miles. This is by no possibility a hybrid." Judging by the stigmas this form is more closely related to S. exigua than to S. sessilifolia. The forms of this part of S. CaHfornia need a special study, and it is almost impossible to express a definite opinion on them as long as S. Parishiana and S. exigua and its varieties are not yet properly understood. 1919] SCHNEIDER— AMERICAN WILLOWS 323 [Willamette], and "we met this species likewise on the bank of the Lewis River of the Shoshonee." The first locality has been visited by Ball, and I follow him in his interpretation of this species. Unfortunately no type specimen exists, and from Nuttall's statement that "the germ is smooth, with 4 sessile stigmas" I believe that he had partly S. melanopsis before him from the second locality quoted, which is on the Snake River in western Idaho. At present the true S. fluviatilis is only known from " the lower part of the Willamette River and adjacent Columbia River" in Oregon, Multnomah County, ranging eastward to Wasco County, The Dalles, where Ball collected it on June 24, 1915 (nos. 1997, m., 1998, 1999, f., 2000, androgyn., 2005, fr., 2007, m., 2015, fr.; C, G.). It has also been found on the opposite bank of the Columbia, in Klickitat County, Wash., by Suksdorf, April 23, May 31, 1881 (no. 6, f., m.; C. [7876]). Other specimens of Ball's (nos. 1857, 1858, 1859; fr. adult.; G.) from northeastern Utah, Cache County, Logan Canyon, above Logan, in my opinion are somewhat uncer- tain. They suggest certain forms of S. melanopsis var. Bolan- deriana, and indeed S. fluviatilis seems in some respects to be quasi intermediate between S. sessilifolia and 5. melanopsis. Ball himself says: "The species is quite different from the true sessili- folia. It is closely related to S. melanopsis Nutt." But he also states: "The style and stigmas indeed are very similar to those of true 5. sessilifolia. " In fact, specimens collected by Ball on the shores of the Umpqua River, near Roseburg, Oregon (no. 1961, 1962, f., fr.; G.), and distributed by him as " ?S. Bolanderiana {X sessilifolia)," are somewhat similar to S . fluviatilis , which, how- ever, seems to be a good species of a very local distribution, quite different in the structure of the male flowers from that of the melanopsis group. 4. S. Parishiana Rowlee in Bull. Torr. Bot. Club 27:249. pi. g. fig. 3. 1900; Abrams, Fl. Los Angeles suppl. ed. loi. 1911. — S. sessilifolia Jepson, Fl. Cahf. 339. 1909, pro parte, non Nutt.; in Mem. Univ. Cahf. 2 : 178 (Silva Calif.) 1910, pro parte.— 5. longifolia var. argyrophylla Jeps. in Mem. I.e. pro parte. —5. argophylla Ahrsims, I.e. 102, pro parte. — This is a peculiar and rather obscure species of which Rowlee has given a somewhat unsatisfactory description. 324 BOTANICAL GAZETTE [april As type is cited Hobby's (recte Frank Hubby'') nos. 54, 55 from Matilija Canyon, Ventura County (not in San Bernardino County or, as is written on the label of the type no. 54 before me, Santa Bar- bara Co.). Besides this there is given on the label for the female specimen "Chff Glen," and for the male ''Ojai Springs," locahties near Matilija. The flowers are young, and the ovaries are not ''densely villous" but, at least partly in no. 55, glabrescent toward the apex and base, and rather silky pubescent. The specimens could easily be taken for S. exigua were it not for the fact that the lobes of the stigmas are narrower, about 3 times as long as thick, and the styles distinct but short. Rowlee also cites a specimen collected by Coville and Funston (no. 263) at Spring Valley, Inyo County, but this is sterile. Only in Herb. W. I have found a few fruits attached to it which look much like those of S. exigua. I find it difficult to express a definite opinion on S. Parishiana, but I wish to enumerate the following specimens which may repre- sent the same form. It looks intermediate between S. exigua (of southern California) and 5. sessilifolia var. leucodendroides, and similar forms seem to occur in the region where var. Hindsiana reaches the southern limit of its range. The question whether we have to do with forms of hybrid origin or with a distinct species can only be solved by careful observation in the field. See also the indications given in the key. Specimens examined. — California: Venturia County: Matilija Canyon (see the remarks given in the preceding text), April 3, 1896, F. W. Hubby (no. 54, m. and f. types; Cor.), April 19, 1896, F. W. Hubby (no. 55, fr.; Cor.); Mt. Pinos Region, Goodenough Meadow, June 28, 1896, W. R. Dudley and A. F. Lamb (no. 4717, fr.; St.; fructibus parvis vix 5 mm. longis probabiliter nondum perfecte maturis); Sespe Creek, near Ten Sycamore Flat, alt. 600- 750 m., June 9, 1908, Abrams and McGregor (no. 169; G., St.); Mt. Pinos Region, below Snedden's, Lockwood Creek, June 23, 1896, Dudley and Lamb (no. 4632, St.; St.; vel exigua). Los Angeles County: Burbank, 1904, /. C. Nevin (m., fr. ; St.; very near S. exigua); Inglewood, April 12, 1901, LeRoy Abrams (no. 1493, f.; St.; glandulis 2, forma incerta); Florence, old bed of the Los Angeles River, April 13, 1903, L. Abrams (no. 3255, m., f.; M., St.; in floribus femineis interdum glandula dorsalis adest); same county?, Leakside, /. B. Grant (no. 6960, f.; St.; "shrub 8 ft. high"); San Antonio 7 For correct statements regarding this name and the following localities I am much indebted to Mr. S. B. Parish. iqiq] SCHNEIDER— AMERICAN WILLOWS 325 Mts., Prairie, fork of San Gabriel River, moist ground in a small open flat, alt. 1700 m., August 23, 1917, /. M. Johnston (no. 1685 m.;St.); San Bernardino County: San Bernardino, May 15, 1913, alt. 400 m., W. L. Jepson (no. 5591 m., fr.; A.). Orange County: Santa Ana, spring 1902, H. D. Geis (no. 653 vel 553; f., fr.; St.). San Diego County: Oneonta, April 24, 1904, H. P. Chandler (no. 51 16, f., fr., m.; N.; porro observanda) ; near Tia Juana, June 1895, 5. G. Stokes (f.; St.; stigmata pl.m. sessilia, forma porro observanda); same place, April 24, 1913,^. Eastwood (no. 2926, m.; A.) ; Tia Juana River, August 1902, A. C. Herre (fr.; St.; ut no. 4632). — Northern Lower California: Causito(?), May 29, 1883, C. R. Orcutt (no. 1180, fr.; M.; ut praecedens, sed amentis duplo brevioribus, ovariis pedicello quam glandula pl.m. sublongiore instructis). Kern County: along the Santa Fe Railroad, in low moist ground about 2 miles west of Bakersfield, April 6, 1905, A. A. Heller (no. 7591, m., f.; A., C, M., St.; looks somewhat like 5. exiguaX^ax. Hindsiana; "shrub 6 or 8 ft. high"). Inyo County: on the old Mitchell Range, resting Spring Valley, alt. 525 m., February 6, 1891, F. V. Coville and F. Funston (no. 263, St.; W.; see preceding remarks). Tulare County: Tule River above Porterville. INIarch 27, 1897, W. R. Dudley (no. 3578, f.; St.; pubescentia foliorum valde juvenilium fere ut in var. Hindsiana, sed ovaria parce pilosa iis 5. Parishianae simillima). 5. S. ARGOPHYLLA Nutt. N. Am. Sylva 1:71. pi. 20. 1843; Rowlee in Bull. Torr. Bot. Club 27:252. 1900, pro parte; Howell, Fl. Northw. Am. 2:618. 1902, pro parte; Piper and Beattie, Fl. Palouse Reg. Wash. 53. 1901; Piper in Contr. U.S. Nat. Herb., 6:213 (Fl. Wash.). 1906, pro parte. — 5. macrostachya Piper, I.e. 214 non Nutt.; Henry, Fl. S. Br. Col. 96. 1915.— 5'. sessilifolia Britt. and Shafer, N. Am. Trees ig6. fig. ij6. 1908, pro parte.— This species, in my opinion, has been misunderstood by almost every later author, owing probably to the inaccurate representation in Nuttall's plate. His Latin description runs: Salix argophylla, foUis lineari-sublanceolatis acutis sessilibus integerrimis utrinque argenteo-sericeis, stipulis obsoletis, amentis serotinis diandris, capsulis villosis lanceolatis. Besides this he says: "This species becomes a small tree from 12 to 15 ft. in height, as silvery and white as the Leucodendron argenteiim, the branches are brown, but the twigs are hoary with villous hairs. The leaves are very much crowded, soft, with whitish shining silky down, so abun- dant on either side as wholly to hide the veins, and nearly the midrib; they are also nearly without footstalks, entire on the margin, of a narrow linear outline and sharply acute, with a distinct bristly point, i . 5 to 2 inches long, and only about 3 lines wide. Stipules small and linear, seldom seen. The aments come out late with the leaves, and the flower branches produce 4-7 leaves. The male ament is small and narrow, with the scales lanceolate and villous, the female 326 BOTANICAL GAZETTE [april aments are oblong, the capsules lanceolate and villous We perceive no affinity that this species bears, except perhaps to the 5. angustifolia of the borders of the Caspian, from which at the same time it is probably very distinct. Nuttall's statements indicate that the main character of S. argophylla is the soft, villous, white pubescence which is also a char- acteristic of S. sessilifolia and S. macrostachya. He does not indi- cate the shape of the stigmas, owing probably to the fact that he collected only plants with mature capsules. The type locality is ''one of the branches of the Oregon [Columbia], the river Boisee, toward its junction with the Shoshonee" [Snake River] in western Idaho, Canyon County. So far as I know there is no type in exist- ence, but Nelson and Machride's no. 1057 and Machride's no. 228 from the same county seem identical with Nuttall's species. Andersson mentioned it first in his monograph in 1867 as follows: "5. longifolia **argyrophylla: (Nutt. Sylva Amer. p. 87 ?) : foliis et capsulis tomento argenteo tomentoso-micantibus. — In regionibus meridionalibus, ut in Mexico, etc.," and he adds a forma "angustissima: foliis anguste linearibus." "Hab. in ripis in Cahfornia (Fremont); Rocky Mountains (Nuttall)," giving as a synonym "S. brachycarpa Nutt. Amer. Sylva p. 85 ?. " In the Prodromus (1868) Andersson cites under his S. longifolia, argyrophylla Berlandier^s no. 2371 (recte 2341) and WrigMs no. 1873, and adds a forma opaca. He certainly misunderstood Nuttall's species entirely, and owing to the changed spelling of the name we may regard his var. argyrophylla as quite a new form which has nothing at all to do with S. argophylla. For a further explana- tion of Andersson's plant see under S. longifolia var. angustissima. S. longifolia argyrophylla of Bebb and other authors as well as S. fluviatilis argyrophylla Sargent are names applied to forms of very different origin, and may sometimes include the true S. argophylla, but mostly seem to refer to S. longifolia var. Wheeleri. Rowlee (1900) mixed with it S. Hindsiana Benth. and also forms which belong to S. exigua and 5. sessilifolia leucodendroides . Piper (1906) and Ball (in different herbaria) referred the forms I take for S. argophylla mostly to S. macrostachya, but Nuttall's type of this species belongs to S. sessilifolia, as previously explained. Male or sterile specimens of S. argophylla are not always easily separated from S. sessilifolia, as for instance those collected by igig] SCHNEIDER— AMERICAN WILLOWS 327 Jack and also by Rehder on Grant's Pass, Oregon. The female plants show almost the same stigmas as in S. exigua, and S. argo- phylla looks often quite intermediate between this species and S. sessilifolia. So far as can be judged at present by the specimens enumerated, its range seems to extend from Bonneville County in eastern Idaho, along Snake River to Canyon, Washington, and Nez Perces coun- ties, and into adjacent Washington (Walla Walla, Whitman, and probably also Franklin and Lincoln counties) as far as western Klickitat County, while in Oregon the species occurs in Sherman and Wasco counties, the forms from Klamath and Josephine counties being rather uncertain. The male specimen from British Columbia cited later looks much like S. sessilifolia, but Professor Piper, with whom I have had an opportunity to dis- cuss the matter, believes it is better referred to S. argophylla for geographical reasons. Only a close study in the field, especially of the forms of southern Washington and northern Oregon in the region of the Columbia and its tributaries, can elucidate the relationship of 5. argophylla with S. sessilifolia and the limits of their geographical distribution. At present I can hardly do more than to indicate what form has to be taken for Nuttall's S. argophylla, and how it seems to be related to and connected with either S. sessilifolia or S. exigua. It would be rather misleading to make too decisive statements as long as one's information is merely based on herbarium material. Specimens examined.— Idaho : Bonneville County: Idaho Falls, among rocks, along river, luly 4, 1901, E. D. Merrill and E. N. Wilcox (no. 803, m.; G.; "4-5 ft."); Canyon County: Falk's Store, slough and creek banks, alt, 660m., July II, 1911, A. Nelson and Macbride (no. 1057, fr.; G., M., St.); along the river, same alt., June 7, 1910, /. F. Macbride (no. 228, m.; G., M., St.); Caldwell, irrigation ditch, October i, 1910, C. R. Ball (no. 1705, fr.; W.; " 10 ft. ") ; Washington County: Weiser, alt. 660 m., July 5, 1899, M. E. Jones (no. 6554, f., fr.; W.); ?Nez Perce County: Clear Water River, June 18, 1894, L. F. Henderson (f., fr.,nonm.; W.; same as no. 2878 in C; forma satis ad S. exiguam spectans). — Washington: Walla Walla County: Waitsburg, June 24, 1897, R. M. Horner (R. 454, B. 451, m.; G., W.); Whitman County: Wawawai, July 9, 1901, C. V. Piper (no. 3592, m.); same place and collector, June 13, 1901 (no. 3595, m.); West Klickitat County: Columbia River, damp or wet places. May 31, July 1884, W. N. Suksdorf (m., f., fr. ; C, 328 BOTANICAL GAZETTE [april M., St.); Franklin County: Pasco, June 1902, jE. P. Baker (no. 70, m.; M.; vel ad 5. sessUifoliam referenda); Lincoln County: Sprague, alt. 560 m., June 3, 1893, /. H. Sandberg and /. B. Leiberg (no. 134; W.; forma porro observanda, pauUo ad S. exiguam spectans). — Oregon: Sherman County: Biggs, along stream, i mile south of Columbia River, August i, 1914, C. R. Ball (no. 1848, fr.; W.); Wasco County: Tygh Valley, June 1881 (vel 1880), T. J. Howell (m. vel androgyn.; A., M.); Hood River County, Hood River, May 25, 1879, J. T. J. Howell {m. vel androgyn.; C); Klamath County: along Sprague River above Yainax Valley, F. V. Coville, August 23, 1902 (no. 1312, St.; forma quamvis incerta) ; Josephine County: Grant's Pass, August 23, 1904, /. G. Jack (st., A.; forma incerta); same place and date, A. Rehder (st.; "large shrub": ut praecedens); ? County: Cache Bar, between Cache and Gordon creeks on Snake River, alt. 380m., June 19, 1897, E. P. Sheldon (no. 8325, m.); east Oregon, without exact locality, stream banks, May 9, June 7, September 1898, W. C. Cusick (no. i860, m., f., fr.; M.; "a straight upright shrub"; forma fohis lanceolatis satis denticulatis) . — British Columbia: Kootenay District, Cascade, near international boundary between Kettle and Columbia rivers, June 26, 1902, /. M. Macoun (no. 68128, O.; m.; G.). 6. S. EXiGUA Nutt. Sylva N. Am. 1:75. 1843; Rowlee in Bull. Torr. Bot. Club 27:255. pi. 9, fig. 15. 1900, pro parte; Piper and Beattie, Fl. Palouse Reg. Wash. 53. 1901; Howell, Fl. Northw. Am. 1:618. 1902; Piper in Contr. U.S. Nat. Herb. 6:213 (Fl. Wash.). 1906; BrittonandShafer, N.Am. Trees 195, ^g. J55. 1908; Ball in Coult. and Nels., New Man. Rocky Mts. Bot. 131. 1909; Garrett, Spring Fl. Wasatch Reg. 10. 1901, pro parte; ed. 2. 16. 1912, pro parte; Rydberg, Fl. Rocky Mts. 192. 1917, pro parte. — ^S*. longijolia var. /3 Hooker, Fl. Bor. Am. 2:149. 1839, quoad specim. Tolmieana. — S. longifolia Wats., Cat. PI. Nev. Utah, in King's Rep. 5:324. 1871, quoad specim. no. 1094, non Muhl.; Bebb in Coult., Man. Rocky Mts. Bot. 335. 1885, pro parte; Jeps. in Mem. Univ. Calif. 2:178 (Silva Calif.) 1910, pro parte. — S. longi- folia var. exigua Bebb in Wats., Bot. Calif. 2:85. 1879; Jones, Willow Fam. 24. 1908, pro parte. — 5. longifolia var. argyrophylla Macoun Cat. Can. PI. 1:450. 1883, pro parte; Jeps. in Mem. I.e. pro parte. — S.fluviatilis var. exigua Sarg., Silva N. Am. 9:124. 1896, pro parte; Sud worth in Bull. U.S. Dept. Agr. Div. For. 14:122 (Nomencl. Arb. FL). 1897, pro parte max; For. Trees Pacif. Slope 223. 1908.— 5. longifolia var. argophylla Jones, Willow Fam. 24. 1908, pro parte. — S. argophylla Henry, Fl. S. Br. Col. 96. 1915; iQiq] SCHNEIDER— AMERICAN WILLOWS 329 Rydbg., Fl. Rocky Mts. 188. 191 7. —The type of this species was collected by NiiUall with his fluvia lilts, probably ''on the banks of the Lewis River of the Shoshonee" (Snake River in Idaho), because at the type locality of S. fluviatilis on the Columbia in the vicinity of Portland, Oregon, this species is apparently the only one of the LoNGiFOLiAE according to Ball (Box. Gaz. 60:45, in note, 1915). Nnltall says: ''This species is also a native of the terri- tory of Oregon, and grew with the preceding, which it strongly resembles" (5. fluvialilis); he does not indicate the exact locaHty. I have a photograph of a so-called cotype of 5. exigua from Herb. P. consisting of a sterile branchlet. The label originally bore the inscription "5. longifolia, Missouri and Arkansas." The name longifolia has been crossed out, and in a similar handwriting is written ''exigua Nutt." Judging by the serration and nervation of the leaves there can be no doubt that the specimen belongs to S. longifolia. I do not know of a true type specimen of S. exigua, but there can hardly be any doubt as to the form Nuttall had in mind. From his phrase "capsulis lanceolatis sessihbus, demum nudiusculis" I infer that the t>^ical S. exigua is a form with, at least in the beginning, hairy ovaries, but Rowlee and other authors ascribe to it glabrous capsules. Ball (1909) is right in stating that it is "variable in fohage characters and sometunes very difficult to distinguish" from S. longifolia. In spite of havmg seen an abun- dant and well collected material, I am still at a loss how to define certain forms and to draw a sharp line between S. exigua on the one hand and such species as 6*. longifolia, S. argophylla, S. Parishiana, and also 5. taxifolia typica on the other. From S. longifolia and its forms it differs chiefly in the opaque color of the canescent leaf- surfaces, bearing a more or less dense appressed tomentum of short silky hairs (especially on the young leaves) of a silvery hue. The leaves are usually smoother with a hardly visible nervation, but in old leaves (for instance in those of the southern form) the veins are sometimes rather well marked ; their margin is mostly entire, but a dentation similar to that of S. longifolia may be observed in the southern forms. The fruiting aments usually are denser and the capsules as a whole shorter. S. argophylla chiefly differs, as pre- viously stated, by its more villous tomentiun, while S. Parishiana, 330 BOTANICAL GAZETTE ' [april which cannot be distinguished by its pubescence, may be recognized by the longer lobes of the stigmas and the more or less distinct style. Male specimens of these species sometimes prove difficult to dis- tinguish. In S. sessilifolia leucodendroides the base of the leaves usually is more obtuse and suddenly contracted in the very short petiole, while in S. exigua as well as in 5. Parishiana the leaves are mostly attenuated at the base, passing gradually into the somewhat longer petioles. S. Parishiana normally has linear leaves, while in S. exigua they are more linear-lanceolate, but all those characters have to be taken cum grano salis. There is a specimen before me from southern New Mexico, Dona Ana County, Mesilla, alt. 1150 m., June 19, 1897, E. O. Wooton (no. 39, m.; G., St., W.), of which the younger leaves are almost sessile, with a pubescence like those of var. leucodendroides, but are more linear; the older ones, which are more glabrescent and measure up to 12 by o. 5 cm., have a distinct petiole 2-3 mm. long. The pubescence and shape of the bracts seem to vary in the same manner in every species. Whether or not the shape and size of the anthers afford a useful character I cannot state. In those regions where the species meet each other hybrid forms are certain to occur. The range of what I call the typical form of S. exigua extends from southern Idaho (from which the type probably came) west- ward to Oregon (where the western line seems to run from about Wasco County to Klamath County) and Washington (where tt hardly reaches the eastern slopes of the Cascades), north- ward to British Columbia (where I did not see it from farther north and west than Clinton on the Fraser River) and southern Alberta (Medicine Hat) , eastward to central Montana and western Wyoming (Yellowstone Park), and southward to southeastern Nevada and southern California. In California it seeins to occur along the eastern border line from Modoc to Inyo County (Pana- mint Range), and in the south (Ventura to San Bernardino, Impe- rial, and San Diego counties). There are also forms very near to it in San Benito, Tulare, and Kern counties, which partly point toward S. sessilifolia var. Hindsiana. From the south I also have seen forms which come very near var. leucodendroides on the one hand and S. Parishiana on the other. As already stated, the limitation of these species is a very difficult task. 19 iq] SCHNEIDER— AMERICAN WILLOWS 331 In Nevada and Utah a form is found in which the female flowers have a ventral and a dorsal gland. To this form belongs S. nevadensis Wats., the type of which came from Nevada, Ormsby County, near Carson City. It is certainly not a good species, but I am inclined to keep it as a variety until it is proved by further observation that the presence of a dorsal gland is a character of no taxonomic value, and that no other character can be detected by studying the plant in the field. In proposing the name S. exigua var. nevadensis, nov. var. (5. nevadensis Watson in Am. Nat. 7 1302. 1873), I provisionally refer to it the following specimens, and wish to draw the attention of collectors to the localities mentioned. The type has glabrous ovaries, with a pedicel nearly as long as the ventral gland, while other forms with two glands have a more or less dense pubescence. Specimens examined.— Nevada: Ormsby County, at the base of the Washoe Mountains, near Carson City, alt. 1500 m., April 1868, 5. Watson (no. 1093, f. type; G.); same region, 1865, C. L. Anderson (no. 196, m., fr.; G.; ovaria pilosa); Washoe County, Franktown Creek, May 18, 1907, C. L. Brown (no. 1677, f.; Reno); Glendale, alt. 1300 m.. May i, 1909, P. B. Kennedy (no. 1743, m.; G.) ; sloughs between Pyramid and Winnemucca lakes, alt. 1250 m., June 2, 1913, P. B. Kennedy (no. 1996, m., fr.; G.; forma quasi ad S. sessili- foliam var. Hindsianam accedens); Truckee River, alt. 1350 m., June 6, 1913, P. B. Kennedy (no. 2010, m., f.; G.); central Nevada, without exact locality, 187 1, Wheeler (m.; syntype; G.).— Cahfornia: Nevada County, along Cold- stream, 3 miles above Truckee, July 17, 1913, A. A. Heller (no. 6953, fr.; forma aliquid incerta).— Utah: Washington County, St. George, alt. 600 m., April 9, 1880, M. E. Jones (no. 1644, m., f.; A., C); without date,£. Palmer (no. 8, m., f.; M.); Redsand, alt. 900m., April 24, 1894, M. E. Jones (no. 5117, m., f.; M.); Santa Clara, 1874, C. C. Parry (no. 8, m., f.; M.); Beaver County, Milford, along a stream, June 4, 1902, L. N. Goodding (no. 1018, fr.; W.); plains and mountains east of Milford, June 22, 1905, P. .4. Rydberg and E. C. Carlton (no. 6318, fr.; G.); Salt Lake County, Salt Lake City, 1350 m.. May 1869, 5. Watson (no. 1091, fr.; G.); same place. May 12, 1880, M. E. Jones (no. 1710, m., f.; A., C.) ; Davis County, Lagoon, common, alt. 1500 m., July 7-8, 1901, Pammel, Johnson, Buchanan, and Lummis (fr. adult.; M.; probably var. typica). — Idaho: Bear Lake County, Montpelier, creek banks, May 20, 1910, /. F. Macbride (no. 207, f.; G.); Power County, north of Arbon, bridge over Bannock River, August 6, 191 5, C. R. Ball (no. 2020, St.; G.; forma incerta). There are also the following 2 specimens from southern California which resemble 5. exigua and possess 2 glands in the female flowers: San Bernardino 332 BOTANICAL GAZETTE [april County, Cushenberry^ Spr[ing], Mojave Desert, June 2, 1901, 5. B. Parish (no. 4931; N., St.), and Los Angeles County, Los Angeles, April 1901, G. B. Grant (no. 1156; M.; apparently the same as Parish's plant). Both need further observation. In 1900 RowLEE described a S. exigua var. virens (Bull. Torr. Bot. Club 27:255), for the type of which a specimen collected by Rothrock in Arizona has to be taken. So far as can be discovered from the specimens cited by the author, I believe that Rowlee mixed several forms of different affinity, belonging partly to S. melanopsis Bolanderiana (Bolander, no. 5031; Kellogg and Harford, no. 922; W. G. Wright, Kernville [not Kernerville]), and partly to S. sessilifolia leucodendroides {Alder son [not Anderson] no. 700). The type of Rothrock, which is sheet no. 6122 in C, represents a female specimen of which the flowers can hardly be distinguished from those of 5. exigua. In the leaves it agrees well with a male specimen of Orcutt's (San Diego County, in the southwestern part of the Colorado Desert, Dos Cabesas, October 11, 1890, no. 2227; A., C), which number is also cited by Rowlee. Both may be taken for a rather glabrescent variety of S. exigua, but the leaves show under the lens a fine and thin silky pubescence and cannot be called ''nearly glabrous," a character apparently taken by Rowlee from the specimens of var. Bolanderiana. Rothrock' s and Orcutfs specimens come very near the 2 specimens of Parish and Grant with 2 glands in the female flowers. Besides these there is Parish's no. 3194 (San Bernardino County, San Ber- nardino Mountains, Big Morongo, alt. 900 m., June 15, 1894; m.; M.) that hardly differs from Orcutfs plant, and also LeRoy Ahrams' and McGregor's no. 406 (Los Angeles County, Liebre Mountains, Oakgrove Canyon and Elizabeth Lake, June 20-23, 1908; f., fr.; St.) seems to represent such a form the leaves of which become rather greenish at maturity, but the lower surface is rather gla- brescent in Rothrock' s specimens. This form somewhat simulates var. Bolanderiana, and I cannot express at present a definite opinion as to its real taxonomic value and true afi&nity. 8 Rowlee spells the name Cashewberry, but I read it as given, and S. B. Parish writes in a letter to Professor C. S. Sargent that this is the local way of spelhng the name, while on the map of the Geological Survey it is spelled Cushenbury. 1919] SCHNEIDER— AMERICAN WILLOWS 7,7,^ In BoT. Gaz. 65:25. 1918 I have made 5. stenophylla Rydbg. a variety of 5. exigua, referring to it the eastern and southeastern forms of this species. Rydberg's female type and male syntype came from southern Colorado, Huerfano County, Cuchara River, below La Veta {Rydberg and Vreeland, nos. 6393 f., 6392 m.; N.), and the ovaries are only partly glabrous, while most of the forms I take for var. stenophylla have wholly glabrous ovaries and fruits. The main character by which they differ from typical 5. exigua is the longer pedicel, which in the fruit usually surpasses the gland in length. After all, even this character can scarcely be regarded as constant, and var. stenophylla is connected with the typical form by numerous intermediates. As a whole, however, the forms of 5. exigua from Wyoming, Colorado, Arizona, New Mexico, Texas (Randall and El Paso counties), and probably also on the western border of Kansas, in northwestern Oklahoma, and in northern Mexico (northern Chihuahua), seem to present slight variations and may be called var. stenophylla until further studies in the field have led to a more proper understanding of the variability of this species. I suggested in Box. Gaz. 65:25. 1918 that S. Hindsiana var. tenuijolia And. (in K. Sv. Vet.-Akad. Handl. 6:56. 1867) might be identical with var. stenophylla, in which case the name tenuifolia would have to be used. As type a specimen collected by Burke on the banks of the Snake River near Fort Hall in Idaho has to be taken. Judging by a photograph and fragments of the type preserved in Herb. K. I cannot decide whether the male specimen really belongs to what I call var. stenophylla or to the t}^ical S. exigua. It comes from a region where both forms meet. The second specimen cited by Andersson "Nova Mexico (Schur) " is unknown to me, and may probably be referable to var. steno- phylla, which name I prefer to keep so long as the identity of the Snake River form remains uncertain. To var. stenophylla also partly belongs as a synonym S. longifolia * * * opaca And. (in K. Sv. Vet.-Akad. Handl. 6:55. 1867) in so far as it refers to Wright's no. 1873, while Berlandier's no. 2341 represents S. longi- folia angustissima. In western Nebraska and northeastern Colorado another form of 5. exigua has been found which somewhat reminds one of the f. 334 BOTANICAL GAZETTE [april Wheeleri of S. longifolia (see following). Rydberg described this form as S. luteosericea (in Britton, Man. 316. 1901) and kept the name in his Fl. Color. 94. 1906, while he makes it a synonym of his S. exigua in 191 7 (Fl. Rocky Mts. 192), as Ball has already done in 1909. The type came from western Nebraska, Banner County. I think it best at present to keep this form separate under the name S. EXIGUA var. luteosericea, nov. var., and I provisionally refer the following specimens to it in the hope that collectors may pay attention to the localities mentioned and try to get a better under- standing of this variety by studying it carefully in the field a; io its association with typical exigua and with S. longifolia. I can hardly point out a good character by which to recognize this form, but its pubescence is a little more villose, and the aments are more loosely flowered than in typical exigua or var. stenophylla. Specimens examined. — Western Nebraska: Banner County, Lawrence Fork, July 8, 1891, P. A. Rydberg (no. 368 partim, f. type; N.); Kearney County, dry creek, June 13, 1891, P. A. Rydberg (no. 369, m. syntype; N.); Scotts Bluff County, Platte bottom, in Mitchell Valley, August 4, 1891, P. A. Rydberg (no. 368 partim, fr.; N.). — ^Colorado: Weld County, Greeley, July 23, 1896, L. H. Pammel (no. 200, fr., 201, m.; M.) ; Larimer County, without exact locality, plains, alt. 1500 m., June 26, 1895, C. F. Baker {Patterson no. 9842, m., f., rather typical, the male specimen almost identical with exigua typica); Fort Collins, near river, June 26, 1896, L. H. Pammel (no. 202, f.; M.); same locality, meadow near river, August 6, 1898 (Hb. Agr. CoU. Colo., no. 2343, fr.; C); Morgan County, Fort Morgan, June 1896, L. H. Pammel (no. 204, St.; M.) ; Fremont County, Canyon City, banks of the Arkansas River, September 24, 1874, G. Engelmann (st., M.; vel var. stenophylla); Boulder County, August I [and 21 ?], 1884, July 20, 1885, G. W. Letterman (fr. ; M.); Denver County, Denver, August 20, 1884, G. W. Letterman (fr.; M.). — S. Dakota: Butte County, Indian Creek, along flood plain, July 31, 1911, 5. S. Visher (no. 2640, St.; C; f. incerta); Bennett County, Little White River, vaUeys, August 15, 1911, 5. 5. Visher (no. 2274, st.; C; rather uncertain, similar to S. longifolia Wheeleri) . There remains another form the proper interpretation of which raises many difficulties. It was described by Henderson as 5*. longifolia tenerrima from specimens collected by the author in Idaho, Elmore, and Canyon counties. At first sight it can hardly be distinguished from what I call S. longifolia var. pedicellala (see later), especially from such specimens as Easlwood's no. 465, but a iqiq] SCHNEIDER— AMERICAN WILLOWS 335 closer inspection shows that the leaves as a whole are narrower and the fruits shorter. In this respect it agrees more with S. exigiia, of which it would represent an extremely glabrescent form. Heller made it a species, and Ball evidently took the same view, as shown by his determinations of herbarium specimens before me, but Rydberg (191 7) quotes it as a synonym of his S. linearifolia, which is the same as var. pedicellata of longifolia. So far as I know the range of this peculiar form, it seems to be restricted to southwestern Idaho (region of Boise River), northwestern Wyoming (Yellowstone Park and northern Lincoln County), and adjacent southern Montana (Carbon and Big Horn counties). There is, to my present knowledge, no 5. longifolia in this region, but it is within the range of 5. exigua. I am therefore inclined to follow a suggestion of C. V. Piper, with whom I have discussed this question, and to refer var. tenerrima as a variety to S. exigua. S. exigua var. tenerrima, nov. comb. — S. longifolia var. tener- rima Henderson in Bull. Torr. Bot. Club 27:354. 1900. — S. tener- rima Heller, Cat. N. Am. PI. ed. 2. 4. 1900. — S. fluviatilis var. tenerrima Howell, Fl. N.W. Am. 618. 1902. — S. linearifolia Rydbg., Fl. Colo. 94. 1906, ex parte; Fl. Rocky Mts. 192. 1917 ex parte. — A t^-po praecipue differt foliis angustioribus linearibus etiam maximis vix ultra 4 mm. latis juvenilibus ut rami noveUi parce breviter sericeis cito glabris vel pilis parcis difficile recognoscentibus vestitis utrinque satis viridibus vix nervatis vulgo pl.m. distincte denticulatis dentibus brevibus subglandulosis saepe satis distanti- bus, ovariis subsessilibus glabris, bracteis oblanceolatis tantum versus basim pilosis, fructibus vulgo pedicello distincto glandulam duplo superante instructis conico-rostratis pedicello excluso ad 6 mm. longis. Specimens examined. — Idaho: Elmore County, shady rocky banks of mountain rills gone dry, July 12, 1895, L. F. Henderson (fr., type; G.) ; Canyon County, Payette River, sandy bottoms, August i, 1897, L. F. Henderson (fr.; G.) ; Falk's Store, open sandy slopes, alt. 660 m., May 24, 1910, /. F. Macbride (no. 98 m., fr. juv.; G., M., St.; "loose clumps"). — Wyoming: Yellowstone Park, Soda Butte Creek, July 14, 1899, in small clumps on the stony river bottom, A. and E. Nelson (no. 5866, fr.; G., St.); Lincoln County, Jackson's Hole, banks of Gros Ventre River, July 14, 1901 , 5. D. Merrill and E. N. Wilcox (no. 996,fr.; G.,M.; "10 ft."). — Montana: Big Horn County, Crow Agency, 336 BOTANICAL GAZETTE [april August 30, 1871, Coulter (no. 5, st.; C; forma porro observanda) ; Carbon County, near Red Lodge, July 28, 1893. /. N. Rose (no. 50. fr. adult.; forma aliquid incerta); Gallatin County, Bozeman, Gallatin River, low ground, October 4, 1905, /. W. Blankinship (no. 465, St.; A.; forma incerta ad 5. longifoliam pedicellatam spectans) ; Rosebud County, Forsyth, north of town, toward river, 1908, C. R. Ball (no. 1305, St.; G.; "6 ft. high"; forma porro ob- servanda).— Utah: Cache County, Logan Canyon, above Logan, August 8, 1914, C. R. Ball (no. 1864, fr. ; W.; forma glabra pro 5. exigua determinata, porro observanda). This variety needs further observation in the field, and some of the specimens cited are uncertain owing to the lack of fertile material. Some forms of S. longifolia pedicellata are extremely alike, but the leaves show a more or less prominent (often very fine) venation, while in the leaves of var. tenerrima the lateral veinlets are scarcely visible and finely impressed; the fruits of both are sometimes almost identical, and I am not yet sure of the true affinity of var. tenerrima. G. J. Jack, August 16, 1918, collected on the Laramie River, Laramie, Albany County, Wyoming (no. 1017), sterile specimens of a form of which I am not sure whether it is var. tenerrima or var. pedicellata, neither of which has hitherto been reported from southeastern Wyoming. Professor Jack says: "Slender, coarse, grasslike, 2-3 ft. high, covering wide sandy areas," and he told me that it is a very distinct low form. There are now living plants in the Arnold Arboretum which I hope will prove useful in determining its real afiinity. There is still one form which needs a few words. It was col- lected by S. M. Tracy and F. S. Earle in western Texas, Jeft" Davis County, Limpia Canyon, April 24, 1902 (no. 210, fr. ; C, G. ; dis- tributed as " 5. longifolia opaca Ands."), and it seems to be identical with Mexican specimens mentioned by me in Box. Gaz. 65:23. 1918, under 5. taxifolia. The habit and the leaves agree well with those of that species, but the fruits in no. 210 are much more like those of 5. exigua with short sessile stigmas. It looks almost like a new species closely related to S. exigua, which seems to show a variability remarkable even among willows. 7. S. MELANOPSis Nuttall, N. Am. Sylva 78. pi. 21. 1843; Rowlee in Bull. Torr. Bot. Club 27:256. pi. g, fig. 16. 1900, pro parte; Piper and Beattie, Fl. Palouse Reg. Wash. 53. 1901; Piper in Contr. U.S.N. Herb. 11:213 (Fl. Wash.). 1906, pro parte; Ball iQig] SCHNEIDER— AMERICAN WILLOWS 337 in Coult. and Nels., New Man. R. Mt. Bot. 131. 1909; in Piper and Beattie, Fl. Northw. Coast 114. 1915; Henry, Fl. S. Br. Col. 97. 191 5; Rydberg, Fl. R. Mts. 192. 191 7. — S. longifolia Bebb apud Coulter, Man. R. Mt. Bot. 335. 1885, pro parte, non Muhl. — S. fluviatilis Howell, Fl. Northw. Am. 1:618. 1902, pro parte, non Nutt. — This is a well marked species the type of which was found by NuTTALL "at our station called Fort Hall, in the plains of the Rocky Mountains, on alluvial lands of Lewis River of the Sho- shonee." According to Ball (1909), this is old Fort Hall, near Pocatello, in Bannock County, eastern Idaho, south of the present Fort Hall,' near Blackfoot, in Bingham County. I have seen a photograph of a cotype preserved in Herb. P. Ball (1909) gives the range as follows: "Common in northeastern Oregon, eastern Washington, and British Columbia as far east as the Selkirks." I have not seen a specimen from the t^-pe region or other parts of southern Idaho, but only from northern Idaho, Montana (Teton County, Midvale, L. M. Umbach, no. 170), Alberta (Crow Nest Pass and Jasper), where it seems to reach its northern limit at about the 53d parallel, British Columbia (in the Chilliwack Valley and at Revelstoke) , Washington (where I have seen it west of the Cascades only from King County, Snoqualmie), Oregon (where it was collected by Ball in 191 5 as far west as the Umpqua River, Rose- burg, Douglas County, and by Applegate, no. 2224, at Ashland, Jackson County), and northern and northeastern California (see below), where it seems to pass into var. Bolanderiana. According to Ball (Box. Gaz. 60:45, first note, 1915), S. Bolanderiana is asso- ciated with S. sessilifolia at Roseburg and also farther north "on the Willamette River at Corvallis," Benton County. What I have seen from Oregon I take for the true S. melanopsis, which ought to be looked for also in northern Utah and in western Wyoming." Its ' This locality, however, is identical with that given for Fort Hall in Lippincott's Geogr. Diet., ed. of 1855; while on the map in the Century Atlas of 191 1 old Fort Hall is indicated south of the 43d parallel just north of Pocatello. Judging by Rand McNally's map the whole region between the two places is called Fort Hail. '" There is a specimen from eastern Wyoming, Converse County, Rawhide Creek, south of Patrick, August 27, 1901, H. P. Baker (m.; M.), which looks like typical S. melanopsis. In Herb. C. I found a specimen from Colorado, Clear Creek County, damp places along Clear Creek, 1885, H. N. Patterson (fr. adult, [sheets 5523 and 107801 ] ) , which clearly resembles S. melanopsis. I am not sure whether the localities given are correct. 338 BOTANICAL GAZETTE [april occurrence so far north in Alberta is interesting. In the north a form with more hairy, almost shining silky leaves seems to be not infrequent (see /. Macoun's specimen from Lower Arrow Lake, no. 24569, O.). The species has usually been mistaken for S. longifolia or 5. fluviatilis, but apparently it forms with the southern var. Bolanderiana a well marked type in this section, and I am not yet sure to which other group of it S. melanop- sis is most closely related. Ball (Box. Gaz. 60:51. 191 5) speaks of a "6'. fluviatilis-melanopsis''' aggregation in contrast with the S. sessilifolia group, but I think S. melanopsis has very little to do with the true S. fluviatilis. The specimens from Umatilla County, Oregon, western slope of the Blue Mountains, in a swampy meadow at Ukiah, June 24, 1908, W. Cusick (nos. 3260, 3261, fr. juv. ; N., St.), need further observation. The young fruits show a short style and are almost sessile. The main characters of S. melanopsis may be gathered from the key. The species is not even mentioned by Andersson (1858, 1867, 1868), and its identity has first been revealed by Rowlee (1900), who erroneously states that "it is particularly abundant along the Columbia River where Nuttall saw it." I have not seen all the specimens cited by Rowlee, but those of Coville, from Washington, Cowlitz County, north fork of Lewis River, July 16, 1898 (no. 719, fr.; W.), which are not men- tioned in Piper's Flora and which have leaves that measure up to 9:2.2 cm., seem not to represent typical S. melanopsis, and I have not yet been able to identify them properly. In Herb. C. are sim- ilar specimens collected by W. N. Suksdorf in W. Klickitat County, "rocky bank of the Larm River," July 17, 1884. After all thej^may be taken for a form of S. melanopsis with very broad leaves. In Cali- fornia 6*. melanopsis is mostly represented by the following variety: 7b. S. MELANOPSIS var. Bolanderiana, nov. var. — S. longifolia Bebb in Watson, Bot. Calif. 2:84. 1879, pro parte, non Muhl.; Jep- son, Fl. Calif. 2:340. 1909, pro parte; in Mem. Univ. Calif. 2:178 (Silva Calif.). 1910, pro parte. — S. Bolanderiana Rowlee in Bull. Torr. Bot. Club 27:257, pi. g, fig. 12. 1900. — S. exigua var. virens Rowlee, I.e. 255, pi. 9, fig. 11.— S. argophylla Rowlee, I.e. 252, quoad specim. Bolanderii (non Breweri!) no. 5031. — 5. fiuviatilis East- wood, Handb. Trees Calif. 37. 1905, pro parte, non Nutt.; Sudw., 1919] SCHNEIDER— AMERICAN WILLOWS 339 For. Trees Calif. Slope 222. fig. gi. 1908, pro parte.— Of this variety Rowlee has given a very incomplete description, and in citing the specimens he says ''Bolander, nos. 49, 58, 4958, 5031." There are no nos. 49 and 58 of Bolander, but only no. 4958, which has to be taken for the type. No. 5031 is also cited by Rowlee under S. exigua var. virens, of which I previously have spoken, and again under S. argophylla as a number of Brewer, who, so far as I know, never collected a specimen bearing the same number at the same locality from which Bolander's plant came. This variety differs from the tjq^e chiefly by the characters indicated in the key. Rowlee's statement in his key that in S. melanopsis the leaves are ''distinctly glaucous and prominently veiny beneath" while they are "not distinctly glaucous nor veiny beneath" in S. Bolanderiana is not correct. The leaves are some- times rather greenish beneath in both forms. The t>'pical form of var. Bolanderiana is somewhat pubescent, while most of the speci- mens before me belong to a glabrous form. There can also be observed a slight variation with partly hairy ovaries and fruits in the specimens of /. Burtt Davy (no. 5691, from Hoopa Valley, Humboldt County, California) and 5. Watson (no. 1092, Truckee Valley, Washoe County, Nevada). Both need further observation, and may represent hybrids with S. exigua. This seems also the case with A. A. Heller's no. 6953 (along Coldstream, 3 miles above Truckee, July 17, 1908). On the other hand, specimens collected at Sunol Valley, Alameda County, June 29, 191 6, by L. R. Abrams (no. 5692, no. 5693, f.; St.), of which the male plant cannot be distinguished from typical var. Bolanderiana, possess ovaries and fruits which are hairy throughout or become glabrous only to a slight degree. They do not look like hybrids, and seem to repre- sent a distinct form with pubescent ovaries and rather silky tomentose young leaves. The typical S. Bolanderiana has rather broad leaves, but there are before me many very narrow leaved specimens, and further observation in the field must show whether the forms with linear- lanceolate leaves can be separated from the typical form. I do not wish to propose too many new varieties and forms which are only known to me from herbarium specimens, but I beheve that a 340 BOTANICAL GAZETTE [april closer study of many difficult forms which I can only briefly mention will lead to a different conception of them. I have seen specimens of var. Bolanderiana from the following counties in CaUfornia (north to south): Humboldt, Siskiyou (.4. A. Heller, no. 8058, female part not quite typical), Shasta, Lassen, Plumas, Butte, Nevada, ?Men- docino (.4. Kellogg and W. G. W. Hartford, no. 922, ?Ukiah), Lake, Solano, Alameda (Sunol), Amador, Tuolumne, Mariposa (Bolander, no. 4958, type!, Yosemite Park, Slough's Valley), Fresno, Monterey, Tulare, and Kern. It may even occur farther south. There is a specimen from San Bernardino County, near head of San Antonio Canyon, in a narrow rocky canyon, alt. 2250 m., July 5, 1918, /. M. Johnston (no. 2087, flor. abnorm. m. et f. mixtis; A.; "shrub, low, under im."). The leaves are almost wholly glabrous when maturing, at least on the lower surface, which is more or less distinctly glaucescent. The flowers, however, are abnormal, the female ones hard to distinguish from those of S. exigua, but glabrous, or almost so. The form may belong to S. exigua virens, if there is really such a variety, or it may be related to var. Bolanderiana. The normal form is represented by Johnston's nos. 1401 and 1665, from the upper San Antonio Canyon. I am much obliged to Mr. Johnston for the following information: Numbers 1401, 1665, 2087 from near head of San Antonio Canyon. To me this is the most interesting plant I sent you. I have thoroughly explored the San Antonio Mountains, but I have only found the single colony from which all my specimens were obtained. It grows as a dense, low, compact shrub (hardly over a meter in height) on the rocky floor of a very deep gulch. A short distance away is found a large colony of S. flavescens and scattering shrubs of S. Watsoni. The nearest Longifoliae that I know of is 7 miles away and is the colony from which my 1685, which you doubtfully referred to S. Parishiana, was obtained. I have never yet seen in S. California a Longifoliae so high in the mountains and associating with such typically boreal species as this one does. You have probably noted that the aments contain both staminate and pistillate flowers, which may be due to its strange habitat. I noted that a large percentage of the aments were entirely sterile at the tune I collected the specimens. 8. S. LONGiFOLiA Muhl. in Neue Schr. Ges. Natf. Fr. Berlin 4:238. pi. 6. fig. 6. 1803, non Lamarck;" in Ann. Bot. Konig " According to the international rules, Muhlenberg's name can stand because Lamarck's (Fl. Fr. 2:232. 1778) is nothing but a synonjon of S. viminalis L.; in following the Philadelphia Code the name S. interior Rowl. has to be used, and I would not keep Muhlenberg's name if Lamarck's were not an unconditional synonym, and could be applied to a form differing from typical S. viminalis. iqiq] SCHNEIDER— AMERICAN WILLOWS 341 2:66. pi. 5. fig. 6. 1806; Carey in Gray, Man. Bot. N.U.S. 429. 1848; Andersson in K. Sv. Vet.-Akad. Handl. 6:54. pi. 4. fig. 35. 1867, pro parte et excl. var.; in DC, Prodr. 16^:214. 1868, pro parte et excl. var.; Bebb in Coult., Man. Bot. R. Mts. 335. 1885, pro parte; apud Watson and Coulter, Gray Man. ed. 6. 482. 1890; Robinson and Fernald, Gray's New Man. 323. fig. 64Q. 1908. — S.fluviatilis Sargent in Gard. and For. 8:463. 1895, pro parte, non Nutt.; Silva N. Am. 9:123. pi. 4^4. 1896, pro parte et excl. var.; Man. Trees N. Am. 175. 1905, pro parte; Schneider, 111. Handb. Laubh. I. 32, figs. 11 h~l, 12 m-m\ 1904; Ball in Proc. Iowa Ac. Sci. 7:145. 1900; in Coult. and Nels., N. Man. R. Mts. Bot. 131. 1909, pro parte; in Box. Gaz. 60:397. 191 5; Britt. and Brown, 111. Fl. 1:497./^. iiSi. 1896; Sudworth, Nomencl. Arb. Fl. U.S. 122. 1897, pro parte; Rydberg in Britt., Man. Fl. N. St. Can. 316. 1901; Hough, Handb. Trees N. St. Can. S4. figs, gy, g8. 1907, pro parte maxima. — S. interior Rowlee in Bull. Torr. Bot. Club 27:253. pi. 9, figs. 12, I J. 1900; Small, Fl. S.E.U.S. 342. 1903, pro parte; Britt. and Shafer, N. Am. Trees 193. fig. 1^4. 1908; Britt. and Brown, 111. Fl. ed. 2. 1:595./^. 1458. 1913; Rydberg, Fl. R. Mts. 192. 191 7. — This is the type species of the section and the only one known from the central and northeastern states and eastern Canada. The type came from Lancaster, Pennsylvania. It has its head- quarters in the regions of the Mississippi, Arkansas, and Missouri, while toward the east the Ohio seems to form the southern border hne of its range up to Pennsylvania. The mouth of the Mississippi in Louisiana is the southernmost point of the range of S. longifolia; its western boundary runs apparently just south of the Red River in Louisiana and Texas, thence through western Kansas, the north- eastern corner of Colorado, touching Wyoming in its northeastern part, from whence it runs through western Dakota to Manitoba. In Texas, southern New Mexico, and northwestern Mexico it is represented by var. angustissima (see later), while in the northwest from western Dakota and northeastern Wyoming through eastern Montana, Saskatchewan, and eastern Alberta the var. pedicellata seems to be the prevailing form, reaching its northwestern limit in the Yukon Valley (vicinity of Dawson and the adjacent parts of eastern Alaska, Fairbanks) and the upper Mackenzie region in the 342 BOTANICAL GAZETTE [april Northwest Territories. The northern border line of the range of S. longifolia and var. pedicellata is not yet exactly known. Approxi- mately it seems to run in the west from Fairbanks in Alaska to Fort Simpson in the Northwest Territories and through the Athabasca Plains and central (or southern ?) Manitoba and southern Ontario to the south of James Bay and to about Lake St. Johns in Quebec, from where the eastern line turns southeast to western New Brunswick (Woodstock, Pokiok) and then southward to New Hampshire along the Connecticut River to Delaware and the District of Columbia. The species apparently reaches its best development in the rich river bottoms from Louisiana to Indiana, while in Oklahoma, Kansas, Nebraska, and Iowa the form of the sand bars seems to prevail, which has narrower, smaller leaves. In the region of the Great Lakes and in the northeast, but also in other portions of the range under similar ecological conditions, the following variety seems to occur frequently: S. LONGIFOLIA var. Wheeleri, nov. comb. — S. interior var. H^/?ee/m Rowlee in Bull. Torr. Bot. Club 27:253, />/. p,^g. 74. 1900. — S. Wheeleri Rydberg in Britt., Man. ed. 2. 1061. 1905; Britt. and Br., 111. Fl. ed. 2. 1:595. iQ^S- — S. longifolia (vel S. fluviatilis) var. argyrophylla Auct. div. pro parte, non And. — I agree to a certain extent with Schaffner (in Ohio Nat. 14:255. 1914), who regards this variety as an ecological form, and I have already pointed out that similar forms seem to occur in S. exigua (see var. luteo-sericea) , S. melanopsis var. Bolanderiana, etc. Those forms very often look quite distinct, especially in the herbarium. The broad leaved forms of var. Wheeleri can easily be taken for a well marked species if one does not have a very rich set of specimens showing all the intermediates between such forms as we know from Maine (Caribou) and New Brunswick and the narrow leaved forms from Lake Champlain, Lake Superior, etc. It may be that the easternmost forms are not quite identical with the typical var. Wheeleri from the region of the Great Lakes, but to decide this question we need a careful study of this form as it is observed in New Brunswick, Maine, Connecticut, western Quebec, and eastern Ontario. There is a male plant in cultivation in the Arnold igig] SCHNEIDER— AMERICAN WILLOWS 343 Arboretum which was brought by Professor /. G. Jack probably from the St. Lawrence region in Ontario. It is extremely Hke the female specimen of Bissell from Glastonbury, Connecticut, and both agree well with the specimens cited from Maine and New Brunswick. In BisseWs plant the stigmas are rather long and narrow, resembling somewhat those of the western 5. fliiviatilis but without a trace of a style. The leaves too of both plants are not very different in their shape, but var. Wheeleri has a coarser silky pubescence of longer hairs. Rowlee stated that "the silvery vesture of this shrub is much like that of S. argophylla of the Pacific Coast." As I have explained under this species, Rowlee did not interpret it correctly. At present I refer to var. Wheeleri the following specimens, and I hope collectors will pay attention to this plant at the localities given. Eastern North Dakota: Benson County, Pleasant Lake, 99 mer., every- where along watercourses, July 2, 1911, /. Luncll (m., llor. satis abnorm.; A.). — Iowa: Story County, Ames, 1888, A. S. Hitchcock (m.; M.); Fremont County, Hamburg, July 4, 1914, L. H. Pammel and H. B. Clarke (no. 44, m.; A.; a hairy sand-bar form). — Illinois: St. Clair County, Cahokia, July 23, 1895, N. M. Glatf elder (m.; M.); Winnebago County, Fountaindale, 1877, M. S. Bcbb, (fr. ; JNI.; narrow leaved form, probably cultivated); Cook County, Dunnmg, fields. May 16, 1916, F. C. Gales (no. 1428, m.; C). — Indiana: Noble County, near Rome City, June 11, 1916, Deam (no. 20118 A, ex parte, f., fr.; A.); Union County, Liberty, July 1886,/. iV. Rose (st.; C). — Michigan: Wayne County, Belle Isle, July 8, 1903, i). A. Farwell (f.; A.; according to a letter of Farwell this form was named by Rowlee himself var. Wheeleri, but it represents a very glabrescent form difficult to separate from typical longifolia). — Wisconsin: Brown County, Green Bay, south shore, June 1878, /. H. Schuette (m., st.; C). — Minnesota: Buffalo Lake, June 1891, B. C. Taylor, (m.; C). — Ohio: Erie County, Cedar Point, August 2, 189s, E. L. Moseley (st.; G.); September 4, 1898, Moseley (st.; W.; folia ad 8:2 m. magna, elliptico-oblonga) ; July 3, 1908, R. F. Griggs (no. 2, m.; N.; folia ad 8:1.5 cm. magna, distanter ciliato-serrata) ; without exact locality and date, W. S. Sullivant (no. 49, St.; N.); Lake County, near Painesville, May 19, 1892, O. Hacker (no. 431, m.; C); Franklin County, Columbus, 1840, W. S. S. (st.; G.); Ottawa County, Bay Point, sandy shore, August 20, 1914, L. H. MacDaniels and A. J. Fames (fr.); Ross County, ChiUicothe, June 16, 1899, A. D. Selby (no. 120, st.; C). — Pennsylvania: Erie County, Presque Isle, Lake Erie, July 23, 1868, T. C. Porter (st.; N., C); York County, shores of the Susquehanna near 344 BOTANICAL GAZETTE [april McCall's Ferry, September 13, 1864, T. C. Porter (m.; C; "shrub 5-6 ft. high"; forma pecuHaris foliis late oblongo-elUpticis ad 9:2.2 cm. magnis).— New York: Erie County, shores of Lake Erie near Buffalo, June 30, 1899, /. F. Cowell (St.; N.); Clinton County, shore of Lake Champlain, near Plattsburg, August 8, 1902, A. Rehder (st.; A.); Tompkins County, Fall Creek ravine, on rocks, May 29, June 6, 1885, W. R. Dudley (m., St.; C; folia pl.m. oblanceolata).— Vermont: wet shore of Lake Champlain, July 8, 1914, Ch. H. KnowUon (m.; NE.); June 15, 1896,^. /. Grout (f.; NE.; stigmata satis elongata).— Connecticut: Hartford County, Glas- tonbury, banks of Connecticut River, May 18, 1902, C. H. Bissell (f.; G.; "small shrub"; forma distincta porro observanda) ; New London County, Lyme, near Selden's Cove, July 29, 1902, C. B. Graves (st.; G.; "2 ft. high"; ut praecedens).— Maine: Aroostook County, Caribou, gravelly river beach, July 18, 1902, E. F. Williams, J. F.Collins, and M. L. Fcrnald (st.; G.; forma satis distincta porro observanda); same locality and date, E. F. Williams (st.; A., G.).— New Brunswick: Woodstock, on the bars in the St. John River, August 30, 1899, Macoun (no. 22609, 0-; st.; very much like the Connecticut forms); near Pokiok, July 8, 1889, Brittain (no. 24577, O.; St.; ut praecedens); above Fredericton, on island, August 23, 1890, /. Brittain (no. 6, fr.; C; ut praecedens); Keswick, June 6, 1891, J. Brittain (no. 4, f.; C.).— Ontario: Lambton County, Fort Frank, 35 miles from Port Huron, Michigan, July 21, 1905, C. K. Dodge (st.; A.; forma densissime sericea); Welland County, Point Albino, August 28, 1896, C. L. Pollard (st.; W.); James Bay, Moose Factory, July 15, 1904, W. Spreadborough (no. 6262e, O.; St.; forma porro observanda pauUo sericea). Every species inhabiting such a wide area as S. longifolia and growing under so many different ecological conditions will naturally show a great degree of variability. Besides this there are quasi intermediate forms with S. exigua in all the regions where both species meet, and it is difficult to decide whether the northwestern forms of what I call var. pedicellata really belong to S. longifolia or to S. exigua, as Ball seems to believe according to his determina- tions in different herbaria. The synonymy of var. pedicellata may be given as follows: 8b. S. LONGiFOLL^ var. pedicellata Andersson in K. Sv. Vet.- Akad. Handl. 6:55. 1867; in DC, Prodr. 16^:214. 1868.— 5. rubra Richardson in Franklin, Narr. Jour. Polar Sea App. 752. 1823, nom. nud., non Hudson.— 5. longifolia ( ?) Torrey in Ann. Lye. Nat. Hist. N.Y. 2:248 (Coll. PI. R. Mts. James)." 1828; Andersson in " The specimen (preserved in N.) has been collected by James either in eastern Wyoming or eastern Colorado, and seems to belong to this variety. iqiq] SCHNEIDER— AMERICAN WILLOWS 345 Ofv. K. Vet.-Akad. Forh. 15:116. 1858, ex parte; Macoun, Cat. Canad. PI. 450. 1883, ex parte; Sargent, Rep. For. Trees N. Am. loth Census U.S. 9:168. 1884, ex parte.— 5. fluviatilis Sargent in Gard. and For. 8:463. 1895, ex parte, non Nutt.; Rowlee in Bull. Torr. Bot. Club 27:254. 1900, ex parte; Henry, Fl. S. Br. Col. 97. 1915. — 5. interior Rowlee, I.e., 253, ex parte; Britt. and Br., 111. Fl. 1:595.. i9i3» ex parte. — S. linearifolia Rydbg. in Britton, Man. 316. 1901; Fl. Color. 94. 1906, ex parte; Fl. R. Mts. 192. 1917, ex parte; Small, Fl. S.E.U.S. 342. 1903, ex parte. — S. longifolia var. interior Jones, Willow Fam. 25. 1908, ex parte. — I have seen a photograph and fragments of the type of var. pedicellata, collected by E. Bourgeau, "Saskatchewan bords des Lacs, abondant, 21 Juin 1858" and preserved in Herb. K., and also of the type of 5. rubra Rich, from the ''Mackenzie River." This specimen of Richardson's represents the same form as the material from "Cumberland House" in Saskatchewan, which is a syntype of S. linearifolia Rydbg. in Herb. N. This variety differs from typical S. longifolia chiefly in its narrower, linear leaves, and its glabrous ovaries, which are more or less sessile when young but usually distinctly pediceled when in fruit, the pedicels often being twice as long as the ventral gland. As previously stated, var. pedicellata is the prevailing form in the northwestern part of the range of 5. longifolia, but there are also forms near the southern limit of its habitat which can hardly be distinguished from var. pedicellata (for instance Munson's specimens from the Red River near Colbert's Ferry, north of Denison, Texas, April 19, 191 1, f., fr.; A.). As previously stated, the most southern form of S. longifolia is represented by var. angustissima And. (1858'^) with which I have dealt in Bot. Gaz. 65:26. 1918. Besides the Mexican specimens here cited, I refer the following to this variety, which seems too closely connected with the typical S. longifolia to be kept as a distinct species. Specimens examined. — Texas: without exact locality and date, Berlandier (nos. 911, 2341, 2368, 3019, cotypes; G., M.; 1938, f.; M.; nos. 2341 and '3 Later, in Monogr. 1867 and in Prodr. 1868, Andersson used this name for dif- ferent forms, partly belonging to S. sessilifolia var. Hindsiana, partly to 5. exigua (probably var. stenophylla). 346 BOTANICAL GAZETTE [april 2368, of which the last has to be taken as the type of 5. Thurberi, have been erroneously attributed by Rowlee to G. Thurher, to whom only the following specimen belongs) ; Horse Head Cruping ( ?) River, October 1850, G. Thurher (no. 9S; G.; "10-12 ft."); ?Pecos County, banks of the Pecos, 1889, iVea//?y (no. :3^2>^ n^-; W.); September 1881, V. Havard (m., f.; W.; ad var. typicam accedens); Brewster County, Rio Grande, south of Chisos Mountains, August 1883, V. Havard (m., f.; W.); Val Verde County, Del Rio, along streams, October 18, 1916, E. J. Palmer (no. 11069, f.; A.); Potter County, Amarillo, creek banks, July 13, 1917, E. J. Palmer (no. 12539, f-. fr.; A.; ad var. typicam accedens) ; along Rio Grande, near San Vincente, August 26, 191 5, M. S. Young (m., f.; M.); Guadalupe County, in the dry bed of the Cibolo 12 miles east of New Braunfels, August 1851, F. Lindheimer (no. 615 [=1191], f.; G., M.); Comanche County, Comanche Spring, Lindheimer (no. 1190, f.; M.); Mata- gorda County, banks of Peyton Creek near Bay City, May 6, 1916, E. J. Palmer (no. 9689, m.; A.); Cameron County, near Brownsville, November 1888, Nealhy (no. 30, f., fr.; W.); (New Mexico?), Rio Grande, July 1848, C. Wright (m.; G.; "small tree"); without locality, 1849, C. Wright (no. 668, m.; G.,W.). There have also been described the following forms which I have not yet been able to elucidate: S. longifolia var. sericans Nees V. Esenbeck in Wied-Neuwied, Reise In. N. Am. 2:448. 1841; Engl. ed. by Lloyd, Trav. Int. N.A. 518. 1843, collected on the Missouri, probably in eastern Montana about July 8 (see I.e. 1:472 [Engl. ed. p. 211]). I would refer it to S. exigua, but the lower flowers of the male aments are described as "triandri"; otherwise the description agrees with S. exigua. — S. longifolia i. integerrima Kuntze, Rev. Gen. PI. 2:643. 1891, and f. paucidentieu- lata Kuntze, I.e. The first is characterized by the phrase "folia denticulata" and as type is given "U. St., Madisonthal"; while the second has "folia paucidenticulata " and came from " Cheyenne, Nebr." The author adds "Ausserdem kann man eine f. multi- denticulata unterscheiden." I suppose those forms are simply typical 5. longifolia. With the hybrids which doubtless occur only too frequently where different species grow together it is impossible to deal, as long as it has not yet been possible to limit the species in a more satisfactory manner. The main purpose of this paper is to point out the correct application of certain names, and to direct atten- tion to such forms as need a close study in the field. Arnold Arboretum t Jamaica Plain, Mass. RESPIRATION AFTER DEATH' A. R. C. Haas (with three figures) It is commonly stated that when respiration ceases the proto- plasm is no longer alive, but it is uncertain in most cases whether respiration ends as soon as death occurs or whether it continues for some time afterward. It was stated by Johannsen (8) , by Detmer (6) , and by Pfeffer (17) that in general there is no production of CO^ after death, although Reinke (18) and Brenstein (3) held the opposite view. Buchner (4) showed that yeast which had been treated with acetone and ether and which was incapable of cell division, and in all probabihty dead, could produce CO2 by fermentation. Kostytscheff (id) found that an aerobic plant, Aspergillus niger, treated in this manner was still capable of respiration. Since some of the cells appeared to be alive after the treatment, he used heat to kill them. After this the oxidation was extremely small. This is to be expected as the oxidases are, for the most part, injured or destroyed by heat. Similar experiments have been made on bacteria. Warburg (21) obtained a completely sterile acetone preparation of Staphylococcus which respired about one thirty-sixth as much as the living material. Warburg and Meyerhof (21) found that treatment with acetone and ether had little effect on the consumption of oxygen by unfertihzed sea urchin eggs (although they were completely killed), but the same treatment diminished the consumption of oxygen by fertilized eggs by 90 per cent. Numerous experiments have been made with cells killed by mechanical means (finely ground) or by freezing and thawing. Palladin (16) found that finely ground wheat produced less CO2 than the normal amount, while various plants exposed to —20° C. for some time and then thawed out showed a loss of power to absorb oxygen, but continued to produce CO2. ' A preliminary communication appeared in Proc. Nat. Acad. Sci. 347I [Botanical Gazette, vol. 67 348 BOTANICAL GAZETTE [april Batelli and Stern (2) have found oxidation in finely ground tissue and watery extracts. Their results have been criticized by Warburg. Warburg (21) found that the finely ground red blood cor- puscles of birds consumed less oxygen than intact cells. Unfer- tihzed sea urchin eggs, cytolyzed in distilled water, consumed as much oxygen as the intact eggs but produced no CO2. In fertilized eggs cytolysis reduced the oxygen consumption by 90 per cent or more. A fuller account of the literature seems unnecessary, as it has been summarized by Warburg. It will be noticed that in the cases previously reported respira- tion after death is greatly reduced or entirely lacking. The only instance in which post mortem respiration is greater than in normal tissue is that reported by Loeb and Wasteneys (ii), in which unfertilized sea urchin eggs, cytolyzed by saponin, showed from 3 to 7 times the normal rate of respiration. It is of considerable interest therefore to find that the respiration of Laminaria after death may be much greater than when in its normal condition. The determination of the output of CO2 was made in the fol- lowing manner. The increase in the hydrogen ion concentration of sea water containing pieces of Laminaria (in the dark) served as a measure of the respiration of the tissue. The decrease in PH value was determined by the addition of a suitable indicator (phenolsulphone phthalein) by comparing the colors with those of a series of buffer mixtures containing an equal amount of the same indicator. Each piece of Laminaria was kept for about half an hour in sea water before beginning the experiment. This treatment tended to obviate any effects of the shght wounding (19, 20). The material was then rolled into a scroll and inserted into a Pyrex glass tube (7) fused shut at one end and attached to a paraffined rubber tube at the open end. Sea water, of the temperature of the bath (16=1= i°C.), was placed in the tube and the latter inserted in a black enameled collapsible tin tube in the bath. The sea water surrounding the tissue (in the tube) was renewed several times before beginning the experiment. A definite amount of sea water iqiq] HAAS— respiration 349 (6 cc.) was placed in each of the tubes. The tubes were clamped shut in such a way as to include a very small air bubble (always of the same size) which served as a stirrer. This was sufficiently accurate and was more convenient than paraffined glass beads. After a tube had been in the dark at i6=t i°C. for a definite period, it was removed from the bath, and the contents shaken by inverting the tube several times. The sea water was then poured rapidly into an empty tube of equal diameter, to which the same quantity (3 drops of 0.0 1 per cent aqueous phenolsulphone phthalein to 6 cc. of solution) of indicator was added as had been added to the buffer mixtures. The color was then compared with the colors of a series of buft'er mixtures by the use of a constant source of light (the "Daylight" lamp) and the PH value determined. The same amount of sea water was again added to the tissue in the tube and the tube exposed (at i6=ti°C. in the dark) for the same length of time as before, after which it was removed from the bath and the PH value again determined. This was repeated until the respiration in sea water was approximately constant. Then sea water con- taining the killing agent was substituted for the sea water, and the PH values determined as before after a series of successive periods (each of the same length as the original). In some cases (acetone 17.4 and alcohol 24 . 2 per cent) the kilKng agent extracted from the plant a small amount of pigment which interfered with the color of the indicator.^ This difficulty disappeared after the first two periods, however, as was shown by running pure hydrogen through the solution, after which it returned to the color found in normal sea water containing indicator. This method also showed conclusively that the only acid excreted by the plant was carbonic acid. The methods of kiUing the tissue were various. Sea water containing anesthetics (made up to the conductivity of sea water by the addition of concentrated sea water) was employed in many of the experiments. In this case the respiration was determined for several periods of equal length in sea water (the solution being renewed after each period). The sea water was then replaced by ^ This did not occur with low concentrations of these substances. 350 BOTANICAL GAZETTE [april sea water containing anesthetic and the respiration determined after successive equal periods until death ensued, and for some time thereafter. TABLE I Control for I A: 7 periods (27.5 min. each) in sea water; solution RENEWED AT BEGINNING OF EACH PERIOD og Total ■ Sa.aJ" Time Relative rate of Period Change m PH change inPH=A Change calcul from perioc Relative amou respir = A/] in mm. respiration I 8.1-7.6 =0.5 o-S 0.5 1 .00 27-5 0.5 -^o. 5 = 1. 00 2 8.1-7 6 =0.5 I.O 1.0 I .00 55-0 0.5 -=-0.5 = 1.00 3 8.1-7 6 =0.5 i-S 1-5 1. 00 82. 5 0.5 -=-0.5 = 1.00 4 8.1-7 6 =0.5 2.0 2.0 1. 00 IIO.O 0.5 -f-o.5 = i.oo 5 8.1-7 65 = 0.45 2.45 2-5 0.98 137-5 0.45-^0.5=0.90 6 8.1-7 65 = 0.45 2.90 30 0.96 165.0 0.45^0.5=0.90 7 8.1-7 65 = 0.45 3-35 3-5 0.96 192.5 0.45^-0.5=0.90 TA BLE I A L Change in PH value of sea water produced by respiration of Laminaria during 2 PERIODS (22 MIN. each) IN SEA WATER AND DURING 7 SUBSEQUENT PERIODS IN SEA WATER APPROXIMATELY SATURATED WITH ETHYL BROMIDE Solution Period Change in PH Change in PH during saturation with ethyl bromide = A Change in PH calcu- lated from first 2 periods = B c< 3 II 2 c 6.2 Time in min. Time during saturation with ethyl bromide in min. Relative rate of respiration I 2 3 4 S 6 7 8 8.1 —7.8 =0.30 8.1 —7.8 =0.30 7.32— 6.I6* = I.I6 + 7.32— 6.I6*=I.I6-|- 7.32— 6.69 = 0.63 7.32-7.07 =o.2S 7.32 — 7.21 =0.11 7.32 — 7.32 =0.0 22 44 66 88 no 132 154 176 0 0 22 44 66 88 no 132 tc tt Sea water containing ethyl bromide Sea water containing ethyl bromide Sea water containing ethyl bromide Sea water containing ethyl bromide Sea water containing ethyl bromide Sea water containing ethyl bromide 1. 16 2.32 2.95 3.20 3.31 3-31 0.30 0.60 0.90 1 .20 I -SO 1.80 3.8 3.8 3.2 2.6 2.2 1.8 1.16-^0.30=3.8 1 .16-^0.30 = 3.8 0.63-^-0.30 = 2.1 0.25-^0.30 = 0.8 0.11-^0.30 = 0.4 0.0 -7-0.30=0.0 * Approximate (at this point indicator is not very sensitive to slight changes in acidity). As it was important to know the time of death as accurately as possible, determinations of the electrical conductivity of the tissue were made by the method of Osterhout (12, 13). If the I9I9] HAAS—RESPIRA TION 351 electrical resistance of the normal tissue be called 100 per cent, it is found that after killing the resistance falls to about 10 per cent. When the resistance has fallen to 15 per cent the tissue is for all practical purposes dead, as there is no recovery when it is returned to normal conditions.^ The results of the experiments showing the relative amount and relative rate of respiration of tissue of Laminaria when subjected to various treatments are presented in tabular form. In every case 6-12 or more closely agreeing results were obtained and the data TABLE I B Net electrical resistance of Laminaria in sea water AND IN SEA water APPROXIMATELY SATtlRATED WITH ETHYL BROMIDE, EXPRESSED AS PERCENTAGE OF NET RESISTANCE AT START OF EXPERIMENT AT 20° C. Sea water Sea water approximately satu- rated WITH ETHYL BROMIDE Time in min. Percentage net resistance Time in min. Percentage net resistance 0 100 90.9 87.8 0 ... 100 100 I . loS 79 56 39 30 16 II 200 5 10 i» IS 20 35 60 80 IOCX3 10 5 of a typical case presented. Table I A shows the relative amount and relative rate of respiration as influenced by sea water approxi- mately saturated with ethyl bromide. This is the only experiment in which the PH value of the control differed from the PH value of the sea water containing anesthetic (at the start of the experi- ment). It will be noted that after about 130 minutes no respiration was detected. An examination of table I B, which gives the net electrical resistance of Laminaria, shows that the material can be considered dead before the end of 60 minutes. From table I A it is seen that at the end of 60 minutes the relative rate and relative 3 The determinations were made in part by Professor Osterhout and in part by me. 352 BOTANICAL GAZETTE APRIL amount of respiration are approximately double that of the normal, although the tissue is shown by the method of electrical resistance to be dead. This is brought out very strikingly in fig. i where we plot as ordinates the relative amount of respiration (curve C, table I A), relative rate of respiration (curve B, table I A), net resistance as percentage of that at the start (curve A, table I B) respectively (unbroken lines). When the relative rate of respira- tion has practically reached zero (curve B) the relative amount of °/4rest. REL.RATEOFRESP. REL. AMT. OF RESP. MINUTES Fig. I. — Curves showing effect produced by sea water, approximately saturated with ethyl bromide, upon relative amount and relative rate of respiration, and upon net electrical resistance of Laminaria: curve A, ordinates represent net resistance as percentage of that at start; curve B, ordinates represent relative rate of respiration; curve C, ordinates represent relative amount of respiration (unbroken lines) ; controls in sea water (broken lines) ; each control curve bears same symbol and letter (with a prime) as experimental curve; abscissae represent time in minutes. respiration is above unity. At the end of 60 minutes, when the tissue can be considered dead, the relative rate is seen to be about double that of the normal rate. Table II A shows the effect produced by sea water containing 17.4 per cent (by volume) acetone, made up to the electrical conductivity of sea water by the addition of concentrated sea water, upon the relative amount and relative rate of respiration of Laminaria. At the end of 2 . 5 hours the rate of respiration is still above the normal rate, while the relative amount of respira- igig] HAAS—RESPIRA TION 353 tion is nearly 2. Table II B shows the net electrical resistance. It is seen that the material is dead before the end of 100 minutes. TABLE II Control for II A: 9 periods (21 min. each) in sea water; solution RENEWED AT BEGINNING OF EACH PERIOD Period I 2 3 4 S 6 7 8 9 Change in PH •37-7 •37-7 •37-7 ■37-7 •37-7 ■37-7 •37-7 •37-7 •37-7 .67 = 0.70 .67 = 0.70 .70 = 0.67 .73 = 0.64 .73 = 0.64 .73 = 0.64 .72 = 0.65 . 73 = 0-64 .72 = 0.65 Total change inPH=A 0.70 1 .40 2.07 2.71 3-35 3-99 4.64 5-28 S-93 pH-T3 , PQ 2^3 .2 o o. 0.70 1 .40 2. 10 2.80 3 50 4. 20 4.90 5.60 6.30 cr! I .0 I .0 0.99 0,97 0.96 0.95 0.9s 0.94 0.94 Time in min. 21 42 63 84 los 126 147 168 189 Relative rate of respiration 0.70 0.70 0.67 0.64 0.64 0.64 0.65 0.64 0.65 70=1.0 70=1 .0 70 = 0.95 70 = 0.91 70 = 0.91 70 = 0.91 70 = 0.93 70 = 0.91 70 = 0.93 TABLE II A Change in PH value of sea water produced by respiration of Laminaria during 2 PERIODS (24 min. each) IN SEA WATER, AND DURING 7 SUBSEQUENT PERIODS IN sea WATER CONTAINING 1 7. 4 PER CENT (bY VOLUMe) OF ACETONE Solution Sea water. Sea water 17 -4 per Sea water 17.4 per Sea water 17.4 per Sea water 17.4 per Sea water 17.4 per Sea water 17.4 per Sea water 17.4 per contaming cent acetone containing cent acetone containing cent acetone containing cent acetone containing cent acetone containing cent acetone containing cent acetone Period Change in PH 8.37- 8.37- 8.37- " 37- 37- 37- 37- 37- 37- -7.76=0.61 -7.78 = 0.59 -6.80 = 1.57 -6.75 = 1.62 -7.06 = 1.31 -7.50=0.87 -7.54 = 0.83 -7.62=0.75 -7.74=0.63 3 K o<: ^ a II C fc- t) — 3c 0 S O =« J: rt o 1-57 319 4-5° 5-37 6.20 6-95 7.58 r- P ffl MS 6 0.60 1 .20 1.80 2.40 3 00 3 -So 4.20 § II E o « 2.61 2.66 2.50 2.24 2.07 1-93 1.80 Time in min. 24 48 72 96 120 144 168 192 216 3-- E a H o o 24 48 72 96 120 144 168 Relative rate of respiration 1-57-^0.60 = 2.6 1.62^0.60 = 2.7 1 .31 -i-0.6o = 2.2 0.87-^0.60 = 1.5 0.83^0.60 = 1.4 0.75-^0.60 = 1.3 0.63 -=-0.60 = 1.1 The fact that respiration proceeds here at a rate much above the normal (although death has taken place) is very clearly brought 354 BOTANICAL GAZETTE [april out by comparing the curves for table II A and II B as given in fig. 2. The ordinates represent relative amount of respiration (curve A, table II A), relative rate of respiration (curve C, table II A), net resistance as percentage of that at the start (curve B, table II B) respectively (unbroken Hues). The relative rate and relative amount of respiration at the end of over 2 . 5 hours are still much above the normal even though the measurements of electrical resistance have shown the tissue to be dead before 100 minutes. REL.RATEOFRESP. %REST. REL.AMT-OFRESP. 60 120 MINUTES Fig. 2. — Curves showing effect produced by sea water containing 17.4 per cent (by volume) of acetone upon relative amount and relative rate of respiration, and effect produced by sea water containing 16. 2 per cent of acetone upon net electrical resistance of Laminaria: curve A, ordinates represent relative amount of respiration; curve B, ordinates represent net resistance as percentage of that at start; curve C, ordinates represent relative rate of respiration (unbroken hnes) ; controls in sea water (broken hnes); each control curve bears same S3rmbol and letter (with a prime) as experimental curve; abscissae represent time in minutes. Table III A shows the effect produced by sea water containing 24.2 per cent (by volume) of ethyl alcohol (made up to conduc- tivity of sea water by the addition of concentrated sea water). In fig. 3 the ordinates represent: relative amount of respiration (curve A, table III A), relative rate of respiration (curve C, table III A), net resistance as percentage of that at the start (curve B, table III B) respectively (unbroken lines). If we con- sider the material dead at the end of 90 minutes, we find that the I9I9] HAAS—RESPIRA TION 355 rate of respiration is much above the normal rate, while the relative amount of respiration is above 2. TABLE II B Net electrical resistance of Laminaria in sea water AND IN SEA water CONTAINING 1 6. 2 PER CENT (BY volume) of ACETONE, EXPRESSED AS PERCENTAGE OF NET RESISTANCE AT START OF EXPERIMENT AT 15 .4° C. Sea water Sea water containing 16.2 PER cent acetone Time in min. Percentage net resistance Time in min. Percentage net resistance 0 100 90.9 87.8 0 100 100 s 10 105.5 94-3 46.6 28.4 22.7 20.2 18.2 16.2 13.2 II .1 8.7 6.7 200 20 ?o 40 50 60 70 80 . 100 200 ■JOO TABLE III Control for III A: 8 periods (30.25 min. e.\ch) in se.\ water; solution RENEWED AT BEGINNING OF EACH PERIOD Period Change in PH Total change in PH=A Change in PH calculated from first 2 periods = B Relative amount of respiration = A/B Time in min. Relative rate of respiration I 2 3 4 5 6 7 8 7 7 7 7 7 7 7 7 90-7 90-7 90-7 90-7 90-7 90-7 90-7 90-7 53 = 0.37 53 = 0.37 53 = 0-37 54 = 0.36 54 = 0.36 55 = 0.35 55 = 0.35 58 = 0.32 0.37 0.74 I. II 1.47 1.83 2.18 2.53 2.85 0.37 0.74 I . II 1.48 1.85 2.22 2-59 2.96 I.O I .0 I .0 0.99 0.99 0.98 0.97 0.96 30.25 60.50 90.75 121.00 151.25 181.50 211.75 242.00 0.37-^0.37 = 1.0 0.37^0.37 = 1.0 0.374-0.37 = 1.0 0.36-7-0.37 = 0.97 o.36-^o.37 = o.97 0.35-^0.37 = 0.95 0.35^0.37 = 0.95 0.324-0.37 = 0.87 Table IV shows the effect produced upon the relative amount and relative rate of respiration of Laminaria by sea water contain- ing 3 . 2 per cent (by volume) of formaldehyde. The solution was 356 BOTANICAL GAZETTE [april made up to the conductivity of sea water by the addition of con- centrated sea water. The free acid of the formaldehyde was first neutraHzed by the addition of a Httle sodium carbonate. This is allowable for the purposes of the present investigation, for its effect would be to make the amount of CO2 produced appear some- what less than was actually the case. At the end of 4 hours the relative rate of respiration was still above the normal, while the 7oREST. REL.RATEOFRESP. RELJIMT.OJFRESP. MINUTES Fig. 3. — Curves showing effect produced by sea water containing 24.2 per cent (by volume) of ethyl alcohol upon relative amount and relative rate of respiration and upon net electrical resistance of Laminaria: curve A, ordinates represent relative amount of respiration; curve B, ordinates represent net resistance as percentage of that at start; curve C, ordinates represent relative rate of respiration (unbroken lines) ; controls in sea water (broken lines) ; each control curve bears same symbol and letter (with a prime) as experimental curve; abscissae represent time in minutes. relative amount of respiration was much above the normal. At this concentration of formaldehyde Laminaria is practically dead in 180 minutes. In table IV, however, after 280 minutes the rela- tive rate of respiration of Laminaria is still above normal, while at 180 minutes the relative rate is far above normal. For purposes of comparison other methods of kilhng were tried. By making preliminary conductivity experiments with Laminaria, it was found that when it is dried upon cheesecloth in the sunlight in a current of dry air, we can consider the tissue practically dead iqiq] HAAS—RESPIRA TION 357 in 135 minutes. After such treatment the material becomes green. It was placed in sea water for 14 minutes; it lost its crispness and TABLE III A Change in PH value of sea water produced by respiration of Laminaria during 2 PERIODS (31.25 MIN. each) IN SEA WATER AND DURING 6 SUBSEQUENT PERIODS IN SEA WATER CONTAINING 24.2 PER CENT (bY VOLUME) OF ETHYL ALCOHOL Solution Period Change in PH < ^ 0. lU so u Change in PH calcu- lated from first 2 periods = B o« 3 II E.2 Hi Z 'Z ^ Time in min. Time (in min.) of exposure during the 6 later periods Relative rate of respiration Sea water I 2 3 4 5 6 7 8 7.90-7-43 =0.47 7.90-7.43 =0.47 7.90 — 6.i6* = i.74 7.90-6.75 =1.15 7.90-7-25 =0.65 7-90-7-55 =0.35 7.90 — 7.80 =0.10 7.90 — 7.84 =0.06 31-25 62.50 93.75 125.00 156.25 187.50 218.75 250.00 0 0 31-25 62.50 93-75 125.00 156.25 187-SO 11 (( Sea water containing 24.2 per cent ethyl alcohol. . Sea water containing 24:2 per cent ethyl alcohol . . Sea water containing 24 . 2 per cent ethyl alcohol. . Sea water containing 24 . 2 per cent ethyl alcohol . . Sea water containing 24.2 per cent ethyl alcohol. . Sea water containing 24.2 per cent ethyl alcohol. . 1-74 2.89 3-54 3-89 3-99 4-05 0.47 0.94 1. 41 1.88 2.3s 2:82 3-S7 307 2.51 2.07 1.69 1-43 1.74-^0.47=3.7 I.^5-^o.47=2.4 0.65-^0.47 = 1.4 0.35^0.47=0.7 0.10-^0.47 =0.2 0.06-^0.47 =0.1 * Appro.ximate (at this point the indicator is not very sensitive to changes in PH). TABLE III B Net ELECTRICAL RESISTANCE OF Laminaria in sea water AND IN SEA water CONTAINING 24 PER CENT (bY volume) of ethyl alcohol, EXPRESSED AS PER- CENTAGE OF NET RESISTANCE AT START OF EXPERI- MENT AT 13.3° C. Sea water Sea water containing 24 per cent ethyl alcohol Time in min. Percentage net resistance Time in min. Percentage net resistance 0 100 98 0 100 240 10 28.8 30 90 18.6 15-2 14.9 I3-S II. 8 II-3 120 150 180 210 became flaccid. The relative amount and relative rate of respira- tion of pieces of Lammaria were determined before and after the 3S8 BOTANICAL GAZETTE [aprtl drying treatment (which killed the tissue). The results are given in table V A, the control data being given in table V. The rela- tive amount of respiration after the treatment was very high and after 2 hours was still at 3. The relative rate of respiration after the treatment was very high at the start but gradually declined to normal in 2 hours. TABLE IV Control for IV A: 15 periods (21 min. each) in sea water; solution RENEWED AT BEGINNING OF EACH PERIOD Period Change in PH Total change inPH=A Change in PH calculated from first 2 periods = B Relative amount of respiration =A/B Time in min. Relative rate of respiration I 2 3 4 5 6 7 8 9 10 II 12 13 14 IS 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 37-7 37-7 37-7 37-7 37-7 37-7 37-7 37-7 37-7 37-7 27-7 37-7 37-7 37-7 37-7 80 = 0.57 90 = 0.47 90 = 0.47 90 = 0.47 92 = 0.45 92 = 0.45 92=0.45 92 = 0.45 90 = 0.47 90 = 0.47 90 = 0.47 90 = 0.47 90 = 0.47 90 = 0.47 90 = 0.47 057 1.04 1.98 2.43 2.88 3-33 3.78 4.25 4.72 5 19 S-66 6.13 6.60 7.07 0.52 1.04 i.S6 2.08 2.60 3.12 3 64 4.16 4.68 5.20 S-72 6.24 6.76 7.28 7.80 I.I I .0 0.96 0-9S 0-93 0.92 0.92 0.90 0.90 0.91 0.91 0.91 0.91 0.91 0.91 21 42 63 84 126 147 168 189 210 231 252 273 294 3IS 0.57-4-0.52 = 1.10 0.47-4-0.52 = 0.90 0.47^0.52 = 0.90 0.47-4-0.52 = 0.90 0.45 -=-0.52 = 0.87 0.45^0.52 = 0.87 0.45^0.52 = 0.87 0.45^0.52=0.87 0.47-4-0.52 = 0.90 0.47-4-0.52 = 0.90 0.47^0.52=0.90 0.47-4-0.52=0.90 0.47-4-0.52=0.90 0.47-4-0.52=0.90 0.47-4-0.52 = 0.90 Another method used in killing the tissue was by placing it in running tap water. A preliminary determination of the electrical resistance showed that 22 hours were more than sufhcient to kill the tissue. The experiment was begun at 11:18 a.m. and at 7:50 P.M. the tissue was still somewhat ahve, but at 9:25 A.M. next day the tissue had in all probability been dead for some time. The respiration was then determined before and after exposure to running tap water for 19 hours. The results are given in table VI, the data for the control being given in table V A. In table VI it is obvious that no respiration of the tissue was observable after it had been in tap water for 19 hours. There is of course the possibility that the rise and decHne of the respiration after death was so rapid as to escape observation if the tissue had been dead much before the end of 19 hours. iqiq] HAAS—RESPIRA TION 359 TABLE IV A Change in PH value of sea water produced by respiration of Laminaria during 2 PERIODS (23.5 MIN. each) IN SEA WATER AND DURING 13 SUBSEQUENT PERIODS IN SEA WATER CONTAINING 3.2 PER CENT FORMALDEHYDE Solution Period Change in PH Mi_l durin orma 3 u It 0 £2< 3 II 2 c E.2 Time in min. C I- ., c £"0 5;-ri . — 3.ii "> '£. "Z. nge i ted f perio e of rmal min ^S-S S.-^ c -3 i- E^.S u U P« H 23-S 47 -o 70. S 0 0 1-47 0.5s 2.67 23-5 2.94 1 .10 2.67 940 47.0 4.26 1.65 2.58 II7-S 70.5 S.48 2. 20 2.49 141 .0 94.0 6.52 2-75 2.37 164. s 117. 5 7.3g 3-30 2.24 188.0 141 .0 8.21 3.85 2.13 211. s 164.5 9.00 4.40 2.04 23s -o 188.0 9-77 4-95 1.97 258.5 211.5 10.49 5.50 1 .90 282.0 235-0 II. 16 6.0s 1.84 305-5 258.5 11-73 6.60 1-77 329.0 282.0 12. 25 7-15 I. 71 352.5 305.5 Relative rate of respiration Sea water. Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2. per cent formaldehyde' Sea water containing 3 . 2) per cent formaldehydej Sea water containing 3 . 2| per cent formaldehyde^ Sea water containing 3 . 2' per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea w'ater containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde' Sea water containing 3 . 2, per cent formaldehyde; Sea water containing 3 . 2; per cent formaldehyde! 10 II 12 13 14 IS 8.37- 37- 37- 37- 37- 37- 37- 37- 37- 37- 37- 37- 37- 37- 37- 7.82 = ■7.82 = •6.90 = ■6.90 = ■7.05 = •715 = ■7-33- ■7.50 = ■7-55 = •7.58 = ■7.60= ■7.65 = •7.70= ■7.80= ■7.85= ^0.55 = 0.55 = 1.47 = 1-47 = 1.32 = 1.22 -1 .04 = 0.87 ^0.82 = 0.79 = 0.77 = 0.72 = 0.67 = 0.57 = 0.52 1.47^0.55=2 1-47-^0.55 = 2 1.32-^0.55 = 2 I.22-r0.55=2 1.04-^0.55=1 0.87-^-0.55=1 0.82-^0.55=1 0.79^0.55 = 1 0.77-^0.55=1 0.72-^0.55=1 o.67-^o.55=I 0.57-^0.55 = 1 o.52-^o.5s=o 7 7 4 2 9 6 S 4 4 3 2 o 95 TABLE V Control for V A: 3 periods (25.75 min. each) in sea water; between second AND third periods AN INTERVAL OF I9 HOURS DURING WHICH TISSUE WAS BATHED IN RUNNING SEA WATER Period Change in PH ToUl change inPH=A Change in PH calculated from first 2 periods = B Relative amount of respiration = A/B Time in min. Relative rate of respiration I 2 3 8.0—7.58 = 0.42 8.0 — 7.58 = 0.42 8.0 — 7.60 = 0.40 0.42 0.84 1.24 0.42 0.84 1.26 1. 00 I .00 0.98 25-75 51-50 77-25 o.42-i-o.42 = i.oo o.42-f-o.42 = i .00 o.40-j-o.42 = o.95 360 BOTANICAL GAZETTE [april By determining the electrical resistance it was found that Laminar ia is killed by exposure to 35° C. for 70 minutes. The respiration before and after such exposure was then determined. During the treatment at 35° C. the material was removed from the tubes and placed in a large volume of the sea water kept at 35° C. The results are given in table VII A. After the exposure to 35° C, the relative amount and rate of respiration had fallen considerably below the normal. This might be expected on the ground that oxidizing enzymes are injured or destroyed by heat. TABLE V A Change in PH value of sea water produced by respiration of Laminaria during 6 PERIODS (30.5 MIN. each) IN SEA WATER; AT END OF SECOND PERIOD MATERIAL DRIED IN CURRENT OF AIR IN SUN FOR 1 39 MINUTES; MATERIAL THEN PLACED IN SEA WATER AT 22° C. FOR 15 MINUTES BEFORE BEGINNING THIRD PERIOD. Period Change in PH <; II c C u Change in PH before drying, calculated from first 2 periods = B § II E 0 >i T Time in min. Time (in. min.) of exposure after drying Relative rate of • respiration I 8.15 — 8.0 =0.15 8.15-8.0 =0.15 8.15-7.20 = 0.95 8.15-7.70 = 0.45 8.15-7.85 = 0.30 8.15 — 8.00 = 0.15 30.5 61.0 91. 5 122.0 152.5 183.0 0 0 30.5 61.0 915 122.0 2 3 4 s 6 0-9S 1 .40 1.70 1.85 O.I5 0.30 0.45 0.60 6.3 4.6 3-8 3-1 O.954-0. 15 = 6.3 0.454-0.15=3.0 0.304-0.15 = 2.0 0.154-0.15 = 1.0 It is well known that. severe injury causes a considerable rise in the respiration, and it seemed desirable to make such experi- ments with Laminaria. After the normal respiration of a piece of tissue had been determined, the material was removed from the tube and finely macerated (by means of the jagged end of a tube of Pyrex glass) on a piece of tested filter paper. The minced Laminaria was put back into the tube and rinsed 6-10 times with sea water until none of the liberated pigment could be distinguished in the sea water. Fresh sea water was then added and the respira- tion determined. The results are given in table VIII, the control data being given in table V. In table VIII it will be observed that the relative amount and relative rate of respiration are both I9I9] HAAS—RESPIRA TION 361 TABLE VI A Change in PH value of sea water produced by respiration of Laminaria dur- ing 3 PERIODS (23 MIN. each) IN SEA WATER; AT END OF SECOND PERIOD MATERIAL PL.^CED IN RUNNING TAP WATER FOR 1 9 HOURS; MATERIAL THEN PLACED IN SEA WATER AT 22° C. FOR 34 MINUTES BEFORE BEGINNING THIRD PERIOD Period Change in PH Change in PH after exposure to tap water = A Change in PH before exposure to tap water, calculated from first 2 periods = B ■o« V H >.§ ■ri'S. 1; I- Time in min. Time (in min.) of exposure after exposure to tap water Relative rate of respiration I 8.0-7.7=0.3 8.0-7.7 = 0.3 8.0 — 8.0 = 0.0 23 46 69 0 0 23 2 3 0 03 0 0.0-^0.3 = 0.0 TABLE VII Control for VII A: 4 periods (25.5 min. each) in sea water; betw^een second AND THIRD PERIODS MATERIAL KEPT IN SEA WATER AT 1 6° C. FOR 70 MINUTES Period Change in PH Total change inPH=A Change in PH calculated from first 2 periods =B Relative amount of respiration =A/B Time in min. Relative rate of respiration I 2 3 4 8.0-7.57 = 0.43 8.0—7.58 = 0.42 8.0—7.58 = 0.42 8.0—7.58 = 0.42 0.43 0.85 1.27 1.69 0.425 0.85 1-275 1.70 1. 01 1 .00 1. 00 0.99 25-S Si-o 76.5 102.0 0.43-7-0.425 = 1.01 0.42-^0.425=0.99 0.42-^0.425=0.99 0.42-e-0. 425=0. 99 TABLE VII A Change in PH value of sea water produced by respiration of Laminaria during 4 PERIODS (28 MIN. each) IN SEA WATER; BETWEEN SECOND AND THIRD PERIODS MATERIAL PLACED IN LARGE VOLLTME OF SEA WATER AND KEPT AT 35° C. FOR 70 MINUTES Period Change in PH Change in PH after exposure follow- ing second period = A Change in PH im- mediately after second period = B 0^ c< 1" II 13 >- Time in min. Time (in min.) of exposure during last two periods Relative rate of respiration I 8.0-7.25=0.75 8.0 — 7.28 = 0.72 8.0—7.80 = 0.20 8.0-7.85=0.15 28 56 84 112 2 3; 4 0.20 0-35 0.73S 1.470 0.27 0.24 28 S6 0. 20-i-O. 75=0. 26 o.I5-^o.75 = o.2o 362 BOTANICAL GAZETTE APRIL more than doubled, but gradually decline. After i hour the rela- tive rate of respiration was still above the normal. In this case the time of death could not be determined. The experiments show that although the rate of respiration may be maintained for a time after death, it gradually falls off and eventually becomes very small. The question arises whether this falling off is due to exhaustion of the supply of oxidizable material or not. It is clear that when respiration has practically ceased there is a considerable amount of organic material left, but it is by no means certain that this material is such as to be easily oxidized by the ordinary processes which produce CO2. On the TABLE VIII Change in PH value of sea water produced by respiration of Laminaria during 5 PERIODS (31.25 MIN. each) IN SEA WATER; BETWEEN SECOND AND THIRD PERIODS MATERIAL FINELY MINCED Period Change in PH ■*-» (^ II a w) u Change in PH before mincing, calcu- lated from first 2 periods = B a< 3 II 2 = E 0 ->l 13 >- Time in min. Time (in min.) of exposure after mincing Relative rate of respiration I 8.1 — 7.70 = 0.40 8.1 — 7.70 = 0.40 8.1 — 7.05 = 1.05 8.1-7.65 = 0.45 8.1 — 7.90 = 0.20 31-25 62.50 93-75 125.00 156.25 2 3 4 s 1-05 I so 1.70 0.40 0.80 1 . 20 2.6 1-9 1-4 31-25 62.50 93-75 I.05-:-0.40=2.6 0.45-^0.40 = 1.1 0.20-7-0.40 = 0.5 other hand, we must consider the possibiHty that the production of CO2 falls off because the supply of oxidizing enzymes is used up. Various observers have found that these enzymes may be used up (or inactivated) during oxidation (i, 9). If the process of oxida- tion involves the cooperation (or successive action) of various enzymes the inactivation of any one of them might bring the whole process to a standstill. Warburg (21), as the result of extensive study, has come to the conclusion that the rate of oxidation depends on the amount of "structure" which the cell possesses. If the "structure" is par- tially or completely destroyed the oxidation diminishes in propor- iqiq] HAAS— respiration 363 tion, except in rare cases (as in the unfertilized sea urchin egg) where the oxidation is independent of structure. He states that the latter case disposes of the '^ reaction chamber" theory of cell structure, according to which the substances necessary for oxidation are separated by the semipermeable membranes of the cell in such a way as to regulate the speed of oxidation, for these substances can be completely mixed, as in the cytolysis of the unfertilized sea urchin egg, without any change in the rate of oxidation. Warburg's treatment of the "reaction chamber" hypothesis seems to rest upon a misunderstanding. It is quite possible that in the cytolysis of the sea urchin egg the "reaction chambers" are not destroyed, since each of the fine granules into which the egg is resolved by cytolysis may be such a "reaction chamber" sur- rounded by a semipermeable surface.'^ In case some or all of the reaction chambers are destroyed by the treatment, because they are larger or for any reason more sensitive to the treatment, a change in the rate of oxidation may be expected (either an increase or decrease, according to circumstances). Warburg himself states that where an increase of chemical action results from the injury the "reaction chamber" hypothesis seems to be justified. This is precisely what the writer finds. Increase of oxidation as the result of injury (although not as the result of death) has previously been recorded by many observers (5). The "reaction chamber" h}^othesis has much in its favor. An especially good example is the bitter almond, which at once pro- duces HCN upon injury. In this case the reacting substances are known and we cannot escape the conclusion that previous to injury they fail to react because they are kept apart by structures in the cell. In some cases the mingling of substances, owing to the break- ing down of such separating structures, can distinctly be seen under the microscope. This is the case with the marine alga Griffithsia, as described by Osterhout (14, 15). When cells of this alga are injured by poisons (NH4CI), or mechanically, or by cytolysis with dilute sea water, the chromatophores (which contain a soluble red pigment) become permeable and the pigment can be seen passing 4 The existence of an actual membrane is unnecessary. 364 BOTANICAL GAZETTE [april out into the surrounding cytoplasm. It would seem, therefore, that in the absence of a better explanation^ the reaction chamber hypothesis might serve a useful purpose. Summary The respiration of Laminaria after death may be considerably greater than in its normal condition. This is the case when it is killed by alcohol, acetone, formaldehyde, and ethyl bromide, as well as by drying and by other methods. Laboratory of Plant Physiology Harvard University LITERATURE CITED 1. Bach, A., and Chodat, R., Die Oxydationsvorgange in der lebenden Zelle. Biedermann's Zentralbl. Agrik. Chem. 37:168. 1908. 2. Batelli, F., and Stern, L., Einfluss der mechanischen Zerstorung der Zellstruktur auf die verschiedenen Oxydationsprozesse im Tiergewebe. Biochem. Zeit. 67:443. 1914 (see literature quoted). 3. Brenstein, G., Producktion von CO2 durch getotete Pflanzentheile. Dissert. Rostock. 1887. 4. BucHNER, E., BucHNER, H., and Hahn, M., Die Zymasegarung. Berlin. 1903. 5. Czapek, F., Biochemie der Pflanzen. 2:400!?. 1905. 6. Detmer, W., tJber physiologische Oxydation im Protoplasma der Pflanzen- zellen. Bot. Zeit. 46:41. 1888. 7. Haas, A. R., A simple and rapid method of studying respiration by the detection of exceedingly minute quantities of carbon dioxide. Science N.S. 44:105. 1916. 8. JoHANNSEN, W., tJber Fortdauer der "Athmungs Oxydation" nach dem Tode. Bot. Zeit. 45:762. 1887. 9. Kastle, J. H., U.S. Pub. Health and Mar. Hosp. Service. Bull. 51. 1909; Bull. 59. 1910. 10. KosTYTSCHEFF, S., tJber Atmungsenzyme der Schimmelpilze. Ber. Deutsch. Bot. GeseUs. 22:207. 1904. 11. LoEB, J., and Wasteneys, H., The influence of hypertonic solution upon the rate of oxidations in fertilized and imfertilized eggs. Jour. Biol. Chem. 14:469. 1913. 5 The explanation suggested by Warburg (21), that the oxidizing substances are bound up by the structure, seems too vague for discussion. 1919] HAAS— RESPIRATION 365 12. OsTERHOUT, W. J. v., The permeability of protoplasm to ions and the theory of antagonism. Science N.S. 35:112. 1912. 13. , The decrease of permeabiUty produced by anesthetics. BoT. Gaz. 61:148. 1916. 14. , The organization of the cell with respect to permeability. Science N.S. 38:408. 1913. 15. , The nature of mechanical stimulation. Proc. Nat. Acad. Sci. 2:237. 1916. 16. Palladix, V. I., Die Arbeit der Atmungsenzyme der Pflanzen unter verschiedenen Verhaltnissen. Hoppe-Seyler's Zeitschrift Physiol. 47:407. 1906. 17. Pfeffer, W., Oxydationsvorgange in lebenden Zellen. 1889 S. 501. 18. Reinke, J., Zur Kenntniss der Oxydationsvorgange in der Pflanze. Ber. Deutsch. Bot. Gesells. 5:216. 1887 ; Einleitung in die theoretische Biologie. S. 637. 1901. 19. Richards, H. M., The respiration of wounded plants. Ann. Botany 10:551. 1896. 20. , The evolution of heat by wounded plants. Ann. Botany 11:29. 1897. 21. Warburg, 0., Beitrage zur Physiologie der Zelle, inbesondere iiber die Oxydationsgeschwindigkeit in Zellen. Ergebnisse der Physiologie 14:318 fif. 1914. BRIEFER ARTICLES GEORGE FRANCIS ATKINSON (with portrait) In the death of George Francis Atkinson on November 14, 1918, America lost one of her great botanists. Born in the little village of Raisinville, Monroe County, Michigan, on January 26, 1854, he received his preliminary collegiate training in Olivet College in that state. From there he went to Cornell University, where he took the degree of Ph.B. in 1885. Immediately upon graduation he became assistant professor of general zoology at the Uni- versity of North Carolina. The following year he was made associate professor, re- maining there until 1888, when he was called to a full professorship in botany and zoology in the University of South Carolina. In 1889 he was appointed professor of biology and botany in the Alabama Polytechnic Insti- tute, where he remained until 1892. In 1892 he was called to Cornell University as assistant professor of botany, became associate professor in 1893, and upon the death of Professor Pren- tiss in 1896 was made full professor and head of the department. He was also for many years the botanist of the Cornell Agricultural Experiment Station. He con- tinued head of the Department of Botany in the Arts College in Cornell University until his death. Upon the request of an organization of his former students, the Board of Trustees of the University in 191 7 relieved him of all teaching Botanical Gazette, vol. 67] [366 1919] BRIEFER ARTICLES 367 and administrative burdens in order that he might devote his entire time and energies to the completion of his monographic studies on the fleshy fungi of North America. In the vigorous and enthusiastic pursuit of this enterprise he made an extensive collecting trip through the Atlantic seaboard states from Florida to the District of Columbia in the spring and summer of 1918. Returning to Ithaca in September he left after an all too short rest for the Pacific Coast, there to pursue his studies of the fleshy fungus flora of that region. On this trip he was without any assistant and most of the time alone. A former student, Dr. Adeline Ames, spent a few days collecting with him in the region about Tacoma, Washington, shortly before his death. Urged by the wonderful variety and abundance of the forms he found and an indomitable enthusiasm for his work, he apparently labored beyond his strength and exposed himself to unusual hardships. He took a hea\y cold from exposure on a trip into the mountains near Tacoma, Washington, which rapidly developed into influenza followed by pneumonia. He died in the City Hospital at Tacoma far from friends and kindred, another martyr to the cause of botanical science. Professor Frye of the University of Washington upon news of his death went immediately to Tacoma to learn the details and to rescue his notes and collections. Dr. Ames also went again to Tacoma shortly thereafter. Thanks to their generous and painstaking efforts we have a full account of Professor Atkinson's last days. This record gives us a wonderful insight into the man's devotion to his work and a fuller appreciation of his greatness. Interested primarily in entomology in the early days of his career, he soon turned to the botanical field, and especially mycology, in which perhaps he has made his most notable investigations. He was without doubt the greatest American student of the fleshy Basidiomycetes. His numerous contributions in this field and a remarkably large and excep- tionally excellent collection of photographs, together with specimens and notes on these forms not only American but European attest his preeminence in this field. He was, however, a botanist of wide interests and his investigations and writings touch nearly every branch of this broad field. A true philosopher, he gave to his contributions that philosophical character and flavor which is the mark of scientific genius. He was the author of many textbooks, notable among which are several elementary and college textbooks of general botany, " The biology of ferns," and "Mushrooms, edible, poisonous, etc." He made many 368 BOTANICAL GAZETTE [april contributions to the botanical journals, not only of America but also of England, France, and Germany. His travels in Europe, his extensive correspondence, and the stu- dents that came from the ends of the earth to study in his laboratories have made his name familiar in the botanical institutions of every land. As a delegate to the International Botanical Congresses of 1905 and 19 10 held in Vienna and Brussels respectively, he made for conservatism in botanical nomenclature. A charter member of the Botanical Society of America and at one time president, he has been for a generation one of the leaders of American botanical thought and activity. He was a fellow of the American Association for the Advancement of Science, a member of the American Philosophical Society, and in 191 8 was elected to the National Academy of Science. He was for years an associate editor of the Botanical Gazette. He was also a member of the honorary societies of Phi Beta Kappa and Sigma Xi. To those of us who knew him intimately as teacher and friend, our days with him in field and laboratory will ever remain a happy and a grateful memory. He was a master of the highest scientific ideals, unsparing in his criticisms, just and fair in his judgments, generous with help and suggestions, a good friend and a genial companion. — H. H. Whetzel, Cornell University, Ithaca, N.Y. CURRENT LITERATURE BOOK REVIEWS Manual of tree diseases' This first wholly American work in book form on the diseases of forest trees is one of "The Rural Manuals" edited by L. H. Bailey, and in conformity with the general plan of this series has been written primarily for the general public. Insect and other animal injuries are not included. The treatment of the subject throughout is simple and direct ; the diseases are concisely described, and methods of control indicated. The first 4 chapters deal with such maladies of biotic and abiotic origin as are common to many kinds of trees and are respectively entitled "Seedling diseases and injuries," "Leaf diseases and injuries," "Body and branch diseases and injuries," and "Root diseases and injuries." Chapters v-xxxii are devoted to an account of the more "specific" diseases, one chapter to each generic host group, beginning with the alders. The arrangement of the chapters is alphabetical according to the EngHsh host group names. Two chapters foUow, one. on "Tree surgery," the other on "Spraying and dusting for leaf diseases." The book is equipped with a glossary, a general bibliography of tree diseases, and an excellent index. This work, although not intended as a textbook, wiU be welcomed by all students of plant pathology because it is the only summary available of the diseases of the forest trees of the United States and of Canada, and because it includes many classified references to the literature. The writing of the book reveals the limitations of forest pathology in America; the number of workers in this field has been small, the subject matter is as yet largely unexplored, and the applications of the results so far attained have been restricted. The author clearly recognizes these facts, and does not fail to point out the direction investigations should foUow; in so doing he makes a contribution of prime importance. — J. H. Faull. MINOR NOTICES Our national forests. — The period of reconstruction not less than the progress of the war has directed, in a special manner, the attention of our people to their natural resources and to the desirability of properly utilizing and conserving them. Thus no more timely moment could be chosen for the pub- lication of some account of our forest wealth as shown in the establishment and ' Rankin, W. Howard, Manual of tree diseases, pp. 398. figs. 70. 1918. New York: Macmillan Co. 369 370 BOTANICAL GAZETTE [april management of our national forests. Boerker^ has collected and organized a mass of scattered data and presented them in a very readable form. While particularly well suited in its style of presentation to appeal to the general public, it wiU prove equally welcome to foresters and botanists who wish to know the history of the organization of these forests and the different forms of administration under which they have attained their present dimensions. The addition of a bibliography would have added much to the scientific value of the volume without detracting from its popular interest. It may also be criticized because of the lack of a suitable index to facilitate reference; but on the whole the work is well done, the material has been well organized, is attractively presented, and so far as the reviewer is able to judge the data are entirely accurate and reliable. — Geo. D. Fuller. Grasses and grasslands of South Africa. — In order to facilitate the study of the extensive grasslands of South Africa, Bews^ has prepared a series of keys for the identification of the 500 species of grasses which form so conspicuous a portion of the flora of that part of the continent. These keys seem to be well suited to serve the purpose for which they are intended, but the other parts of the volume are of far more interest to the American reader. In them are discussed: (i) the structural and ecological characteristics of the principal species; (2) general character of the grasslands and the development of the various association types; and (3) economic application of the ecological prin- ciples involved. It is interesting to find types comparable to the "short grass," "wire grass," and "prairie grass" of North America, as well as a tall coarse Andropogon association, this last developing upon potential woodland areas, and a mountain tussock grassland. The discussion of the successional relations of these and other association types into which grasses enter gives a compre- hensive general sketch of the plant communities of the major portion of South Africa. In the final chapter the feeding value of the different types of grassland, as weU as the comparative merits of native and introduced species, is discussed. The effect upon the productivity of various types of grassland by various kinds of grazing and the results from grass burning are considered and some of the ecological problems involved are pointed out. An appendix contains a list of Enghsh, Dutch, Zulu, and Sesuto names of the more important species. — Geo. D. Fuller. NOTES FOR STUDENTS Vegetation of Cape Breton. — Separated from the mainland of Nova Scotia by a narrow strait, the island of Cape Breton Ues between the Gulf of St. Lawrence and the Atlantic in latitude 45-47° N. It possesses a climate ^BoERKER, Richard, H.D., Our national forests, pp. lxix+238. ^g5. 80. 1918. New York: Macmillan Co. 3 Bews, J. W., The grasses and grasslands of South Africa. 8vo. pp. if)i. figs. 24. map. Pietermaritzburg: Davis & Sons. $2.00 (postpaid from author). 1919] CURRENT LITERATURE 371 characterized by long winters and short cool summers, the extremes of tem- perature being modified by the close proximity of the ocean. A rainfall of 50 in. per year and frequency of fogs make the water supply sufficient for a luxuriant vegetation, which has been carefully studied by Nichols.'' He finds two climatic forest formations represented, the deciduous type upon the low- lands which fringe the coast, and the coniferous type upon the granite uplands which occupy the entire interior portion of the island. These are about 1000 ft. above sea level and form a slightly undulating glaciated surface. The lowlands show many associations depending upon the stage of develop- ment attained, and these variations and the successive stages which have led to their development are carefully discussed and the climax shown to be a forest dominated by beech, sugar maple, and hemlock, together with small quantities of Betula lutea, Picea canadensis, and Abies balsamea. The abundant details of these studies cannot be noticed in a brief review, but two problems in the relationship between the deciduous and evergreen elements of the vegetation are decidedly interesting. It has been found that upon the destruction of the deciduous forest by culling or burning it is succeeded by a coniferous stand dominated by Abies balsama and Picea canadensis; and further that the climax deciduous forest possesses a very considerable percentage of small Abies balsamea which never seem to succeed in competition with the other tree members of the association. Nichols presents evidence showing that the balsam fir is fairly shade tolerant, and that its lack of success is due to its short life, maturity being attained in about one century, and to its great susceptibility to fungus diseases. It seems evident that the coniferous forest dominated by Abies and Picea is the cHmatic rather than the edaphic climax of all portions of the island exceeding 700 ft. in elevation. The factors which appear to differentiate the climate of the uplands from that of the lowlands are the greater extremes of temperature and the greater humidity due to fogs and low-hung clouds which frequently envelop the more elevated areas. This upland forest is of decided importance in the production of pulp wood, its contents being estimated at 12,000,000 cords. Upon the more exposed parts of the uplands are developed "the barrens," closely resembUng the tundras of the subarctic. The low vegetation of "the barrens" varies from a degenerate coniferous forest of the Krummholz type, where the distorted trees are limited in height to the thickness of the snow cover, to coniferous and ericaceous heaths, and to bogs of varied character. These bogs occupy con- siderable portions both of the lowlands and "the barrens," their most striking form being the raised peat bogs of the latter region, which have received careful attention, so that many problems connected with their development have been elucidated. ^ Nichols, Geo. E., The vegetation of northern Cape Breton Island, Nova Scotia. Trans. Conn. Acad. 22:249-267. figs. 70. 1918. 372 BOTANICAL GAZETTE [april Nichols finds the bogs of the raised type, corresponding to the "Hoch- moors" of Europe, occurring upon this continent in Newfoundland and those parts of eastern Canada and Maine which are in close proximity with the sea coast. The cUmatic factors necessary for their development are abundant precipitation, relatively low atmospheric humidity, cool summers, and the absence of extremely low temperatures such as prevail farther inland. One of the necessary edaphic factors is an impervious substratum, here furnished by the Laurentian rocks. This is vitally important, since the source of water supply is the rainfall and not springs, as some have assumed. The early stages of these raised bogs are not essentially different from those obtaining in bogs of the more common and familiar type, but their subsequent development is dependent upon the presence of distinct types of Sphagnum. Nichols classifies these mosses into 5 ecological groups, beginning with the decidedly aquatic and ending with those of comparatively xerophytic habits. It is upon the growth of the mesophytic and xerophytic sphagnums that the developrtient of the dry raised bog depends. These mosses are cushion- forming in habit, and their successive development elevates the central portions of the bog many feet above its rim. Such a raised bog presents a hmnmocky surface, and except in wet weather a rather firm springy substratum quite dry underfoot. Upon the surface in addition to the xerophytic Sphagnum are other mosses such as Racomitriiim and Polytrichum, some fruticose lichens, and several ericaceous shrubs for the most part less than a foot high. Scattered and dwarfed specimens of Larix and Picea mariana also occur. A typical specimen of the former, scarcely a foot high, possessed a trunk i inch in diameter showing more than 50 annual rings. Another striking feature of the region seems to be the development of sub- sequent ponds within the bog area. These differ decidedly from the marginal trenches described by Stallard,s which are due to fire consuming the peat in the shallow marginal portions of the bog during periods of unusual drought. The ponds in the Cape Breton bogs are due to the impervious nature of the peat from some of the sphagnums forming barriers and dams which obstruct the drainage on gentle slopes. Such ponds function as storage reservoirs, retaining much of the water which accumulates in them during wet periods and thus insuring to adjacent areas a constant supply throughout the season. The development of the raised bogs, the subsequent bog ponds, and other features of the vegetation are illustrated by diagrams and photographs. The various successions are carefully traced and clearly described, the various communities being classified according to the system already noted.^ In its comprehensive character, its abundance of detail, and its notable contributions to various phases of ecology, including the relationships between deciduous and coniferous forests, the ecology of the sphagnums and of the development of 5 Stallard, Harvy, The origin of Sphagnum atolls. New Phytol. 15:250-256. 1918. 6 BoT. Gaz. 66:385-388. 1918. iqiq] current literature 373 raised bogs, this report stands as one of the most notable of recent years. — Geo. D. Fuller. A new fixative for paraffin sections. — Dr. Koloman Szombathy^ describes a new method of fixing paraftm sections to the sHde. The fixative is claimed to have the advantage of not being dissolved by alkaline stains, and furthermore in not being stained by hematins and aniline stains such as eosin fuchsin, orange G., etc. The formula given by him is as follows: gelatin i gm., distilled water loo cc, salicylate of soda (a 2 per cent solution) i cc, pure glycerine 15 cc. Dissolve the gelatin in water at 30°, add the salicylate of soda, shake well, cool, and filter. To this add 15 gm. of pure glycerine. The solution obtained should be perfectly clear. A small amount of the fixative together with a drop or two of a 2 per cent formalin solution is placed on the slide, smeared evenly over the surface, and rubbed in well. Care should be observed that the formahn is mixed with the fixative. The sections or parafiin ribbons are then placed on the fixative and permitted to dry in the thermostat or any other warm place which is protected from dust. The formalin "tans" the gelatin and makes it insoluble. A modification of the method consists in exposing the slides, which have been mounted without the use of formalin solution, to vapors of con- centrated formahn in a thermostat. The effect of the formahn is identical. A third method consists in preparing a solution of equal parts of i per cent gelatin in water and 2 per cent formalin. The fixative is then used as recom- mended for albumen fixative. The writer has tested the fixative recommended by Szombathy and finds it to be an excellent one. Material known to be difficult to retain on the shde was tried out. Sections of grass leaves and moss archegonia adhered to the slide even when the latter were left in running water for several days or exposed to a strong solution of hydrogen peroxide. Alkahne stains do not dissolve the gelatin nor do the stains tested stain the background to an appreciable extent. Of the 3 methods originally recommended, the following modification gives the most satisfactory results. Make up the fixative according to the first formula, put a drop on the slide, and smear it evenly over the surface. Float the paraffin ribbon on the shde on a 2 per cent formahn solution. Warm the sUde gently on the usual copper plate and, after the ribbon has straightened and become smooth, drain off the surplus water and let the preparation dry. When one is deahng with material which does not stick to the shde easily, it will be found of advantage to put a small dish of formahn in the thermostat where the preparations are drying, since the formalin vapors help in rendering the gelatin insoluble. This new fixative is very easily prepared, keeps well, and does hold the sections to the slide. It should come into general use especially for material which does not adhere to the slide under ordinary conditions and when stains 7 Szombathy, Koloman, Neue Methode zum Aufkleben von Parafi&nschmitten . Zeitschr. Wiss. Mikr. 34:334-336. 1918. 374 BOTANICAL GAZETTE [april are employed which are alkaline in nature and have the objection of staining the background. — Ernest F. Artschwager. Size variation in secondary xylem. — Bailey and Tupper^ have applied the comparative method in thoroughgoing fashion to an attack on the problem of cell size. Confining themselves to a study of the length of the tracheary elements in the secondary xylem of trees and shrubs among vascular cryp- togams, gymnosperms, and angiosperms, they present data derived from thousands of measurements on some 440 species belonging to 124 famihes. The most conspicuous fact brought out by this reconnaissance survey is that the length of these elements is roughly correlated with phylogenetic position, being greatest in vascular cryptogams, somewhat less in gymnosperms, and least in angiosperms. This progressive reduction in the length of the wood cells has been associated with the reduction in amount of the primary xylem in the passage from lower to higher forms, but is probably due in greatest measure to the evolution and differentiation of vessels. These elements have become progressively shorter and broader, thus losing their resemblance to the primitive tracheid; and the fibers and tracheids associated with them have also grown shorter, although naturally to a less extent. Notable exceptions to the general rule are the vessel-less Magnoliaceae and Trochodendraceae, represented by Drimys and Trochodendron, which possess tracheary elements far longer than other angiosperms, and thus resemble the gymnosperms. Evidence from this source obviously supports the view that these genera are primitive rather than reduced types. The authors have also made a preliminary study of the relations between the length of the tracheary elements and the age of the plant, its growth habit, and the environment under which it lives. So far as the cells studied are concerned, there is no definite correlation between body size and cell size. The tracheary elements may increase in length for a few years as the plant grows larger, but they soon reach a constant size. Dwarfed and depauperate plants tend to have somewhat smaller elements than normal individuals. The authors point to the need of more intensive investigations in this hitherto almost unexplored field; and in particular call for a careful study of the activities of the cambium and the factors which direct these activities. Indeed the growing point of plants, once so enthusiastically studied as the key to histology and then for so long neglected, bids fair to be once more a center of interest as one of the keys to a knowledge of morphogenesis. — E. W. Sinnott. Sap concentration in epiphytes. — Continuing the studies already noted' upon the concentration of tissue fluids, Harris^" has found in several species of * Bailey, I. W., and Tupper, W. W., Size variation in tracheary cells. I. A comparison between the secondary xylems of vascular cryptogams, gymnosperms, and angiosperms. Proc. Amer. Acad. 54:149-204. ^g^. 6. 1918. 9 BoT. Gaz. 65:285-286. 1918. '» Harris, J. Arthur, On the osmotic concentration of the tissue fluids of desert Loranthaceae. Mem. Torr. Bot. Club 17:307-315. 1918. iqiq] , CURRENT LITERATURE 375 Phoradcndron growing in the Arizona deserts upon various hosts, such as species of Acacia, Querciis, Fraxinus, and Popiilus, that the osmotic concen- tration of the tissue flmds of the parasite is generally greater than that of the host. The concentration of the fluids of such parasites in this semidesert region is also greater than and usually about twice as great as that of similar plants found in the mountain rain-forests of Jamaica. These results quite agree with our expectations, but in a further paper the same investigator" clearly demonstrates the errors that would be involved in generalizing broadly on insufficient data. The later investigations have to do with the tissue fluids of epiphytic Bromeliaceae, Orchidaceae, Piperaceae, and Gesneraceae, and these are shown to possess a decidedly lower concentration than those from terrestrial vegeta- tion. In the mountain rain-forests of Jamaica the epiphytes show 37-60 per cent of the concentration commonly found in herbaceous terrestrial vegetation and 28-45 P^r cent of the concentration characteristic of ligneous soil plants. The epiphytes of the Jamaican rain-forests show lower concentrations than related plants of the same habit growing in the subtropical forests of Florida. The exactness of the data and quantitative character of the comparisons make these investigations important, and lead us to look forward for the further results promised in the study of parasitism by quantitative methods. — Geo. D. Fuller. Bennettitales. — Two cones of the Bennettitales from the British Cre- taceous, one of them a new species, have just been described by Stopes." The first and most important is the one upon which she has founded the new species B. albianus, the specific name referring to the strata in which the specimen was found. Only a small piece of a single cone was found, but it was very well preserved. After a study of the topography, the entire fragment was cut, yielding 2 longitudinal and 5 transverse sections, the latter passing through the seeds and the former through their stalks. The most striking feature of the cone is its large size, not less than 70 mm. in diameter and probably more. The seeds are innumerable, as many as 600 showing in a single transverse section of the fragment. The seeds are 5-6 mm. long and i . 2 mm. in diameter, thus contrasting with the more or less ovoid seeds already described. The interseminal scales are fused around the apex of the seed. The embryo has 2 cotyledons and a rather massive hypocotyl and radicle. The other specimen, B. maxinms, was described from superficial characters by Caiiruthers in 1870. The present study shows that the vascular axis is very small for such a large plant and the cones are bisporangiate, the first petrified bisporangiate cones which have been found in England. The cones " Harris J. Arthur, On the osmotic concentration of the tissue fluids of phanerogamic epiphytes. Amer. Jour. Bot. 5:490-506. 1918. " Stores, Marie C, New Bennettitean cones from the British Cretaceous. Phil. Trans. Roy. Soc. London 208:389-440. pis. ig-24. 1918. 376 BOTANICAL GAZETTE ^ [april are very young but do not seem to have been well preserved. If material in this stage and somewhat older stages could be secured, it would help immensely in comparing the Bennettitales and the Cycadales. — C. J. Chamberlain. Cytology of the basidium. — A cytological investigation of the basidium of Eocronartium muscicola, one of the Auriculariales parasitic upon mosses, was undertaken by Fitzpatrick'^ because he had noticed that the nuclei are of unusual size, and because very little cytological work has been done in this order. The mycelium, which is intracellular and extends throughout the host, is composed of binucleate cells. The cells of the sporophore are also binucleate, and, during division, it is seen that the number of chromosomes in each of the 2 nuclei is 4. During the development of the basidium, the 2 nuclei fuse, the resulting nucleus passes into synapsis, and in later stages of division shows 4 chromosomes, which is also the number at the second division, so that the total number of chromosomes in the cell is reduced. Toward the close of the second division a transverse wall appears in the middle of the basidium and is soon followed by two more walls, so that the basidium consists of a filament of 4 cells. The sterigmata, which are large in proportion to the cells from which they arise, are not quite simultaneous in their appearance. The chromatin becomes drawn out into a slender thread as the nuclei pass into the young spores, and there is no connection with the centrosomes, as has been reported for some basidia. How the binucleate mycehum arises from the uninucleate spore has not yet been determined. — ^C. J. Chamberlain. Orientation of roots. — Holman"! has investigated the influence of the medium upon the orientation of primary terrestrial roots. He shows that the failure of roots grown in air to reach a vertical position is due to lack of mechan- ical resistance to the advance of the root tip after the flattening of the primary geotropic curvature, rather than to differences in water content in the medium, or changes in geotropic sensitiveness, or to thigmotropism. His observations have been extended to secondary roots,'s and here also he finds that when they have been displaced from normal position with respect to gravity, and the first curvature of response has been flattened, mechanical resistance is necessary to a complete reaction to normal position. The mechanical resistance hinders flattening of the primary curvature of the root tip, and passively depresses the tip as it moves forward, thus reinforcing and completing the geotropic response. — C. A. Shull. '3 FiTZPATRiCK, H. M., The cytology of Eocronartium muscicola. Amer. Jour. Bot. 5:397-419. ph. 30-32. 1918. "• HOLMAN, Richard M., The orientation of primary terrestrial roots with particu- lar reference to the medium in which they are grown. Amer. Jour. Bot. 3:274-318. 1916. 's , Influence of the medium upon the orientation of secondary terrestrial roots. Amer. Jour. Bot. 3:407-414. 1916. T AST summer, America expected a long -*— ^ war. She prepared for it, and pre- pared so tremendously that the knowledge of what America was doing broke the spirit of the German resistance. • The war collapsed. But those preparations must be paid for. Such a sudden victory is expensive in money, hut a slower victory would have cost the lives of thousands more ol Americans boys, and even more treasure. Americans will show their gratitude by helping to pay for those preparations. / The money for the Victory Liberty Loan will pay the bills. How much will you subscribe ? Subscribe to the Victory Liberty Loan Space contributed by THE UNIVERSITY OF CHICAGO Prepared by American Ai^sociatioii of Advertising Ageni;ies Starved Rock State Park and Its Environs By CARL 0. SAUER Department of Geology, University of Michigan GILBERT H. CADY Illinois State Geological Survey HENRY C. COWLES Professor of Ecology, University of Chicago This thorough and interesting geographical study of the best-known feature of the Illi- nois Valley has been prepared for the benefit of those who are interested in knowing more about the region than can be gathered by personal observation in a few hours or a few days spent in the park. The physical geog- raphy of the park and its surroundings has a number of features, such as the beautiful little canyons, which are unusual in this part of the country. Part I is given up to the geography of the park, its surface features and their origin, and the exploration, settle- ment, and development of the region; Part II, to the geology of the park; and Part III, to its botany. 130 pages, cloth, $2.00, postage extra {weight 1 lb. 8 oz.) THE UNIVERSITY OF CHICAGO PRESS CHICAGO .... ILLINOIS MORPHOLOGY of GYMNOSPERMS By John M. Coulter and Charles J. Chamberlain A revised edition of Coulter and Chamberlain's well-known work on Morphology of Gymnosperms is now offered, not only for the use of investiga- tors, but for advanced students in this field. Each of the seven great groups is presented in detail, and the final chapter discusses the problem of phylogeny and points out the evolutionary tend- ency. _ Although some corrections and additions occur in the presentation of all the groups, the most numerous changes will be found in the chapter on Cycadales, the group which has received the most attention since the publication of the former edition. The most important addition is to the Bibliography, 150 titles being added to the 484 titles of the former book. This list will put the student in touch with the work of the last seven years. The book is equipped with an index and appendixes. 'The most important general work on the gymnosperms which has ever appeared." — E. C. Jeffrey in Science. 478 pages, 462 illustrations, 8vo, cloth; $5.00, postage extra {weight 3 lbs.) The University of Chicago Press Chicago .... Illinois PLANT GENETICS By JOHN M. COULTER Head of the Department of Botany in the University of Chicago and MERLE M. COULTER Instructor in Plant Genetics in the University of Chicago fl This book has been written to meet an increasing need among botanical students. Such students in these days, in whatever phase of botany they may be specializing, find it neces- sary to read with understanding much of the literature of plant genetics, because it is becoming increasingly significant in all botanical problems. This means that teachers and investigators must be able to command the literature of plant genetics, much of which has been so complex as to be a closed book for the uninitiated. Plant Genetics is an attempt to open this subject to botanical students. fl The book is not intended to be a thorough, authoritative text, but a relatively simple presentation of the more significant investigations on plant genetics which will initiate the student into the subject. Material dealing with some highly specialized phases of genetics and material that is very complex has been purposely omitted for pedagogical reasons. In short, the book is an easy introduction to plant genetics. x-\-2i4 pages, i2mo, cloth; $1 .50, postage extra {weight i lb. 2 oz.) THE UNIVERSITY OF CHICAGO PRESS CHICAGO ILLINOIS A NEW VOLUME OF THE UNIVERSITY OF CHICAGO SCIENCE SERIES The Living Cycads By Charles Joseph Chamberlain Professor of Botany in the University of Chicago This study includes investigations extending over fifteen years, that necessitated trips to Cuba, New Mexico, Australia, and Africa. The author studied all the genera and many of the species in the field, and preserved much material for later study in the laboratory. In the first part of the book is given an account of the distribution, general appearance, and field conditions of the cycads, together with some of the experiences of an investigation involving much travel in distant and varied countries. In the second part is presented the life-history of the group based largely on the author's own observations in the field and laboratory. Part III is devoted to the evolution and phylogeny of the cycads, the opportunity -for such a study being exceptionally favor- able because the ancestry can be traced back through geological periods, and because the extinct predecessors of the cycads are the best known of the fossil plants. xiv-\-i72 pages, cloth; $1.50, postage extra TO BE PUBLISHED APRIL 15 Problems of Fertilization By Frank R. Lillie Professor of Embryology and Chairman of the Department of Zoology in the University of Chicago As the problem of fertilization deals with the union of the male and the female sex elements, which forms the beginning of each individual's life-history, it has important bearings upon the problems of heredity. It also has close relations to several other fundamental biological problems. Cloth; $!.'/§, postage extra OTHER PUBLISHED VOLUMES IN THE SERIES The Evolution of Sex in Plants. By John Merle A Chemical Sign of Life. By Shiro Tashiro. Coulter, x-l-140 pages, i2mo, doth; $1.00, post- x-l-142 pages, small i2mo, cloth; $1.00, postage age extra (weight 14 oz.). extra (weight 14 oz.). kidioiduality in Organisms. By Charles Manning ^^^ Biology of Twins. By Horatio Hackett CmLD. X-I-2I2 pages, small i2mo, cloth; $1.25, Newman. x+i86 pages, small i2mo, cloth; postage extra (weight i lb. 6 oz.). *^-^5, postage extra (weight i lb.). TL n • • r iL c- .L T,^ ^ '^°**'' Poisoning. By Edwin Oakes Jordan. I he Ungtn of the Earth. By Thomas C. viii-|-u6 pages, small i2mo, cloth; $1.00, postage Lhaiiberlin. xii -H 27 2 pages, small 1 2mo, cloth; extra (weight 11 oz ) S1.50, postage extra (weight i lb. 6 oz.). ff,^ Electron: Its Isolation and Measurement, and Finite Collineation Groups. ByHANS F.Blichfeldt. the Determination of Some of Its Properties. xii-Hi94 pages, small i2mo, cloth; $1.50, postage By Robert Andrews Millikan. xii 4- 268 pages, extra (weight i lb.). small i2mo, cloth; $1.50, postage extra (weight I lb. 3 oz.). VOLUMES PLANNED OR IN PREPARATION Black Body Radiation. By Charles E. Mendenhall. Mechanics of Delayed Germination in Seeds. By W. Crocker. The Rigidity of the Earth and of Materials. By A. A. Michelson. Linear Integral Equations in General Analysis. By Eliakim Hastings Moore. THE UNIVERSITY OF CHICAGO PRESS CHICAGO ILLINOIS Men of Science The signs and symbols used in your everyday work are provided for in special keyboards prepared for the .Home Typewriter, the Self-Starting Remington Tell us your needs and we will sub- mit a keyboard layout for your work. No obligation, of course, on your part. Remington Typewriter Company, Inc. 374 Broadway, New York City FINE INKS AND ADHESIVES For those who KNOW Higgins' Are the Finest and Best Drawing Ink« eternal Writinc Ink Engrossing: ink Taurine Mucilage Photo Mounter Paste Drawing Board Paste Liquid Paste Office Paste Vegetable Qlue, Etc. Inks and Adhesives Emancipate yourself from the use of corroiive and ill-smelling inks and adhesives and adopt the Higgins Inks and Adhesives. They will bea revelation to you, they are so sweet, clean, well put up, and withal so efficient. At Dealers Qeneraliy CHAS. M. HIGQINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street Brooklyn, N.Y. Combined Balopticon For Lantern Slides and Opaque Objects Price $135 ^ausch^^lomb BALOPTICON THE PERFECT STEREOPTICON for every practical form of optical projection is again in the market, backed by increased facilities w^hich have been occupied in wrar service of arnost significant character. Our release from this service is of course gradual. As the demands made upon us by the government are lessened from month to month, however, we are enabled to in- crease deliveries to our other patrons. Write for Balopticon catalog, with price list of our re- vised line, and inform us regarding your requirements. Bausch ^ Ipmb Optical (5. 654 ST, PAUL STREET, ROCHESTER, N.Y. NEW YORK WASHINGTON CHICAGO SAN FRANCISCO Leading American Makers of Photographic Lenses, Microscopes, Projection Apparatus {Balopticora) , Ophthalmic Lenses and Instruments, Photomicrographic Apparatus, Range Finders and Gun Sights for Armu and Navu, Searchlight Reflectors, Stereo-Prism Binoculars, Magnifiers and Other High-Grade Optical Products. I ^ Volume LXVII Number 5 THE Botanical Gazette Editor: JOHN M. COULTER MAY 1919 Effect of Anesthetics upon Respiration - (With seven figures) A. R. C. Haas 377 Basis of Succulence in Plants D. T. MacDougal, H. M. Richards, and H. A. Spoehr 405 A Coniferous Sand Dune in Cape Breton Island (With eight figures) - LeRoy H. Harvey 417 Embryo Sac and Embryo of Pentstemon secundiflorus. Contributions from the Hull Botanical Laboratory 248 - With Plate Xn Briefer Articles Depressed Segments of Oak Stems . (With four figures) Importance of Epidermal Coverings (With two figures) Current Literature Notes for Students ----=■ Arthur T. Evans 427 I. W. Bailey 438 R. B. Harvey 44 1 - 445 The University of Chicago Press CHICAGO, ILLINOIS, U.S.A. THE CAMBRIDGE UNIVERSITY PRESS, London and Edinburgh THE MARUZEN-KABUSHIKI-KAISHA. Tokyo, Osaka. Kyoto. Fukuoka. Smndai THE MISSION BOOK COMPANY. Shanghai volume LAVU Number 5 THE Botanical Gazette A MONTHLY JOURNAL EMBRACING ALL DEPARTMENTS OF BOTANICAL SCIENCE EDITED BY JOHN M. COULTER With the assistance of other members of the botanical staff of the University of Chicago Issued May 19, 1919 The Botanical Gazette is published monthly by the University of Chicago at the University Press, 5750 Ellis Avenue, Chicago, HI. HThe subscription price is $7.00 per year; the price of single copies is 75 cents. Orders for service of less than a half-year will be charged at the single-copy rate. ^[Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, Republic of Panama, Hawaiian Islands, Phihppine Islands, Guam, Samoan Islands, Shanghai. IfPostage is charged extra as follows: For Canada, 35 cents on annual subscriptions (total $7 -35), on single copies, 3 cents (total 78 cents); for all other countries in the Postal Union, 84 cents on annual subscriptions (total $7.84),*on single copies II cents (total 86 cents). If Patrons are requested to make all remittances payable to The University of Chicago Press in postal or express money orders or bank drafts. The following are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. 4. Yearly subscriptions, including postage, £1 125. 6d. each; single copies, including postage, 35. 6d. each. For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage, Yen 15.75 each; single copies, including postage. Yen i . 75 each. For China: The Mission Book Company, 18 Peking Road, Shanghai. Yearly sub- scriptions, $7.00; single copies, 75 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 84 cents, on single copies II cents. Claims for missing numbers should be made within the month following the regular month of pubUcation. The pubUshers expect to supply missing numbers free only when losses have been sustained in transit and when the reserve stock will permit. DeKveries^ are not guaranteed within the war zones of Europe, nor in other countries where transportation is disturbed by con- ditions incident to war. Business correspondence should be addressed to The University of Chicago Press, Chicago, lU. Communications for the editor and manuscripts should be addressed to the Editor of the Botanical Gazette, The University of Chicago, Chicago, 111. Contributors are requested to write scientific and proper names with particular care, to use the metric system of weights and measures, and in citations to foUow the form shown in the pages of the Botanical Gazette. Papers in excess of thirty-two printed pages are not accepted imless the author is wilL'ng to pay the cost of the additional pages, in which case the number of the pages in the volume is correspondingly increased. Illustrations are furnished without cost to author only when suitable originals are supphed. A copy of the suggestions made in the January number, 1907, will be sent on application. It is advisable to confer with the editors as to illustrations required in any article to be ofiFered. Separates, if desired, must be ordered in advance of pubUcation. A table showing approxi- mate cost of separates is printed on an order blank which accompanies the proof; a copy will be sent on request. Entered as second-class matter August 21, 1896, at the Post-Office at Chicago, Illinois, under the Act of March 3, 1879. Acceptance for mailing at special rate of postage provided for in Section 1103, Act of October 3, 1917, authorized on July IS, 1918. VOLUME LXVII NUMBER 5 THE Botanical Gazette MAY igig effect of anesthetics upon respiration A. R. C. Haas (with seven figures) The special interest which this subject has acquired, as the result of certain modern theories, makes it desirable to give a brief review of some of the more important contributions (5, 10) to our knowledge of it. ' Meyer (19) and Overton (21) independently concluded that the effect of a narcotic increases with its solubility in substances of a lipoid nature. According to them narcosis does not appear until the lipoids of the cells have absorbed the narcotic to a definite molecular concentration (21). The theory has been criti- cized (5, 10) because it fails to explain why narcotics, such as benzamide and monacetin, which at higher temperature are less soluble in fat, have an eft'ect which increases with the temperature. Both Meyer and Overton recognize the fact that often there is no relation between the narcotic power of a substance and its rela- tive solubility in oil; the partition quotient for isobutyl alcohol is about 180 times greater than for ethyl alcohol, but its power to cause narcosis is only about 6 times as great as that of ethyl alcohol. They state that such cases cannot be used as arguments against the lipoid theory in case the narcotic is not chemically indift'erent, but has a special reaction affinity, as is the case, for example, with the basic narcotics. Verworn (29) has advocated the view that narcotics interfere with the oxygen carriers of the cell and render them incapable of 377 378 BOTANICAL GAZETTE [may activating the molecular oxygen. As a result oxidation cannot take place and disintegration occurs, the cells thereby being asphyxiated. In this connection he makes the statement that a more or less com- plete recovery from narcosis, which may occur even in an oxygen- free medium, is at the cost of the oxygen contained within the living substance, which (on account of the suppression of the oxidation processes) could not be consumed. Mansfeld's (i6) view is not essentially different from that of Verworn. He beheves that because the lipoids take up the narcotic, their power to absorb oxygen is decreased. Narcotized cells cannot take up sufficient oxygen for their needs and hence irritability is decreased by lack of oxygen. That narcotics do decrease the ability of olive oil to dissolve oxygen, has been asserted by Hamburger (8) , although objection to his experiments has been made by Winterstein (31). In many cases it is certain that narcosis has nothing to do with absorption of oxygen. Thus Winterstein has observed that on anesthetizing the anaerobic worm Ascaris in absence of oxygen, it comes to rest very quickly under the influence of the anesthetic. Experiments by numerous investigators have shown that nar- cosis and decrease of oxidation are not parallel. Warburg (30) has observed that some narcotics decrease the oxidation of the erythrocytes of geese as much as 30-70 per cent. He has also observed, however, that narcotics do not always decrease the con- sumption of oxygen, for he has found that the segmentation of fertilized sea urchin eggs can be inhibited by phenylurethane with- out a perceptible decrease of the consumption of oxygen. Loeb and Wasteneys (14) have obtained similar results with chloral hydrate, chloroform, and alcohol. Nothmann-Zuckerkandl (20) observed that the protoplasmic streaming of plant cells is quickly brought to a standstill by narcotics. Such cells have been shown to be rather insensitive to lack of oxygen, in that the streaming is stopped in the absence of oxygen only after several weeks. BtJRKER (4) was of the opinion that because of the great solu- bihty of narcotics in the lipoid of the cells, there is a competition between the lipoids and other substances in the protoplasm for the active oxygen, such that the cells are more or less in a state of 1 9 1 9] HAA S~RESPIRA TION 379 asphyxiation. His h^'pothesis maintains that the absorption of oxygen is not reduced during narcosis, but that the oxygen is prevented from going to its usual point of attack. This would necessitate the assumption that oxygen is more readily absorbed by the lipoids than by the other substances of the cell, which is not the case. An objection to all theories which make narcosis con- ditional on lipoid solubility is the fact that magnesium sulphate and carbon dioxide, which are not soluble in lipoid, produce typical anesthesia. HoBER (10) has formulated the h\^othesis that narcosis is due to inhibition of enzymatic processes, brought about by a decrease in dispersion. This hypothesis is favored by the investigations of Batelli and Stern (2), who found that proteins, such as the nucleo-proteins, are influenced by narcotics in approximately the same relative and absolute concentrations as those at which the enzymes are affected. In this connection Vernon (28) found that most narcotics are harmless to oxidases up to a definite limiting concentration, beyond which injury occurs. Hober explains the retardation or inhibition of enzyme action as due to the fact that narcotics go into the surface between the enzymes and the medium in which they are dispersed, thereby displacing the substratum on which the enzymes act. The investigations hitherto discussed are largely based on measurements of the absorption of oxygen. In many cases such measurements could be made more accurately than determinations of the amount of carbon dioxide produced. Since the writer has recently been able to develop a method for the measurement of minute amounts of carbon dioxide in solution, it seemed to him that a fresh investigation of the subject by means of this method was desirable. Previous investigations on the effect of narcotics on the produc- tion of carbon dioxide have yielded somewhat contradictory results. Appleman (i) found that vapor of ethyl bromide approxi- mately doubles the respiration of potatoes. Mayer (18) found that 0.25 per cent prussic acid stops the respiration of higher plants entirely. Schroeder (24) observed a decrease in respiration when Aspergillus was treated with ether. He found that prussic acid 380 BOTANICAL GAZETTE [may inhibited the production of CO2, while the absorption of oxygen continued. Kosinski ■ (12) found that low concentrations of ether increased the respiration of Aspergillus, while higher concen- trations decreased it. Lauren (13) and also Irving (ii) have noted that respiration increases during anesthesia produced by ether and chloroform. Tashiro (26) found that anesthetics greatly reduce the output of CO2 by dry seeds. Bonnier and Mangin (3), as the result of experimental work upon the influence of measured quantities of ether upon flowering plants, concluded that the respiratory activity is unaffected by anesthetics. It has since been shown by Ewaet (6) that chloroform increases the respir- atory activity in Elodea. The effect of anesthetics upon the respiration of marine plants has received very httle attention. Harder (9) has made determi- nations of the respiration of marine algae, but in no case has he studied the effect of anesthetics. Pantanelli (22) has observed that sea water, when half saturated with chloroform, reduces the excretion of- CO2 to about one-half of the normal. His experiments were very few, and no duplicate or control experiments were made. The methods employed by Pantanelli and Harder often required that plants be shut up air-tight, in flasks completely filled with sea water, and left in this condition for several hours, analyses being made at the beginning and end of these long periods. In the experiments of the writer on the effect of anesthetics upon the production of CO2, the marine alga Laminaria was found to be well suited to the purpose. Fronds were cut up into pieces about 2 inches long. Each piece was rolled up loosely and inserted into a piece of Pyrex glass tubing. This was closed at one end, while a piece of paraffined rubber tubing was attached to the open end. Sea water, of approximately the same temperature as the material, was then added to the tubes and the rubber tube closed by a spring clamp. The tubes were then brought very gradually to the temperature (16° C.) of the constant temperature bath. The tubes were kept dark by inserting each tube in a black- enameled, collapsible tin tube submerged in the bath. Several tubes were used as controls in each experiment. These contained material in sea water without aoiy addition. The reagents were 1 9 1 9] HA A S—RESPIRA TION 3 8 1 always used at the same temperature as that of the water bath. After the tubes containing the Laminaria in sea water had been at 16° C. from half an hour to an hour, the solution was poured out of each tube and replaced by fresh sea water at 16° C. This was repeated several times before beginning an experiment. After a tube containing Laminaria in sea water had been clamped off so as to include a small bubble of air which could be used as a stirrer, and had been exposed to 16° C. for a definite period, it was removed from the bath. The solution was slightly stirred by inverting the tube a few times; the clamp was then opened and the solution rapidly poured into another empty tube to which the same number of drops of indicator (phenolsulphone- phthalein) had been added as was added to the buffer solutions (7). In order to mix the solution with the indicator it was stirred as described, and was then compared with buffer solutions of a known PH value which contained the same amount of indicator. The use of a small bubble of air as a stirring agent was found to be very convenient, and when compared with the use of paraffined glass globules as stirrers was found to introduce no error of any importance. The buffer solutions (27) were made up by mixing M/15 Na2HP04 and M/15 K.H2PO4 in various proportions. They were of the same diameter as those containing the alga. A o.oi per cent aqueous solution of phenolsulphonephthalein served as the indicator of the PH values. The indicator solution was used at the rate of 5 drops in 10 cc. of solution. A correction (30) of o . 30 for the salt error of the indicator in sea water was subtracted from the observed readings of the PH value of sea water. The use of a constant source of light (the "Daylight" lamp) permitted observations to be made in a uniform manner. The decrease in PH, which results from the production of CO2, served as a criterion of the amount of respiration. In all the experiments, each tube was observed for a number of periods (always of the same length) for each piece of material, until it was evident that the rate of respiration had become prac- tically constant. Several of the tubes were then used as con- trols, while to the others was added the sea water containing the 382 BOTANICAL GAZETTE [may anesthetic. In every case tubes which contained only sea water showed that the apparatus was not responsible for any of the changes observed in the experiments in which tissue was used. The solutions used in any one experiment were always of the same PH value, so that the results were comparable. Many of the anesthetics, especially when in small concentrations, do not appreciably affect the PH value of the sea water. A large number of carboys of sea water, obtained from Woods Hole at the same time, were filtered to remove organic matter in varying degrees, the unfiltered material settUng to the bottom. When an anes- thetic decreased the PH value only slightly, it was possible by a suitable selection of the sea water from the carboys to obtain a bottle of sea water of a PH value equal to that of the sea water containing the anesthetic. Whenever large amounts of an anesthetic were used in the sea water, it was considered advisable to add concentrated sea water to bring the electrical conductivity of the solution up to that of sea water. This was done with each of the following solutions: sea water containing 16. i and 24.2 per cent alcohol respectively; sea water containing 17.4 per cent acetone; and sea water con- taining 3 . 2 per cent formaldehyde. Many of the experiments were repeated at Woods Hole in the summer of 191 7. In these experiments the solutions containing alcohol (all above i per cent) were made of the same electrical conductivity as sea water. The results were practically the same as when no concentrated sea water was added. In the case of the higher concentration of formaldehyde, the free acid was first neutralized with sodium carbonate. This is allowable for the purposes of the present investigation, as its only effect would be to make the amount of CO2 produced appear some- what less than was actually the case. It should be understood that the percentages of liquid anesthetics in the following descriptions are percentages by volume. In an experiment in which there is considerable dilution, as when absolute alcohol is added to sea water to make a 10 per cent alcoholic solution, it might be con- ceivable that the dilution might be the cause of the increased rate of respiration. For that reason experiments were made with 1919] HAAS— RESPIRATION 383 pieces of material in normal sea water until constant values for the respiration were obtained; then sea water diluted with 10 per cent of tap water was used, and it was found that the dilution produced no appreciable effect. In each experiment 6 cc. of solution was used. At the end of each period, after noting the new PH value, 6 cc. of fresh solution was poured into the tube, which was then treated as before. It was at first thought that in pouring out the solution from the tube containing the tissue sufficient might be held back by the tissue to affect the observed PH value. Actual determinations proved that this was not the case. In view of the fact that there may be an increase of respiration as the result of injury (23), preliminary experiments were made to ascertain whether the cutting of the Laminaria had any appre- ciable effect upon the respiration. It was found that the change in the respiration due to cutting was negligible (the cutting was always reduced to a minimum). In some of the experiments there was noted a very slight decrease in the production of CO2 as time went on, although not enough to be of significance for the present investigation. This phenomenon has been observed by Miss Matth.a.ei (17) in determinations of CO2 in connection with experiments on photosynthesis. Such a decrease in respiration has been attributed by her to the gradual decrease of substances available for oxidation. The results of the experiments are given both in tables and figures. In each experiment 6-12 (or more) closely agreeing deter- minations were obtained and the results of a typical case were taken. The plant was always placed for a definite time in sea water or sea water containing the reagent, and at the end of this time the PH value was determined. This interval is in every case shown in the tables as well as by the points on the curves. At the end of each interval the solution was renewed. The alteration of the PH value is an index of the amount of CO, produced, since the greater the amount of CO^ the greater the decrease in PH. Since this relation is approximately linear in this range of PH values, it is not necessary to translate the decrease in PH value into cc. of CO2 produced, as the form of the respiration 384 BOTANICAL GAZETTE MAY curve would be practically the same whether we use as ordi- nates PH values or cc. of CO2 produced. The curves are plotted in such a manner that the abscissae represent time (in minutes) and the ordinates represent the change in PH value corresponding to either the relative amount of respiration, the relative rate of respira- tion, or to both (as designated in the figures). The relative amount of respiration is obtained by dividing the total change in PH during exposure to the reagent by the total change in PH caused by the REL.RATEOFRESP REL. AMT.OF RESP. 2 AB DD '^c OA a-e^-K--ar D' E' A' C 60 MINUTfeS Fig. I a. — Points showing relative rate and relative amount (identical in this case) of respiration of Laminar ia produced by sea water containing A, 0.1 per cent chloral hydrate; B, 0.1 per cent novocain; C, i per cent ether; D, o.i per cent cafifeine; E, ethyl-bromide; controls in sea water (broken lines); see tables I A to E; each con- trol bears the same letter (with a prime) as the experimental curve. same material under normal conditions during the same length of time. The relative rate is obtained by dividing the change in PH during one period by the change produced during a similar period by the same material under normal conditions. The broken lines in each case represent controls in sea water. The curve of each control bears the same symbol and letter as the experimental curve for sea water plus anesthetic, except that in the control the letters are primed. Inspection of the results for o . i per cent chloral hydrate (fig. la [A]; table I A), o.i per cent novocain (fig. la [B]; table I Bj, iqiq] HA AS—RESPIRA TION 385 TABLE I A* Period of 47 . 75 min. ix sea water and equal period ix sea water con- taining O . I PER cent chloral HYDRATE Solution Change in PH Relative rate of respiration Sea water 7-90-7-55 = o-35 7.90-7-35=0.55 Sea water containing 0 . i per cent chloral hydrate 0.55-^0.35 = 1.57 * In the control, with 3 periods (38.75 min. each) in sea water, the change in PH in each period was 7.90-7.55=0.35. TABLE I B* Period of 33.25 mix. in sea water followed by equal period in sea WATER containing O . I PER CENT NOVOCAIN Solution Change in PH Relative rate of respiration Sea water 7.65-7.40 = 0.25 7.65-6.90 = 0.75 Sea water containing 0 . i per cent novocain 0.75-^0.25 = 3 * In the control, with 2 periods (30.75 min. each) in sea water, the change in PH in each period was 7.65-7.40=0.25. TABLE I C* Period of 45.25 mix. ix sea water and equal period ix sea water CONTAINING I PER CENT (bY VOLUME) OF ETHER Solution Change in PH Relative rate of respiration Sea water 7.90-7.57 = 0.33 7.90-7.25=0.65 Sea water containing i per cent ether 0.65^0.33 = 1.97 * In the control, with 3 periods (42 min. each) in sea water, the change in PH in each period was 7.90 — 7.60 = 0.30. TABLE ID* Two PERIODS (32.5 MIN. EACH) IN SEA WATER AND SAME LENGTH OF TIME IX SEA WATER CONTAINING O . I PER CENT CAFFEINE Solution Change in PH Relative rate of respiration Sea water 7.65-7-43 = 0.22 7.65-7.43 = 0.22 7.65-7.25=0.40 1 <( u i Sea water containing 0 . i per cent caffeine 0.40-^0.22 = 1 .8 * In the control, with 3 periods (29.5 min. each) in sea water, the change in PH in each period was 7.65*-7.5o=o.i5. 386 BOTANICAL GAZETTE [may TABLE I E* Period (43 . 25 min.) in sea w.ater followed by equal period in sea water .approximately saturated with ethyl bromide Solution Change in PH Relative rate of respiration Spa water 7-90-7-35 = o.55 7.90-6.75 = 1.15 Sea water saturated with ethyl bromide 1.15-7-0.55 = 2.1 * In the control, with 2 periods (49. 25 min. each) in sea water, the change in PH in each period was 7.90 — 7.30 = 0.60. TABLE I F Three periods (22 min. each) in sea water and 7 equal periods in sea water approximately saturated with ethyl bromide Solution Change in PH Relative rate of respiration Sea water 8 I —7.8 =o.?o u u 8 7 7 7 7 7 7 I -7.8 =0.30 32 — 6. i6*= 1 . 16+ 32 — 6.i6*=i . 16+ 32 — 6.69 =0.63 32-7.07 =0.25 32-7.21 =0.11 32 — 7.32 =0.0 Sea water containing ethyl bromide u u u '< « U U U K U u u u u u U U U U tl u u u u u I.I6-^o.30 = 3.8 1.16-7-0.30 = 3.8 0.63^0.30=2.1 0.25-4-0.30 = 0.8 0. 11-^0.30 = 0.4 0.0 ^0.30 = 0.0 ■ Approximate (at this point the indicator is not very sensitive to slight changes in acidity). TABLE I F CONTROL Seven periods (27.5 min.) in sea water Period I 2 3 4 5 6 7 Change in PH Relative rate of respiration 8 8 8 8 8 8 8 1-7.6 =0.5 1-7.6 =0.5 1-7.6 =0.5 1-7.6 =0.5 1-765 = 0. 45 1-7-65 = 0.45 1-7-65=0-45 0.5 ^0.5 = 1.00 0.5 -^O. 5 = 1. 00 0.5 -5-0.5 = 1. 00 0.5 -j-o.5 = i.oo 0.45-7-0.5 = 0.90 o.45-^o. 5 = 0.90 0.45-7-0.5 = 0.90 I per cent ether (fig. la [C]; table I C), o. i per cent caffeine (fig. la [D]; table I D), and for sea water approximately saturated with ethyl bromide (fig. la [E]; table I E; fig. ib [A and B]; table I F) shows that these anesthetics increase both the relative amount and relative rate of respiration. I9I9] HAAS— RES P IRA TIOX 387 In fig. 16 (A) it will be seen that in sea water containing ethyl bromide the relative rate of respiration is greatly increased during the first 2 periods and then drops to the normal rate in about 90 min. After 132 min. no excretion of CO^ could be detected. When we plot the curve B (fig. ib), in which the ordinates represent the relative amount of respiration, we find that the curve is far above the base line even when the respiration cannot be detected. REL. RATEOFRESP. REL. AMT.OFRESP MINUTES Fig. 16. — Cunes showing effect of sea water approximately saturated with ethyl bromide upon A, relative rate of respiration; B, relative amount of respiration of Laminaria (unbroken lines j; controls in sea water (broken linesj; see table I F; each control bears the same letter (with a prime; as the experimental curve. The fact that ethyl bromide appears to have a marked accelerat- ing effect upon the respiration confirms the results of Applemax d), who obsers'ed that when potato tubers are exposed to ethyl bromide vapor the respiration is greatly increased. Fig. ib makes it evident that there is no initial decrease in the respiration. The ethyl bromide referred to in table I F was acid, and since it was not neutralized by adding sodium carbonate it caused the PH of the sea water plus anesthetic to be lower than that of the sea water alone. In experiments with all other substances both sea water and sea water plus anesthetic were of the same PH value at the beginning of the experiment. 388 BOTANICAL GAZETTE [may The respiration of Laminaria was followed for several hours, when the tissue was placed in solutions of sea water containing 3.2 per cent (fig. 2a [A], fig. ih [A]; table II A) and 0.8 per cent REL.AMT.OFKESP. 2 B~~r~^~~~~-A- A £, 60 120 MINUTES Fig. 2a. — Curves (unbroken lines) showing the effect upon the relative amount of respiration of Laminaria of sea water containing A, 3.2 per cent formaldehyde; £,0.8 per cent formaldehyde; C, 0.3 per cent chloroform; D, 0.05 per cent chloro- form; controls in sea water (broken lines); see tables II A to II D; each control bears the same letter (with a prime) as the experimental curve. REL.RATEOFRESP. ^^^s^^:^.^^ 60 120 MINUTES Fig. 26. — Curves (unbroken lines) showing effect upon relative rate of respiration of Laminaria of sea water containing A, ^.2 per cent formaldehyde; B, 0.8 per cent formaldehyde; C, 0.3 per cent chloroform; Z), 0.05 per cent chloroform; controls in sea water (broken lines); each control bears the same letter (with a prime) as the e.xperimental curve. iqiq] HAAS—RESPIRA TION 389 TABLE II A Two PERIODS (23.5 MIN. EACh) IN SEA WATER AXD 13 EQUAL PERIODS IX SEA WATER CONTAINING 3 . 2 PER CENT FORMALDEHYDE (8 PER CENT BY VOLUME OF 40 PER CENT FORMALDEHYDE) Solution Sea water. Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water ^containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3.2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde ' Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Sea water containing 3 . 2 per cent formaldehyde Change in PH 37-7 37-7 37-6 37-6 37-7 37-7 37-7 37-7 37-7 37-7 37-7 37-7 37-7 37-7 02 = 0.55 82 = 0.55 90=1.47 90=1.47 05 = 1.32 15 = 1.22 33 = 1 04 50 = 0.87 55 = 0.82 58 = 0.79 60 = 0.77 65=0.72 70 = 0.67 80 = 0.57 Relative rate of respiration 37-7-85 = 0.52 1-47-^0.55 = 2 7 1.47-^0.55 = 2 7 1.32^0.55 = 2 4 1.22^0.55 = 2 2 1.04-^0.55 = 1 9 0.87^-0.55 = 1 6 0.82^0.55 = 1 5 0.79-7-0.55 = 1 4 0-77-^0.55 = 1 4 0.72-^0.55 = 1 3 0.67-7-0.55 = 1 2 0.57^0.55 = 1 0 0.52^0.55=0 95 TABLE II A CONTROL Fifteen periods (21 min.) in sea water I . 2. 3- 4- 5- 6. 7- 8. 9- 10. II. 12. 13- 14- 15- Period Change inPH Relative rate of respiration 8.37-7.80 = 0.57 0.57- -0.52 = I . 10 8.37-7 90 = 0.47 0.47- -0.52=0.90 8.37-7 90 = 0.47 0.47- -0.52=0.90 8.37-7 90 = 0.47 0.47- -0.52=0.90 8.37-7 92=0.45 0.45- -0.52=0.87 8.37-7 92=0.45 0.45- -0.52=0.87 8.37-7 92=0.45 0.45- -0.52=0.87 8.37-7 92 = 0.45 0.45- -0.52 = 0.87 8.37-7 90 = 0.47 0.47- -0.52 = 0.90 8.37-7 90 = 0.47 0.47- -0.52 = 0.90 8.37-7 90 = 0.47 0.47- -0.52 = 0.90 8^37-7 90 = 0.47 0.47- -0.52=0.90 8.37-7 90 = 0.47 0.47- -0.52 = 0.90 8.37-7 90 = 0.47 0.47- -0.52 = 0.90 8.37-7 90 = 0.47 0.47- -0.52 = 0.90 39° BOTANICAL GAZETTE [may TABLE II B Two PERIODS (22.25 MIN. EACH) IN SEA WATER FOLLOWED BY 5 EQUAL PERIODS IN SEA WATER CONTAINING 0.8 PER CENT FORMALDEHYDE (2 PERCENT BY VOLUME OF 40 PER CENT FORMALDEHYDE) Solution Change in PH Relative rate of respiration Sea water . u a Sea water containing 0.8 per cent formaldehyde Sea water containing 0.8 per cent formaldehyde Sea water containing 0.8 per cent formaldehyde Sea water containing 0.8 per cent formaldehyde Sea water containing 0.8 per cent formaldehyde 90-7 90-7 90—6 90—6 90-7 90 — 6 90—6 55 = 0.35 55 = 0.35 8 I = I . 09 96 = 0.94 03 = 0.87 95 = 0.95 95 = 0.95 1.094-0.35 = 3.1 0.94-^0.35 = 2.7 o.87-f-o.35 = 2.5 0.95^0.35 = 2.7 0.95-7-0.35 = 2.7 TABLE II B CONTROL Five periods (20 min.) in sea water Period Change in PH Relative rate of respiration I 7.90—7.50 = 0.40 7.90-7.50 = 0.40 7-90-7-55 = 0.35 7.90-7.55=0.35 7-90-7-55=0-35 0.40-5-0.40=1.00 0.40-4-0.40=1 .00 0.35-^0.40 = 0.87 0.35^0.40 = 0.87 0.354-0.40 = 0.87 2 a 4 c TABLE II C* Two periods (35.5 min. each) in sea w.ater and during 6 equal periods in SEA water containing O . 3 PER CENT (bY VOLUME) OF CHLOROFORM Solution Change in PH 03- 03- -7 -7 43 40 = 0 = 0 60 63 03- -6 3ot = I 73 03- -6 78 = 1 25 03- -7 IS = 0 88 03- -7 55 = 0 48 03- -7 65 = 0 38 03- -7 85 = 0 18 Relative rate of respiration Sea water . Sea water containing 0.3 per cent chloroform Sea water containing 0.3 per cent chloroform Sea water containing 0.3 per cent chloroform Sea water containing 0.3 per cent chloroform Sea water containing 0.3 per cent chloroform Sea water containing 0.3 per cent chloroform 1.734-0.615 = 2.8 1 .254-0.615 = 2 .0 0.884-0.615 = 1 .4 0.484-0.615 = 0.78 0.384-0.615 = 0.62 o. 18-^0.615=0.29 * In the control, with 8 periods (31.5 min. each) in sea water, the change in PH in each period was 8.03 — 7.80 = 0.23. t Approximate (at this point the indicator is not very sensitive to changes in PH). 1 9 1 9] HAA S—RESPIRA TION 3 9 1 TABLE II D* One period (44 -5 ^i^'-) in sea water followed by equal period ix sea water CONTAINING O.5 PER CENT (bY VOLUME) OF CHLOROFORM Solution Change in PH Relative rate of respiration Sea water 8.10-7.35=0.75 8. 10—6.97 = 1 . 13 Sea water containing 0.05 per cent chloroform 1.13^0.75 = 1.5 * In the control, with 2 periods (39. 25 min. each) in sea water, the change in PH in each period was 8.10-7.35=0.75- (fig. 2a [B], fig. ih [B]; table II B) of formaldehyde respectively. In both figs. 2a and 26 (A) and (B), there is a marked increase in the respiration during the first period. The curves for the weaker concentration tend to become approximately horizontal in the later periods. The curves for the stronger concentration of formaldehyde present a somewhat different case. Here the respiration reaches its maximum during the first period and maintains this rate during the second period. Following this, the respiratory rate steadily becomes smaller. The curves for 0.3 per cent (fig. 2a [C] and fig. 2b [C]; table II C) and o. 5 per cent (fig. 2a [D] and fig. 26 [D] ; table II D) chloro- form respectively, each show that respiration is increased during the first period. The curves for the 0.3 per cent chloroform indicate that the rate steadily becomes smaller, until at the end of about 2.25 hours the respiratory rate falls below what it was normally (when in sea water). The observation has often been made (11) that in human beings and in mammals during prolonged anesthesia there are typical products of incomplete oxidation such as fatty acids, lactic acid, and above all acetone (in not inconsiderable quantities) eliminated, as the case may be, into the urine or into the respired air. It seemed of interest in this connection to study the effect of acetone upon the respiration of Laminaria. It will be seen from the curves that when sea water contains o . i per cent (fig. 3 [A] ; table III A) or 0.51 per cent (fig. 3 [B]; table III B) of acetone respectively, the respiration is practically unaffected. When, however, the sea water contains 17.4 per cent of acetone (fig. 3 [C] and [D]; table III C), a pecuHar condition results. During the first period the 392 BOTANICAL GAZETTE [may respiratory rate (curve D) is greatly increased and reaches its maximum during the second period. This is followed by a rapid decrease in the rate during the third and fourth periods, although the rate is still above the normal. After the fourth period a more REL, RATE OF RESP. REL.AMT.OFRESP. 60 120 MINUTES Fig. 3. — Curves (unbroken lines) showing effect of sea water containing A, 0.1 per cent acetone; B, 0.51 per cent acetone upon relative rate and relative amount of respiration of Laminaria (identical for these substances); C, effect of sea water con- taining 17.4 per cent acetone (unbroken line) upon relative amount of respiration; Z), effect of sea water containing 17.4 per cent acetone (unbroken line) upon relative rate of respiration; controls in sea water (broken lines); see tables III A to IIIC; each control bears the same letter (with a prime) as the experimental curve. gradual decline begins, so that even at the end of the experiment covering 2 hours and 48 min. the respiratory rate is still sHghtly above the normal. TABLE III A AND B Two PERIODS (23 MIN. EACH) IN SEA WATER FOLLOWED BY 2 EQUAL PERIODS IN SEA WATER CONTAINING O . I PER CENT ACETONE, FOLLOWED BY 2 SIMILAR PERIODS IN SEA WATER CONTAINING O.51 PER CENT ACETONE Solution Sea water Sea water containing o . i per cent acetone Sea water containing o . i per cent acetone Sea water containing 0.51 per cent acetone Sea water containing 0.51 per cent acetone Change in PH Relative rate of respiration 37-7.64 = 0.73 37-7.64 = 0.73 37-7-65 = 0.72 37-7-65 = 0.72 37-7-65=0.72 37-7-63 = 0.74 o.72-e-o.73=o.99 0.72-^0.73 = 0.99 0.72-^0.73 = 0.99 0.74-^0.73 = 1.01 jgig] HAAS— RES P IRA TION 393 TABLE III A AND B CONTROL Six periods (20.25 ^^in. each) in sea water Period Change in PH Relative rate of respiration I 8.37-7.60 = 0.77 8.37-7-60 = 0.77 8.37-7.61=0.76 8.37-7.60 = 0.77 8.37-7.60 = 0.77 8.37-7.60 = 0.77 0.77-5-0.77 = 1.00 2 0.77-5-0.77=1.00 1! 0.76-5-0.77 = 0.99 4. 0.77-7-0.77 = 1.00 c 0.77-^-0.77 = 1.00 6 0.77-^-0.77= I .00 TABLE III C Two PERIODS (24 MIN. EACH) IN SEA WATER AND 7 EQUAL PERIODS IN SEA WATER CONTAINING 1 7. 4 PER CENT (BY VOLUME) OF ACETONE Solution Change in PH Relative rate of respiration Sea water . Sea water containing 17.4 per cent acetone Sea water containing 17.4 per cent acetone Sea water containing 17.4 per cent acetone Sea water containing 17.4 per cent acetone Sea water containing 17.4 per cent acetone Sea water containing 17.4 per cent acetone Sea water containing 17.4 per cent acetone 8 37- -7 76 = = 0 61 8 37- -7 78 = :o 59 8 37- -6 80 = = I 57 8 37- -6 75 = = I 62 8 37- -7 06 = = I 31 8 37- -7 50 = = 0 87 8 37- -7 54 = = 0 .83 8 37- -7 62 = = 0 -75 8 37- -7 74 = = 0 -63 1. 57-7-0. 60 = 2. 6 1 .62-7-0.60 = 2 .7 1 .31-5-0.60 = 2 . 2 o.87-f-o.6o= 1 .5 0.83-^-0.60=1 .4 o.75-=-o.6o=i.3 0.63-7-0.60= X . I TABLE III C CONTROL Nine periods (21 min. each) in sea water Period Change in PH Relative rate of respiration 3- 4. 5. 6. 7- 8.37-7- 8 37-7- 8 37-7- 8 37-7- 8 37-7- 8 37-7- 8 37-7- 8 37-7- 8 37-7- .67 = 0. 70 .67 = 0.70 .70 = 0.67 .73=0.64 .73=0.64 .73=0.64 .72 = 0.65 .73=0.64 .72 = 0.65 o. 70 o. 70 0.67 0.64- 0.64 0.64- 0.6s 0.64 0.65 70=1 .00 70 = 1 . 00 70=0.95 70 = 0.91 70 = 0.91 70 = 0.91 70=0.93 70 = 0.91 70=0.93 394 BOTANICAL GAZETTE [mav In view of the fact that alcohol is considered to be formed dur- ing respiration, it was deertied important to study the effect of varying concentrations of alcohol upon the rate of respiration. When sea water contains i per cent of Squibb 's absolute alcohol (figs. 4(2 [F] and 46 [F]; table IV F), the respiratory rate remains normal for 3 periods, after which there is a gradual decline to below REL.AMT.OF RESP. 6 4JI^j jj^rg:5z:rg^::zpc 90 180 MINUTES Fig. 4a. — Curves (unbroken lines) showing effect upon relative amount of respira- tion of Laminaria of sea water containing ^, 24. 2 per cent ethyl alcohol; B, 16. i per cent ethyl alcohol; C, 10 per cent ethyl alcohol; D, 5 per cent ethyl alcohol; £, 2 per cent ethyl alcohol; F, i per cent ethyl alcohol; controls in sea water (broken lines); each control bears same letter (with prime) as the experimental curve, except that D' serves as control for curves D, E, and F; see tables IV A to IV F. the normal rate, whereas the relative amount of respiration remains nearly constant. The curves for sea water containing 2 per cent alcohol (figs. 4a [E] and 4b [E] ; table IV E) show a slight increase in the relative rate during the first period, followed by a smaller increase for the second period, after which there is a decline below the normal. As would be expected, 5 per cent alcohol (figs. 4a [D] and 46 [D]; table IV D) gives a much greater increase in the rela- iqiq] HAAS—RESPIRA TION 395 tive respiratory rate and amount than does 2 per cent alcohol (curve E). Unlike the 2 per cent alcohol, the maximum increase is maintained during the second period. The rate becomes less during the third period, after which it becomes quite constant. The curves for 10 per cent alcohol (figs. 4a [C] and 46 [C] ; table IV C) reach their maximum during the first period. Approximately the KEL.RATEOFRESP. 6 ac 90 180 MIN17TES Fig. 46. — Curves (unbroken lines) showing efifect upon relative rate of respiration of Laminaria of sea water containing A, 24 . 2 per cent ethyl alcohol; B, 16 . i per cent ethyl alcohol; C, 10 per cent ethyl alcohol; D, 5 per cent ethyl alcohol; E, 2 per cent ethyl alcohol; F, i per cent ethyl alcohol; controls in sea water (broken lines); each control bears same letter (with prime) as the experimental curve, e.xcept that D' serves as control for curves D, E, and F; see tables IV A to IV F. same rate of respiration is maintained for 3 periods, after which the increased rate rapidly becomes smaller. The relative amount of respiration remains approximately constant for 4 periods and then falls off very gradually. With 16 . i per cent alcohol (figs. 4a [B] and 46 [B]; table IV B) the maximum rate is not reached until the second period, after which the decline is more rapid than that for any of the lower concentrations. It will be seen that at the end of 396 BOTANICAL GAZETTE [may 224 min. the relative amount is about 1.5, while the relative rate falls below the normal after about 130 min. The maximum increased rate, when 24.2 per cent alcohol is used (figs. 4a [A] and 4b [A]; table IV A), is reached during the first period and then becomes smaller very much more rapidly than at any of the lower TABLE IV A Two PERIODS (31 . 25 MIN. EACh) IN SEA WATER AND 6 EQUAL PERIODS IN SEA WATER CONTAINING 24.2 PER CENT (bY VOLUME) OF ETHYL ALCOHOL Solution Sea water . u u Sea water containing 24 . 2 per cent ethyl alcohol Sea water containing 24 . 2 per cent ethyl alcohol Sea water containing 24 . 2 per cent ethyl alcohol Sea water containing 24 . 2 per cent ethyl alcohol Sea water containing 24.2 per cent ethyl alcohol Sea water containing 24.2 per cent ethyl alcohol Change in PH Relative rate of respiration 7.90-7.43 =0.47 43 =0-47 90-7 90 — 6 90 — 6 90-7 90-7 90-7 90-7 i6* = 75 = 25 = = 1.74 ^115 = 0.65 55 =0.35 80 =0.10 84 =0.06 1.74-4-0.47 = 3.7 i.i5-=-o.47 = 2.4 0.65^0.47 = 1.4 o-35-^o-47 = o.7 o. io-i-o.47 = o.2 0.06-^0.47 = 0. 1 * Approximate (at this point the indicator is not very sensitive to changes in PH). TABLE IV A CONTROL Eight periods (30.25 min. each) in sea water Period I. 2. 3- 4- 5- 6. 7- Change inPH 7.90-7.53=0.37 7 90-7 53 = o-37 7 90-7 53 = 037 7 90-7 54 = 0.36 7 90-7 54 = 0.36 7 90-7 55=0.35 7 90-7 55=0.35 7 90-7 58 = 0.32 Relative rate of respiration 0.37-^0.37 = 1.0 0.37-^0.37 = 1.0 0.37^0.37=1.0 0.36-7-0.37 = 0.97 0.36-7-0.37=0.97 0.35^0.37 = 0.95 0.35-^0.37 = 0.95 0.32-^0.37 = 0.87 concentrations, falling below the normal rate at about no min. The curve for relative amount of respiration at the end of 3 hours is about 1 . 4, even though respiration has nearly ceased. These curves, showing the effect of varying concentrations of alcohol upon the rate of respiration, indicate that for concentrations above i per cent there is a marked increase in the respiratory I9I9] HAAS—RESPIRA TION 397 activity. Furthermore, there is a maximum increase which is usually reached in the first, but sometimes not until the second period. The relative amount and relative rate of respiration in- crease with increasing concentrations of ethyl alcohol up to 10 TABLE IV B Two PERIODS (28 MIN. EACh) IN SEA WATER FOLLOWED BY 8 EQUAL PERIODS IN SEA WATER CONTAINING 1 6. 1 PER CENT ETHYL ALCOHOL Solution Sea water Sea water containing 16.1 per cent ethyl alcohol Sea water containing 16. i per cent ethyl alcohol Sea water containing 16. i per cent ethyl alcohol Sea water containing 16. i per cent ethyl alcohol Sea water containing 16. i per cent ethyl alcohol Sea water containing 16. i per cent ethyl alcohol Sea water containing 16. i per cent ethyl alcohol Sea water containing 16.1 per cent ethvl alcohol Change in PH Relative rate of respiration 9S-7-50 = o-45 95-7-52=0.43 95-6 95-6 95-6 95-7 95-7 95-7 95-7 95-7 76=1.19 55 = 140 75 = 1.20 20 = 0.75 57 = 0.38 70 = 0.25 85=0. 10 90 = 0.05 i.i9-=-o.44=2.7 1.4 -=-0.44 = 3.2 1.2 -^0.44=2.7 0.75-^0.44=1.7 0.38-^0.44 = 0.86 o.25-^o.44 = o.57 o. 10 -^ 0.44 = 0. 23 0.05-4-0.44 = 0. II TABLE IV B CONTROL Nine periods (28.5 mix. each) in sea water Period I 2 3 4 5 6 ■7 8 9 Change in PH Relative rate of respiration 7 95-7- 7 95-7- 7 95-7- 7 95-7- 7 95-7- 7 95-7- 7 95-7 7 95-7 7 95-7 55 = 0.40 o.4o-=-o.395 = i.oi 50 = 0.39 0.39-^-0.395 = 0.99 56 = 0.39 0.39^0.395 = 0.99 57 = 0.38 0.38^0.395 = 0.96 58 = 0.37 0.37^0.395 = 0.94 58 = 0.37 0.37^0.395=0.94 58 = 0.37 0.37-^0.395 = 0.94 58 = 0.37 0.37-^0.395 = 0.94 58 = 0.37 0.37^0.395=0.94 per cent. At concentrations from 2 to 10 per cent the relative rate of respiration remained far above the normal during the entire experiment. When larger concentrations such as 16. i or 24 per cent are used, however, the decline, once the maximum rate has been reached, becomes more, rapid with increasing concentrations 398 BOTANICAL GAZETTE [may of alcohol. The curves of the relative rate or respiration for such higher concentrations of alcohol fall quite rapidly below the normal, whereas the curves for the relative amount of respiration for the same experiment may remain far above unity. TABLE IV C Two PERIODS (39.75 MIN. EACh) IN SEA WATER FOLLOWED BY 6 EQUAL PERIODS IN SEA WATER CONTAINING lO PER CENT ETHYL ALCOHOL Solution Change in PH Sea water u u Sea water containing 10 per cent ethyl alcohol Sea water containing 10 per cent ethyl alcohol Sea water containing 10 per cent ethyl alcohol Sea water containing 10 per cent ethyl alcohol Sea water containing 10 per cent ethyl alcohol Sea water containing 10 per cent ethyl alcohol 7.70-7.55 = 0.15 7.7o-7.55 = o-i5 7.70-6.77 = 0.93 7 . 70 — 6.80 = 0.90 7 . 70 — 6.80 = 0.90 7.70—6.80 = 0.90 7 .70—6.90 = 0.80 7.70-7.05 = 0.65 Relative rate of respiration 0.93^0.15 = 6.2 0.90-i-O. 15 = 6.0 0.90^0. 15 =6.0 0.90-T-O. 15 = 6.0 0.80 -=-0.15 = 5.3 0.65 -=-0.15=4.3 TABLE IV C CONTROL Eight periods (38 . 5 min. each) in sea water Period Change in PH Relative rate of respiration I 7 70—7. 55 = 0.15 0. 15-i-0. 15 = I .00 2 7 7 7 7 7 7 7 70-7 70-7 70-7 70-7 70-7 70-7 70—7 55 = 0.15 53 = 0.17 55 = 0.15 55 = 0.15 55 = 0.15 55 = 0.15 55 = 0.15 0.15-^0. 15 = 1 .00 ■2 0.17-^0.15 = 1.13 A , 0.15^-0.15 = 1.00 e 0.15-5-0.15 = 1 .00 6 0.15-^0.15 = 1.00 7 o.I5-^-o.I5 = I .00 8 0. 15-j-o. 15 = 1 .00 It is important to know whether or not the decrease in PH value is due to the excretion of CO^ or to other acids, such as organic acids which are products of incomplete oxidation. To determine this pure hydrogen was bubbled through the solution (which has been made more acid by Laminaria) until the excess of CO2 was expelled. The solution was then allowed to come into equilibrium with the CO2 of the air. It was found that in all the experiments I9I9] HAAS— RES P IRA TION 399 (with certain exceptions to be mentioned) the color of the indicator was reversible by this means. This showed conclusively that the increased acidity was actually due to CO2. In the case of sea water containing 0.3 per cent chloroform, the color of the indicator TABLE IV D Two PERIODS (37.5 MIN. EACh) IN SEA WATER FOLLOWED BY 6 EQUAL PERIODS IN SEA WATER CONTAINING $ PER CENT ETHYL ALCOHOL Solution Sea water Sea water containing 5 per cent ethyl alcohol Sea water containing 5 per cent ethyl alcohol Sea water containing 5 per cent ethyl alcohol Sea water containing 5 per cent ethyl alcohol Sea water containing 5 per cent ethyl alcohol Sea water containing 5 per cent ethyl alcohol Change in PH 7 7 ^5- 65- ~7 ■ -7- 7 65- -6. 7 65- -6. 7 65- -7- 7 65- -7 7 65- ~7 7 65- -7 43=0.22 95=0.70 95=0.70 22=0.43 25=0.40 25=0.40 25=0.40 Relative rate of respiration O. 70-i-0.22 = 3. 2 o. 70^0. 22 = 3.2 0.43^0.22 = 1.9 0.40-7-0.22 = 1 .8 0.40-7-0. 22 = 1.8 O. 40 = 0.22 = 1.8 TABLE IV E Two PERIODS (39.25 MIN. E.'iCH) IN SEA WATER FOLLOWED BY 4 EQUAL PERIODS IN SEA WATER CONTAINING 2 PER CENT ETHYL ALCOHOL Solution Change in PH Relative rate of respiration Sea water . Sea water containing 2 per cent ethyl alcohol Sea water containing 2 per cent ethyl alcohol Sea water containing 2 per cent ethyl alcohol Sea water containing 2 per cent ethyl alcohol 765-7.35 = 0.30 •35 = 0.30 65-7 .65-7 65-7 65-7 65-7 20 = 0.45 28 = 0.37 43 = 0.22 52 = 0.13 0-45 -^0.30=1. 5 0-37-^0.30=1.2 0.22^0.30 = 0. 73 0.13-7-0.30 = 0.43 was not reversible at the end of the first period, although after any of the succeeding periods the color was fully reversible. The color of the indicator in sea water containing 17.4 per cent acetone (made up to the conductivity of sea water) was not reversible at the end of the first period of exposure. At the end of the second period of 400 BOTANICAL GAZETTE [may exposure the color was almost entirely reversible, and after the third period (and succeeding periods) the color was completely reversible. When sea water containing 24 . 2 per cent ethyl alcohol (made up to conductivity of sea water) was used, the color of the indicator after the first period of exposure was not reversible. The TABLE IV F Two PERIODS (34.75 MIN. EACh) IN SEA WATER FOLLOWED BY 5 EQUAL PERIODS IN SEA WATER CONTAINING I PER CENT ETHYL ALCOHOL Solution Sea water u u Sea water containing i per cent ethyl alcohol Sea water containing i per cent ethyl alcohol Sea water containing i per cent ethyl alcohol Sea water containing i per cent ethyl alcohol Sea water containing i per cent ethyl alcohol Change in PH Relative rate of respiration 65-7-40 = 0.25 6o-/ 65-7 65-7 65-7 65-7 65-7 35 = 0.30 35 = 0.30 35 = 0.30 35 = 0.30 40 = 0.25 43=0.22 o.3o-=-o.275 = i.i o.30-=-o.275 = i.i o.30-^o.275 = I.I 0.25-^0.275=0.91 O. 22-^0. 275=0.80 TABLE IV D, E, F CONTROL Six PERIODS (35.5 min. each) in sea water Period Change in PH Relative rate of respiration I 7-65-7-45 = o.20 7-65-7-45 = o.2o 7 65-7 -45=0. 20 7-6s-7-43 = o.22 7.65-7.43 = 0.22 7.65-7.43 = 0.22 0. 20 -j- 0.20= 1 .00 2 0. 20 — 0.20= I .00 •J 0. 20-i-O. 20=1 .00 4. 0, 22-^-0. 20= 1 . 10 e 0.22^0.20=1 .10 6 0.22-i-0.20=I.IO color of the indicator at the end of the second period of exposure was almost completely reversible, but after each subsequent period the color was fully reversible. The reason for these instances of irreversibility can easily be explained. Phenolsulphonephthalein in the acid end of its range is yellow. Laminaria has a yellowish brown pigment which may come out when the cells are injured by very strong concentra- tions of the anesthetic, while the green pigment does not come out. When the concentration of the anesthetic is so great as to rapidly injure the cells (as is the case with all of the exceptions just noted) , igig] HAAS— RESPIRATION 401 the yellowish pigment comes out so rapidly as to interfere with the indicator. As soon as the extraction of the pigment ceases (usually ending during the first 2 periods), the color of the indicator is not interfered with. When the concentrations of anesthetics are lower than in those instances just mentioned, the pigment comes out (if at all) so slowly as not to affect the indicator. This can be proven by matching the solution obtained after any period (without the addition of indicator) with the color of sea water (containing no indicator) in a tube of equal diameter. In such cases no pigment is detectable. Further evidence that it is CO2 that is being measured in the experiments rather than other acids is the fact that by the use of the gas chain it was found that Laminaria, after being 2 weeks in a small quantity of unchanged sea water, had given off no acid other than CO2. It might be supposed that the addition of so much alcohol as 24 . 2 per cent would dilute the buffer substances of the sea water so that a given amount of CO2 added to the mixture would produce more change in PH value than would be the case in sea water. This was largely avoided by concentrating the sea water before adding the alcohol, so that the amount of buffer substance remained the same in the mixture as in the sea water alone. Tests made by adding measured amounts of CO2 to sea water and to sea water plus 24 . 2 per cent alcohol (made up to the electrical conductivity of sea water) showed that there was not sufficient difference in this respect to be of importance in this investigation. Such experiments enable us to follow the respiration of the same piece of tissue during shorter or longer periods of exposure to various concentrations of anesthetics. They further show that in no instance was there an initial decrease in the rate of respiration. It will be observed that when the concentration of anesthetic is strong enough to produce any measurable result, the first effect is an increase of respiration, which gradually declines and may eventually fall below the normal. This decline is interpreted by the writer as a toxic eft"ect. These results are not in accord with the statements of Tashiro and Adams (27), according to whom anesthetics do not produce 402 BOTANICAL GAZETTE [may an increase of respiration except when their concentration is so low that they have only a stimulating action. They state that when the concentration is increased to the point where anesthesia occurs, the rate of respiration falls below the normal. It is evident that this is not the case with Laminar ia, for in no instance was the respiration observed to fall below the normal except after prolonged exposure to high concentrations which produced death. Further investigation will be necessary to deter- mine the cause of these discrepancies. It is evident that these experiments directly contradict the idea, advocated by Verw-orn (29) and his pupils, that anesthesia is a kind of asphyxia and that anesthetics act by reducing respiration. Summary When Laminaria is exposed to anesthetics (in sufficiently high concentration to produce any result) the initial effect is an increase of respiration. This may be followed by a decrease if the anesthetic is sufficiently toxic. No decrease of respiration is observed when the concentration is too low to be toxic. These results directly contradict the idea advocated by Verworn and his pupils that anesthetics act by decreasing respiration. Laboratory of Plant Physiology Harvard University LITERATURE CITED 1. Appleman, C. O., Relation of oxidases and catalase to respiration in plants. Amer. Jour. Bot. 3:223. 1916. 2. Battelli, F., and Stern, L., Einfluss der Anasthetika auf die Oxydone. Biochem. Zeitschr. 52:226. 1913. 3. Bonnier, G., and Mangin, L., Recherches sur Taction chlorophylliene separee de la respiration. Ann. Sci. Nat. \TI. 3:5. 18S6. 4. Burker, K., Eine neue Theorie der Narkose. Miinch. Mediz. Wochen- schr. 57:1443. 1910. 5. Czapek, F., Biochemie der Pflanzen. 1913 (195 ff.). 6. Ewart, a. J., The action of chloroform on CO2 assimilation. Ann. Botany 12:415. 1898. 7. Haas, A. R., A simple and rapid method of studying respiration by the detection of exceedingly minute quantities of carbon dioxide. Science N.S. 44:105. 1916. iQig] HAAS— RESPIRATION 403 8. Hamburger, E., Narkose und Sauerstoffmangel. Pfliiger's Arch. 143: 186. 1912. 9. Harder, R., Beitrage zur Kenntniss des Gaswechsels der Meeresalgen. Jahrb. Wiss. Bot. 56:254. 1915. 10. HoBER, R., Physik. Chem. der Zelle und der Gewebe, Chaps. 8, 9. 1914. 11. Irving, A. A., The effect of chloroform upon respiration and assimilation. Ann. Botany 25: 1077. 1912. 12. KosiNSKi. I., Die Athmung bei Hungerstanden und unter Einwirkung von mechanischen und chemischen Reizmitteln bei Aspergillus niger. Jahrb. Wiss. Bot. 37:137. 1902. 13. Lauren, W., Uber den Einfluss von Aetherdampfen auf die Atmung von Keimhngen. Diss. Helsingfors 1891. Cited from Just's Bot. Jahresb. 20:92. 1892. 14. LoEB, J., and Wasteneys, H., Is narcosis due to asphyxiation? Jour. Biol. Chem. 14:517. 1913. 15. jNIcClendon, J. F., The composition, especially the hydrogen ion con- centration, of sea water in' relation to marine organisms. Jour. Biol. Chem. 28:135. 1916. 16. Mansfeld, G., Narkose und Sauerstoffmangel. Pfliiger's Arch. 129:69. 1909. 17. Matthaei, G. L. C, Experimental researches on vegetable assimilation and respiration. III. On the effect of temperature on carbon dioxide assimi- fation. Trans. Roy. Soc. London B. 197:71. 1904. 18. JMayer, a., Landw. Versuchsstat. 1879. S. 335. 19. Meyer, H. H., Zur Theorie der Alkoholnarcose. Welche Eigenschaft der Anesthetica bedingt ihre narcotische Wirkung ? Arch. Exper. Pathol. 42:109. 1899. 20. NoTHMANN-ZucKERANDL, H., Die Wirkung der Narkotica auf die Plas- mastromung. Biochem. Zeitschr. 45:412. 191 2. 21. Overton, E., Studien iiber die Narkose. 1901. S. 51. 22. Pantanelli, E., Athmung der Meeresalgen. Ber. Deutsch. Bot. Gesells. 32:488. 1914. 23. Richards, H. M., The respiration of wounded plants. Ann. Botany 10:551. 1896; The evolution of heat by wounded plants. Ann. Botany 11:29. 1897. 24. Schroeder, H., tjber den Einfluss des Cyankaliums auf die Atmung von Aspergillus niger nebst Bemerkungen iiber die Mechanik der Blausaure- Wirkung. Jahrb. Wiss. Bot. 44:409. 1907. 25. SoRENSEN, S.P.L., Compt. Rend. Lab. Carlsberg 8:1. 1909. 26. Tashiro, S., Carbon dioxide production from nerve fibers when resting and when stimulated; a contribution to the chemical basis of irritability. Amer. Jour. Physiol. 32:107. 1913. 27. Tashiro, S., and Adams, H. S., Studies in narcosis. I. Effect of ethyl urethane and chloral hydrate on the CO2 production of the nerve fiber. Internal. Zeitschr. Phys.-Chem. Biol. 1:450. 1914. 404 BOTANICAL GAZETTE [may 28. Vernon, H. M., The function of lipoids in tissue respiration and in the activity of oxidases. Jour. Physiol. 45: 197. 1912; Die Abhangigkeit der Oxydasewirkung von Lipoiden. Biochem. Zeitschr. 47:374. 1912. 29. Verworn, M., Irritabihty. 1913 (Chap. IX). 30. Warburg, O., tjber Beeinflussung der Sauerstoffatmung. Zeitschr. Physiol. Chemie. 70:413. 1911; see also Miinch. Mediz. Wochenschr. 1911, No. 6; Oxydationen in lebenden zellen nach Versuchen an rothen Blutkorperchen. Zeitschr. Physiol. Chemie 66:305. 1910. 31. WiNTERSTEiN, H., Beitrage zur Kenntniss der Narkose. I Mittheilung. Kritische tJbersicht iiber die Beziehungen zwischen Narkose und Sauer- stoffatmung. Biochem. Zeitschr. 51:143. 1913. BASIS OF SUCCULENCE IN PLANTS D. T. MacDougal, H. M. Richards, and H. A. Spoehr Succulents may be characterized as plants in which the paren- ch>Tnatous elements show an exaggerated development with rela- tion to the more rigid tissues, and, unlike pith or medullary tracts, the masses of thin-walled cells remain distended and turgid. The liquid contents of such cells may or may not contain much dissolved material. The disposition of the water-holding tracts varies from leaves to stems and roots, but in all cases the most important general effect is one of massiveness, and the surfaces of succulent plants may in such forms as the barrel cacti bear the smallest possible proportion to the mass, that of a globe. The ecologist recognizes two general types of succulents, those of the arid regions, which are of a xerophytic character, exemphfied by the cacti; and the halophytes or fleshy seashore plants, also at home in alkaUne areas. The plants of the two types are quite unlike in their transpiratory relations. The desert succulents may lose water so slowly that an existence of several years may be main- tained upon the water in the thin-walled tracts.' On the other hand, the halophytes or fleshy shore plants may flag and wilt as readily as any thin-leaved form, due to the rapid loss of water from the surfaces. The origination of these striking forms has been the subject of much speculation, but all attempts to connect suc- culency in a causal way with the presence of salts in the soil, or in the plant with the well known high acidity of many of these forms, or with any purposeful development of water storage capacity, have been inadequate. Our concurrent observations and experiments may be briefly summarized as follows: I. A Castilleja native to the region about the Coastal Labora- tory, at Carmel, California, includes two habitat forms, geneticaUy identical, one with thin leaves growing in the open forest formation, ' MacDougal, D. T., et al., End results of desiccation and respiration in succulent plants. Physiol. Researches i : 2 89-3 25. 1 9 1 5 . 405 4o6 BOTANICAL GAZETTE [may and another with fleshy leaves growing on the sandy foreshores under arid, but not sahne, soil conditions. The succulent leaves owe their increased thickness to the enlargement of elongated cells vertical to the surface of the leaves. 2. The thin leaves show an acidity double that of the fleshy t>pe, and have a relatively greater dry weight. 3. The fleshy leaves, fresh and in a dried condition, present swelling reactions similar to those of sections of the joints of platy- opuntias, indicative of cells high in pentosans, or mucilages. The behavior of these organs is different in many important particulars from that of thin leaves, which swell more in acid than in alkaline solutions, the reverse taking place in succulent leaves, in parallelism to Opuntia. 4. Differences in the swelling reactions of dried leaves of both kinds are to be ascribed to the adsorption of the contained acids and salts of different amounts in the two cases on cell colloids, high in pentosans in one case and hence presenting characteristic coagulatory effects. 5. It has been established by researches not described in this paper that the reduction of the water content of the cell below a certain point results in the conversion of polysaccharides, which do not show a high imbibition capacity, to pentosans, which mixed with nitrogenous substances have an enormous hydration capacity. 6. Succulence, therefore, may originate as it is seen to occur in Castilleja as a direct result of aridity. Species of Ericameria and Erigeron with a distribution similar to Castilleja display thin and succulent leaves corresponding in the same manner to the environment. 7. High acidity may not be taken as a result of succulence. It is probably more nearly correct to assume that succulence may develop only in plants which have a carbohydrate metabolism characterized by large acid residues. The bulk and durability of succulents have made them readily available for chemical studies, and these features are responsible in part for the fact that carbohydrate metabolism and respiration, photosynthesis, the formation and fate of acids, the oxygen-carbon- dioxide ratio, and other features have all received contributions iQigl MacDOUGAL, RICHARDS, 6= SPOEHR— SUCCULENCE 407 based upon researches carried out with the fleshy plants.^ The amount of detailed and systematic information concerning these plants as a t>pe is probably greater than that of any other ecological group, and it is evident that their metabolism presents some definite characteristic aspects. It is upon the basis of such knowl- edge that it becomes possible to formulate the generalizations set forth in this paper. The only comparisons between succulents and non-succulents that have been possible have lacked directness because the reactions of different species could not be rated against each other with accuracy. The final and necessary conditions for a critical dis- cussion of the matter, that of succulent and non-succulent individ- uals of the same species, finally came to the attention of the authors in the case of Castilleja lalijolia at Carmel, California, in the summer of 19 1 8. One form of this plant which grows on the edge of the bluff overlooking the beach, or within 25 ft. of if, has succulent leaves of considerable thickness which are usually pale green. The other, which grows farther back on the foreshore, has a thinner, mesophytic type of leaf which is darker green and more hirsute than the succulent type. This is probably the t>'pical form of the manuals and is similar to the one which grows farther back in the pine woods. A notable exception as to the relative region of growth of these 2 forms was found in a luxuriant growth of the thin-leaved type at the base of the beach bluff on the edge of the sand. Exami- nation showed that this was unquestionably due simply to the ample water supply from the seepage at the base of the cHff. It becomes evident, therefore, that the difference in the succulent and mesophytic habit is not a case of even partial halophytism, for if salt were present anywhere it would be at the chft" base. The contrast- ing habit is one rather of xerophytism versus a mesophytic growth. The members of the genus are reported to be parasitic, and individuals with thin and others with succulent leaves were found ^ Spoehr, H. a., Photochemische Vorgange bei der Diurnalen Entsaurung der Succulenten. Biochem. Zeitsch. 57:95-111. 1913. Richards, H. M., Acidity and gas interchange in cacti. Publ. no. 209. Carnegie Inst. Wash. 1915. Hempel, Jexxy, Buffer process in the metabolism of succulent plants. Compt. Rend. Carlsberg. 13: no. i. 191 7. 4o8 BOTANICAL GAZETTE [may with minute roots attached to the older tapering roots of Artemisia pycnocephala, in a manner indicating that the dependent nutritive relation is not an important one. That the appearance of suc- culence in this plant has no connection with its parasitism is supported by the fact that a similar state was found in Erigeron glaucus and Ericameria ericoides which are found in the locality near Castilleja. Measurements of the thickness of the leaves of the two types show that the thin leaf averages about o .5 mm., while the succulent one ranges from i to i . 5 mm. Examination of sections shows that the structure of the leaves is mainly differentiated by the size of the cells. While nearly dorsiventral in position when young, structurally the leaves appear almost bifacial, with 2 or 3 rows of vertically elongated cells on each face. In the thin-leaved type these cells are about 35X25 ju- while similar cells of the succulent leaf are 110X30 /x. Thus the increased thickness of the leaf is due chiefly to the enlargement of the cells vertical to the flat surface of the leaf. There is naturally a disparity in the relation of fresh to dry weight in the 2 forms. Averages of a number of determinations show the following figures: mesophytic t^^^e, i gm. fresh young leaves, o. 193 dry weight; succulent type, i gm. fresh young leaves, o . 1 13 dry weight. In general the succulent type yields only three- fifths dry substance per unit fresh substance compared with the other form. The acidity relations are also different. The succulent leaves are much less acid than the thinner ones. In table I averages of 10 or more determinations indicate the difference in acid extracted, and also the amount of water absorbed after 24 hours' immersion. As might be expected, the acidity relation of the 2 forms approxi- mates more closely when reckoned according to dry weight than on the basis of fresh weight. Even then, however, the young active leaves show a considerably greater acidity in the thin type, a notable departure from our preconceived conceptions of acidity in relation to succulence. It may be said, however, that direct comparisons of acidity of a plant of the same species in a mesophytic and succulent condition do not seem to have been made. igigl MacDOUGAL, RICHARDS, 6- SPOEHR—SUCCULENCE 409 In connection with the swelling measurements described later, it became important to ascertain how much acid would leach out during immersion in water. Table II indicates the averages of a considerable number of series. It is noticeable that the total amount of acid is somewhat greater than the figures previously given, due very possibly to the forma- tion of acid during immersion in water, which might be caused TABLE I Material Total acidity Water absorbed, Per gm. fresh Per gm. dry cc. PER GM. FRESH WEIGHT Thin young leaves . Succulent young leaves 1.50CC. N/20 KOH 0.62 N/20 1 . 00 N/20 0.65 N/20 7.78 CC. N/20 KOH 5.37 N/20 5.20 N/20 5 . 74 N/20 I. 14 0-57 0.85 0 35 Thin old leaves. . . . Succulent old leaves TABLE II Leaves immersed in w.\ter for 24 hours at i7°c. Material Acid diflfused out per gm. fresh weight Acid retained in tissue per gm. fresh weight Thin young leaves I.30CC. N/20 KOH 0.55 N/20 0.62 N/20 0.30 N/20 0.39 CC. N/20 KOH 0.28 N/20 0.66 N/20 0.41 • N/20 Succulent voung leaves Thin old leaves Succulent old leaves by the exclusion of oxygen. It is also to be observed that the residual acid in the young leaves is closely the same in both the succulent and thin-leaved type. It appears that in the old leaves proportionately less of the acid leaches out and more is retained in the tissues. Other series of experiments were undertaken to determine the rate at which the acid diffuses out, the results of which are given in table III. During the first hours of immersion the amount of acid which passes out is small and nearly equal in each case. As prolonged immersion in water kills the leaves, it seems probable that very little acid escapes as long as the cells are alive.^ Table IV indicates 3 Lauk, E., Die Bedeutung der Elektrolyten fuer Quellungsprocess. Biochem. Zeitsch. 37: 15-58. 1916. 4IO BOTANICAL GAZETTE [may the length of immersion which the 2 forms can withstand and recover to a seemingly normal condition. After 12 hours' immersion in water all the thin-leaved shoots were killed; some of the succulent leaved shoots survived partly. TABLE III Rate of diffusion of acid in terms of cc. N/20 koh per gm. fresh weight AT I7°C. Material 3 hours 6 hours 9 hours 1 2 hours IS hours 24 hours Residual acid Total Thin young leaves Succulent young leaves O.IO O.IO 0.20 0.12 0.27 0.18 0.30 0.15 015 0.05 0.05 Trace 0.30 0.23 1-37 0.80 TABLE IV Material 1 . 5 hours 3 hours 6 hours 9 hours Thin young leaves Succulent young leaves . All recov- ered All recov- ered All recov- ered All recov- ered Some killed Some killed More than half killed About half killed TABLE V Swelling of leaves of Castilleja at 16° c. Material Water Citric acid O.OI N Potassium hydrate O.OI M Potassium nitrate O.OI M Thin Percentage 149 143 Percentage 120 95 Percentage 60 125 Percentage 76 Succulent 90 The reactions of the succulent leaves are seen to present the general aspects of sections of joints of Opuntia discata grown at Carmel and tested at 16° C. at Carmel, July 1918. Swellings of dried slices are as follows: Material Water Citric acid O.OI N Potassium hydrate O.OI M Potassium nitrate O.OI M Fresh sections. . . Dried slices Percentage 9 509 Percentage 10 218 Percentage II. 7 413 Percentage 10.4 417 I9I9] MacDOUGAL, RICHARDS, b- SPOEHRSUCCULENCE 411 A set of sections were now prepared, and, after being swelled with reactions parallel to the preceding, were dried, and the expan- sion when immersed was calculated on the original thickness, as follows : Material Water Citric acid 0.01 N Potassium hydrate 0.01 M Potassium nitrate 0 . 01 M Fresh sections . . Dried sections . . Percentage II. 4 17s Percentage 6.4 26.4 Percentage 6.6 24s Percentage 9.2 22.4 Determinations were also made of the escape of acid from dried leaves, which, as shown in table VI, is much more rapid than with the living material. It will be seen that the total of the acid extracted from the dried leaves is much less than that obtained from fresh leaves, which, as might be expected, indicates that some of the TABLE VI Escape of acid from dried leaves in terms of cc. N/20 koh per gm. fresh WEIGHT AT I7°C. Material 2 hours 6 hours 12 hours IS hours 24 hours Residue Total Succulent young leaves . . Thin young leaves 0.30 0.54 O.IO 0.20 0-5 0.8 Trace Trace Trace Trace 0.12 0.40 O.S7 1 .22 acid salts are absorbed and held in the irreversible aggregation phenomena connected with the processes of drying. It may be mentioned that two series of both kinds of leaves, which by chance were dried much more slowly than the others, showed a difference in the rate at which they yielded up their acid; in both cases the amount which escaped in 2 hours was much less than in the case of the rapidly dried leaves. When trios of leaves of the 2 types were placed under the auxo- graph to determine their unsatisfied hydration capacity, sweUings as follows (table VII) were displayed at 16° C. As will be seen by comparisons with data obtained from Castil- leja, the dried mass behaves like the succulent leaves by showing but little expansion after immersion and drying. Similar tests were appHed to sections and to dried median slices of an unknown Opuntia which appeared to be less muci- laginous than 0. discata. Dried slices came down to a thickness of 412 BOTANICAL GAZETTE MAY about 0.2 mm. and these gave swellings at i6° C. which are to be compared with the swellings of fresh material. The second swelling produced an expansion scarcely more than half that of the first in all solutions, and being still further decreased in alkali, furnishing striking parallels with the action of succulent leaves of Castilleja. TABLE VII Process Water Citric acid O.OI N Potassium hydrate O.OI M Potassium nitrate 0.01 M After first drvine Percentage 361 42 9-7 Percentage 306 56 7 Percentage 250 100 Percentage ^2S After second drying on basis of re- duced thickness 75 Fresh sections through joints 0- ^ TABLE VIII Hydration reactions of succulent and thin leaves of Castilleja; JULY 28-31; at 16° c. Material Water Citric acid O.OI N Potassium hydrate O.OI M Potassium nitrate O.OI M Succulent Fresh Swelled and dried . . . Fresh dried Percentage 143 20 76 140 132 100 Percentage 95 16 60 120 12 20 Percentage 125 20 55 60 67 118 Percentage 90 60 65 Thin leaves Fresh 76 Swelled and dried . . . Fresh dried 153 130 Two additional treatments of the leaves were given to test the effects of hydration on the swelling capacity of the contained colloids. In one case the trios of sections which had swelled were dried on filter paper for a day at 20° C, with only enough pressure to prevent warping or curling, then again hydrated in water or the identical solutions of the first swelHng. The second case included a swelling of leaves which had been simply dried for a day at 20° C, in which process they came down to about half the original thick- ness. The measurements at 17-18° C, calculated on dried thick- ness, which was usually about one-half that of living material, are given in table VIII. 1919] MacDOUGAL, RICHARDS, ^- SPOEHR— SUCCULENCE 413 The comparisons which may be made upon the basis of such data are almost endless, and a citation of even the salient features of interest cannot be made briefly. The proportionate hydration of the succulent and thin leaves are reversed in acid and alkali. The succulent leaf, which proves to be one-half as acid as the thin leaf, swells most in the alkaline solution; while the thin leaves, with an acidity double that of the thick succulent ones, have an equivalent maximum in hundredth normal citric acid, and take up only half as much water in the alkaline solution, the disproportion between the two expansions being greater than that of the acid alkali ratio in the succulent. The thin leaves are characterized by a uniformly high hydration capacity in water in the 3 cases, although reaching a maximum in the salt, a high swelling capacity in acid when fresh, which undergoes a great reduction after drying, while the swelling capacity increases in alkali in parallel treatments. The maximum swelling of the succulent leaves is in water, with great variation in the 3 conditions in which leaves were tested, and with but little variation in the reactions in the salt. The thin leaves, on the other hand, show the maximum and greatest diversity in the salts and more uniformity in water. The variations in swelling in the acid solution presented such unusual features that an additional series was planned in which thin and succulent leaves in fresh condition were swelled, then such leaves fully hydrated in water and in various solutions were dried and swelled a second time for comparison with the reactions of leaves dried directly from the living condition. Table IX shows the swellings in o.oi normal citric acid at 15° C. The chief departure from the original series is in the matter of the swelling of the fresh succulent leaves, which in this case appear to have been in such a highly hydrated condition as only to be capable of slight expansion. This assumption is in accordance with the fact that after being immersed and then dried they assumed approximately the original thickness on a second swelling. The thin leaves of this series were consistent in their reactions with those previously examined, showing a relatively small expansion from a dried condition. The conditions making possible the greater variations are evidently those recognizable in the succulent type, 414 BOTANICAL GAZETTE [may not the least important feature being the greatly enlarged paren- chymatous cells. It is to be seen that immersion and drying, and also simple drying, reduce the swelling capacity of thin leaves in acid, but no such decrease occurs in the succulent leaf. The principal changes which take place in swelling consist in the extraction of acids and acid salts, as indicated on the previous pages, and of the hexoses as yet undetermined. Any mucilages or pentosans present would of course diffuse at a rate so slow as to be of no consequence in the present experiments, TABLE IX Material Fresh leaves Above leaves dried and rehy- hydrated* Fresh dried leaves* Thin Thickness mm. / 0.4 \ 0.41 / 0.23 \ 0.25 (0.38) O. 2 (0.38) 0.2 Swelling Percentage 125 184 42 20 25 62 Succulent Thickness mm. 1-4 1-4 0.5-0.6 0.63 (1.2) 0.5 (l.i) 0.38 Swelling Percentage 21 25 95 91 120 92 * Expansion in terms of dried thickness. The swelling of fresh leaves of both types in water reaches the limit in less than 2 hours, the rate of extraction of acid in the 2 types of leaves being equivalent, and the proportionate expansions not widely different. When such leaves are dried the thin leaves attain the limit in water inside of 2 hours, while the succulent leaves continue to expand for 6 hours with an escape of acid about half that from the thin leaf during the same time. If attention be turned to the reactions in acids, it is seen that thin leaves swell more than succulents in such solutions when fresh, and that the swelling extends over a greater length of time, while the total swelling in a dried condition is accomplished in a few minutes. The succulent leaves, on the other hand, require a period of as much as 6 hours to reach full hydration from a dried condition. While the effect of the residual acidity is discernible in some of these relations, it is evident that this factor is not the dominating iqiq] MacDOUGAL, RICHARDS, &- SPOEHR— SUCCULENCE 415 one. Two other features remain to be considered, that of the composition of the plasmatic colloids, and of the salts dissolved in the water of hydration. The colloids of living leaves are highly kydrated, and the salts, acids, etc., are also in a highly dilute condition, in which case their effect would be at a minimum. Death and desiccation would be accompanied by a concentration of these compounds, until finally they would be adsorbed by the cell walls and plasmatic colloids in their most concentrated condition with resulting coagulations, some of which in all probability are irre- versible. The thin leaves have a higher acid content, and, to anticipate, a smaller proportion of pentosans which would accen- tuate this effect, hence the relatively low coefficient of swelling from a dried state. A long series of experiments with sections of dried colloids and of living and dried plants of known composition make it appear that the water relations of active tissues show the behavior of a biocolloid consisting largely of pentosans, of which agar or plant mucilages would be an example, a small proportion of protein or protein derivatives, and some salts and free acids.'' It is to these features, therefore, that one would naturally turn for the factors which might increase the water-holding capacity of the cell or organ, and in so doing the pentosans would claim attention first. These substances probably are always present in some proportion in cells, and their occurrence is therefore not significant. Any action or condition which brings about a notable increase in their proportion in the cell would have most important consequences however. Such increase does result from a depletion of the water of a cell, for the polysaccharides under such conditions are reduced to the pentosans, and the reduction of the water content of a cell results in the conversion of the polysaccharides, which do not show marked imbibition, to pentosans, which take the form of an elastic gel with an enormous capacity for expansion, particularly when mLxed with nitrogenous material, and upon this rests the hyper- trophies or hyperplasias of thin- walled tracts in the development of '' MacDougal, D. T., Imbibitional swelling of plants and colloidal mixtures. Science 44:502. 1916. MacDougal D. T., and Spoehr, H. A., The effect of acids and salts on biocolloids. Science 45:269-272. 1917. , Growth and imbibition. Proc. Amer. Phil. Soc. 56:289-352. 1917. , The behavior of certain gels useful in the interpretation of the action of plants. Science 45:484-488. 1917. 4i6 BOTANICAL GAZETTE ' [may succulence in an organ. Briefly restated, whenever the water content of a cell becomes low, some of the hexose-polysaccharides, which have a low imbibition capacity, are converted into pentosans, which have a high hydration capacity, the action having the force of a regulatory adjustment, and as the change is irreversible, the pentosans are accompanied by a permanent succulence, with all of the implied alterations in metabohsm,^ including a very striking change in the type of respiration, or of transformations in the carbohydrates.^ It is notable that, while this change in the sugars takes place in the cell, the type of transformations of energy changes completely, but the approximate rate of respiration is not materially affected. The nature and amount of the end products, however, may dift'er materially from those of a respiration in thin leaves, notably in the acid residues. It is in the mesh of reactions indicated that the origin and the nature of succulence will be found, and whatever causal value is attributed to the action of soil salts or of arid con- ditions will rest upon their part in the conversion of the poly- saccharides to pentosans. Acidity in succulents has been attributed by many writers, including the authors of this article, to the imperfect oxidations resulting from the lessened aeration of massive tissues, leaving a residue of mahc acid, for example. Castilkja, however, presents the example of highly acid thin leaves, which become succulent under conditions similar to those which favor the transformation of polysaccharide to pentoses in other plants. Instead of acidity being a direct result of succulence, it is much more reasonable to conclude that high acid residues may be characteristic of plants which present a metabolic complex favorable to pentose formation and to the development of succulence under certain environic conditions. s Spoehr, H. a., The pentose sugars in plant metabolism. Plant World 20:365. 1918. 'MacDougal, D. T., and Spoehr, H. A., The origination of xerophytism. Plant World 2 1 : 245-249. 1918. Coastal Laboratory Carmel, California A CONIFEROUS SAND DUNE IN CAPE BRETON ISLAND LeRoyH. Harvey (with eight figures) Nova Scotia has been called "the long wharf of Canada." Cape Breton Island, which is cut off from the mainland by the Gut of Canso, may be likened to its outermost pier. The island (fig. i), which is about loo miles long by 30 miles wide in its northern portion, extends in a northeasterly-southwesterly direction, re- straining the waters of the Gulf of St. Lawrence on the west and separating them from the Atlantic Ocean on the east. The latitude of 47° north cuts the northern end of the island a few miles to the north of Aspy Bay, on whose shores the coniferous sand dune is located. Nova Scotia lies in the coniferous belt, which occupies the upland with the mixed hardwood formation occupying the most favorable situations along the narrow coastal strip. The interior is occupied by a vast expanse of wet and dry tundra-like formations, bordered by gnarled and twisted dwarf spruce, the entire vegeta- tional aspect being decidedly coastal rather than alpine.' The country is extremely rugged and the coastline jagged. Along the east coast is a narrow strip of sloping land, rarely a mile wide and often entirely lacking, which soon rises abruptly into an upland about 1000 ft. above sea level. In some places this upland plunges precipitously into the sea and the coast is very wild and bleak. This old Atlantic upland, which forms the backbone of the island, is the northern extension of the Piedmont Plateau. This upland has been cut during eons of erosion into deep gulches which extend far back into the central plateau. Down these gulches run swift and bowlder-bedded streams to the sea. At the mouth of these streams intervales are formed. Storm and tidal action have thrown' shingle beaches and sand spits across the mouths of many of these reentrant bays (fig. 2). Upon one of these sand spits inclosing the South Pond of Aspy Bay is located the sand dune area which forms 'Nichols, G. E., The vegetation of Cape Breton Island, Nova Scotia. Trans. Conn. Acad. Sci. 22:251-467. 1918. 417 4i8 BOTANICAL GAZETTE [may the subject of this study. It is the only sand dune noted on the island. The area is located some 5 miles from the crest of the up- land and is fully exposed on its western side to the terrific northwest I I Fig. I. — Cape Breton Island: Atlantic upland represented by dark shading; sloping marginal strip of lowland in light; Aspy Bay indicated by arrow. winds, as well as to the cutting action of the outrunning tides from South Pond. The sand spit juts out to the southeast from a rocky upland reaching a total length of approximately § mile, with a maximum width at the present time of less than 650 ft., and with a I9I9] HARVEY—CONIFEROUS SAND DUNE 419 minimum of less than 400 ft. But for the swift running tidal currents the spit would completely impound South Pond, extending across to the other headland; at best only a shallow and narrow channel now exists. Fig. 2. — Aspy Bay region: sand spit on which sand dune is located marked by arrow; light shading indicates upland; dark shading indicates lowland. The present condition of the area may be seen from fig. 3, which attempts to show the distribution of the existing plant asso- ciations. The dune complex, which is some 300 ft. wide, occupies less than one-half the length of the spit, being replaced in part at its southeastern extremity by a middle beach 400 ft. in width. 420 BOTANICAL GAZETTE [may On the east the dune complex is fronted by a middle and lower beach, each with a width of approximately loo ft. The lower S B o u u o en en di a a o 3 ^ en -a 1=1 (U en U >.< CO o beach maintains its average width on the east side of the duneless end of the spit, where it has a gentle slope, but on its west side the iqiq] HARVEY— coniferous SAND DUNE 421 lower beach occupies a narrow steep margin only a few feet wide. To the immediate west of the spit in South Pond are several iso- lated sand islands (S.I.) mostly covered at high tide. To the west of the dune complex a transitional zone some 50 ft. wide separates it from an extensive salt marsh which is about a foot lower, and whose average width is estimated at 1 200 ft. Several narrow salt water lagoons traverse this area in a north-south direction. Ex- tending out into the marsh some 600 ft. along the eastern border is an area occupied by 10 or 15 old white pine stumps, approximately 100 years old, with well exposed roots, and standing in rows more or less parallel to the axis of the spit. The south end of the salt marsh and dune complex are suffering very active erosion under the daily outgoing tidal currents. High tides and occasional storms ap- parently sweep completely over the low duneless extremity of the spit, greatly augmenting this erosion. We may now consider each of these associations in greater detail. Middle beach The middle beach, which is extremely barren, is composed mostly of a fine sand, but shingle of a coarser nature is not wanting. The usual debris of the middle beach is encountered only to the east of the dune complex (fig. 4). The 3 principal plants are Mertensia' maritima, Euphorbia polygonifolia, and Ammophila arenaria, but all are exceedingly scattered. The only other species noted were Glaux maritima, Lathyrus maritimus, Salsola Kali, Cakile edentula, and these are represented only by occasional individuals. Dune complex The facies of the dune complex is Picea canadensis with occa- sional Abies balsamea. At the northern end the older trees were estimated at 75 years, while those at the southern, eroding end are scarcely 30-40 years of age. The southern (fig. 5) and eastern fringes of the complex include the highest dunes, which range from 3 to 15 ft. above sea level, and are generally margined by a narrow "grassy foredune'' (fig. 6) with its precipitous slope oceanward. The sand binder is Poa compressa, a most unique condition. Am- mophila, although scatteringly present, is of little importance in 2 Nomenclature of Gray's New Manual, 7th ed. 422 BOTANICAL GAZETTE [may Fig. 4. — Southeast end of dune complex, from east, showing middle beach, the Picea-Abies stand, and grassy foredune; South Pond and highland beyond is shown at extreme left; photograph by Dr. G. E. Nichols. Fig. 5. — Southern and eroding end of dune complex, from south: South Pond and highland to west seen at left; middle beach is well shown; evidence of recent and rapid erosion plainly evident. iQig] HARVEY— CONIFEROUS SAND DUNE 423 this respect. Associated species are Euphorbia polygonij'olia, Taraxacum, Iris versicolor, Rhus Toxicodendron, Rubus sp., and Ribes oxycanthoides, all of which occur sporadically. On the thickly wooded lea slope of these grassy dunes are found numerous woody and herbaceous species. The most prominent Fig. 6. — Grassy foredune held by Poa conipressa, from north: middle beach and naked spit seen to south; highland shown to south of South Pond (barely visible); photograph by Dr. G. E. Nichols. species are Iris versicolor, Campanula rotundifolia, Vaccinium Vitis-Idea, V. pennsylvanicum, Maianthemum canadense, Rhus Toxi- codendron, Ribes, and Rubus. The occurrence of these forms is sporadic. In some places this outer range of dunes passes toward the west into low areas of considerable extent occupied by a unique associa- tion (fig. 3, blank areas in dune complex). Its aspect is grassy, 424 BOTANICAL GAZETTE [may determined mainly by Festuca rubra, Danthonia spicata, Agrostis maritima, and Panicum implicatum in rather open formation. Lecliea intermedia, Vaccinium Vitis-Idaea, Potentilla trident ata, Fra- garia virginiana terra-novae, Juniperus horizontalis , Empetrum ni- grum, Barbula, and a species of moss form more or less extensive mats. Other more scattered species are Campanula rotundifolia, EupJirasia americana (?), Cerastium arvense (?), Solidago hicolor, Plantago maritima, Iris setosa (?), Veronica serpyllifolia, Arenaria lateriflora, Plantago major, and several ruderals. Fig. 7. — Picea canadensis showing layering; individual trees plainly seen in center of fig. 5. A second and in some places a third series of much lower dunes is met in transect to the west. At the northern end of the complex practically all the white spruce is excessively infected with Ar- ceuthohium pusillum, presenting the most remarkable development of witches' brooms it has ever been my privilege to see. We have here a most remarkable physiographic condition of a dune moving seaward. The trees have mostly germinated at a lower level, and as the sand blows over the rounded top of the "grassy foredune" it forms a gentle lee slope to the west among these trees. As the trees are covered, abundant layering takes place, giving a long-lived and self -perpetuating stand (fig. 7) . There I9I9] HARVEY— CONIFEROUS SAND DUNE 425 is some evidence, however, that germination actually takes place on these grassy dunes. Through layering and germination the com- plex slowly moves ocean ward. Salt marsh The aspect of the salt marsh is determined by Spartina glabra, S. patens, and Disticblis spicata. J uncus balticus littoralis is very abundant along the drier margins. Other common species are Salicornia eiiropaea, Polentilla pacifica, Ranunculus Cymbalaria; Fig. 8. — Salt marsh from eastern margin; Piniis Strobiis stump in foreground; lagoons and South Pond in background; photograph by Dr. G. E. Nichols. while Vaucheria and Cladophora occur in great mats on the margins of pools. The most striking feature, however, is the presence of numerous white pine stumps (iig. 8), remnants of a lumbering opera- tion, whose distribution simulates rows parallel to the axis of the spit extending from the eastern shore out into the marsh to a dis- tance of several hundred feet. The roots of these stumps are well exposed. It is evident that they must have germinated upon land possibly a foot or more higher than this, somewhat over a century ago. It is also evident that the salt marsh has encroached from the west and is moving eastward. Coastal elevation or denudation 426 BOTANICAL GAZETTE [may could account for this encroachment, and I beheve the latter more probable. The erosive force might well have been wind, acting subsequent to the removal of the white pine forest which evidently existed here. High tidal action has undoubtedly cooperated in the removal of the upper part of what appears to have been an extensive sand plain, as indicated by the north and south extensions of the lagoons. Restoration of original condition Within a century it seems probable that an area somewhat more extensive to the south than that now occupied by the salt marsh, and lying in the lee of the highlands to the north, was covered with a stand of Pinus Strobus. The present area of the dune complex, lacking this protection from marine influences, was covered with a stand of Picea canadensis and Abies, and this stand extended in its full width to the present end of the spit. It seems probable that the dune complex is a relatively recent phenomenon, developing subsequent to the removal of the stand of white pine. At about this same time the channel of the tidal current was changed and began cutting at the south end of the spit, eroding the Picea stand (fig. 5). According to a native, about one-third of this erosion has been accomphshed in the last 35 years, or at the rate of 25 ft. per annum. If this rate has been approximately constant, the Picea stand was intact within a century to the end of the spit, which now lies bare for about one-half mile (fig. 3). Summary It is the purpose of this paper to put on record several facts of ecological interest: (i) a coniferous sand dune with Picea cana- densis as its facies located at the latitude of 47° north; (2) Poa compressa as a sand binder; (3) abundant layering in Picea canadensis and Abies balsamea; (4) the anomalous condition of a sand dune moving seaward; (5) a phenomenal development of Arceuthobiiim pusillum on Picea canadensis; (6) the decisive value of ecological data in the interpretation of physiographic phenomena. Western State Normal School Kalamazoo, Mich. EMBRYO SAC AND EMBRYO OF PENTSTEMON SECUNDIFLORUS CONTRIBUTIONS FROM THE HLILL BOTANICAL LABORATORY 248 Arthur T. Evans (with PLATE XIl) Two genera of the Scrophulariaceae were the first plants in which the development of the embryo sac and the embryo were correctly investigated. In 1851 Hofmeister (8), working on Lathraea squamaria and Pedicularis sylvalica, proved that the embryo was formed as a result of the fertilization of the egg, and not from the end of the pollen tube as was believed by Schleiden and his followers. Deecke (5) in 1855 reinvestigated Pedicula- ris sylvatica. He insisted that Hofmeister was wrong and that the embryo really did develop from the end of the pollen tube. He was supported in his assertions by Schacht (ii). Later, how- ever, Hofmeister (9) proved that what Deecke really saw was the proembryo. In his paper on Lathraea and Pedicularis, Hof- meister discusses the beginning of the endosperm and the haustoria. No further work of importance was done upon the Scrophulariaceae until 1874, when Chatin (3) studied the devel- opment of the ovule and the seeds in a number of genera. Four years later Vesque (14) worked on the embryo sac of a number of families, among which were included several of the Scrophulariaceae. Even as late as his publication of this paper Vesque believed that Schleiden's theory of the formation of the embryo was correct, and criticized Hofmeister's interpretation as inaccurate. One of the best contributions to our knowledge of the embryo sac situation in this family is by Balicka-Iwanowska (2) in 1899. The account includes a study of a number of families of the Sym- petalae, but deals especially with several genera of the Scrophu- lariaceae, particularly taking up the question of nutrition in the embryo sac. The haustoria are believed to have an absorptive power, and thus conduct nourishment into the embryo sac, the 427 428 BOTANICAL GAZETTE [may conclusions being based on the fact that the haustoria are always found in contact with parts which are well supplied with nourish- ment. Miss Balicka-Iwanowska disagrees with Hegelmaier (7) as to the function of the tapetum. Hegelmaier beheved it to act as a protective covering, while the former seemed to prove that it serves to pass nutritive substances on to the embryo sac, and that it possibly has a digestive function, since the cells of the integu- ment adjoining it are found constantly breaking down. In 1906 ScHMiD (12) investigated numerous species of the Scrophulariaceae. He discusses the formation of the embryo sac, fertilization, endo- sperm formation, and the development of the haustoria. He has done very little with the development of the embryo. In 191 5 Miss Mitchell (id) investigated the embryo sac and the embryo in Striga lutea, a semi-parasitic plant found in South Africa. In this form she has noted the lack of a tapetal layer. The material for this study was collected near Boulder, Colo- rado, where the species is abundant. The indeterminate inflores- cence affords flowers in all stages of development on the same plant. Such material was killed in chrom-acetic acid, cut in paraffin, and stained in various stains, safranin-gentian violet proving the best. Longitudinal sections 10 m thick proved quite satisfactory for study. The writer is indebted to Dr. Francis Ramaley of the Uni- versity of Colorado for advice during the early stages of the work, and especially to Dr. Charles J. Chamberlain of this laboratory, under whose direction the work was completed, for his kindly aid and criticism. Ovary and embryo sac The ovary of Pentstemon secundifloriis Benth. is of the ordinary bilocular scrophulariaceous type, with the partition somewhat swollen in the median Hne forming the placenta, which bears the numerous crowded anatropous ovules. Longitudinal sections of such an ovary at right angles to the partition afford a large number of ovules in each section for study. Sections of very young ovaries show the ovules beginning as slight swellings of the placenta. The megaspore mother cell is not distinguishable in such an early stage, becoming apparent only after the beginning of integument forma- iQigl EVANS— PEN TSTEMON 429 tion. It appears as a single enlarged and darker staining sub- epidermal cell, which functions directly. Growth is quite rapid and the cell soon becomes elongated. It is surrounded by a single layered nucellus. The single integument forms along the sides of the nucellus and soon surrounds it. A short time before the integument has completely surrounded the ovule the megaspore mother-cell has entered synapsis (fig. i). Miss Mitchell has estimated that about 10 per cent of the ovules of Striga lutea have reached synapsis at the same time. This percentage may safely be placed much higher for P. secundifforus, probably more than 75 per cent of the ovules of a single ovary showing the same stage of development. By the time the nucleus of the megaspore mother-cell has entered synapsis the young ovule is rapidly assuming its anatropous form, which is reached by the time the reduction divisions are completed. The first reduction division occurs about the time the integument has surrounded the ovule completely. This division is soon fol- lowed by the second, forming the row of 4 megaspores. Either the third or the fourth megaspore of the row may function in forming the embryo sac (figs. 2, 3). The other 3 disintegrate rapidly and become crushed by the growth of the one functioning. The megaspore which functions increases rapidly in size, the micropylar end becoming bulbous while the chalazal end remains narrowed (fig. 4). The chalazal end, however, lengthens rapidly until it is 2-4 times as long as the bulbous portion. This growth carries it to a point in contact with the end of the vascular system. The nucellus early disappears, but by the time the embryo sac is formed another nutritive layer, the tapetum, has formed from the integument. During the growth of the embryo sac the single nucleus by 3 divisions has formed the 8-nucleate sac. The rapid growth of the sac causes the protoplasm to be much vacuolated. In the earliest stage of the 8-nucleate sac 4 nuclei are found grouped at each end (fig. 4). Soon, however, a nucleus from each end migrates toward the opposite end. Eventually they meet and form the polar fusion nucleus (fig. 5). Balicka-Iwanowska and ScHMiD have commented upon the place of this fusion. The former says that it occurs near the middle of the sac, while the 43° BOTANICAL GAZETTE , [may latter finds that it may occur anywhere in the sac of the Scrophu- lariaceae studied by him. In P. secundiflorus polar fusion was found to take place anywhere, seeming to be more a matter of chance than any regulated procedure. Regardless of where polar fusion takes place, the polar fusion nucleus is always found in the bulbous micropylar end of the sac at the time of fertiliza- tion (figs. 6, 7). It is here that the triple fusion is completed. By the time polar fusion is completed the egg apparatus is well formed and the antipodals have begun to disintegrate. In only one case were the antipodals observed to form anything resembling cell walls. The mature embryo sac (fig. 5) is one of the commonest stages of the sac to be found. This is probably due to failure to pollinate at once. A short period of inactivity always seems to occur. The mature embryo sac is interesting in that it is always well filled with starch (figs. 6, 7). As soon as the megaspore begins its development into the embryo sac, traces of starch are to be found in it, although it is not until the embryo sac is well matured that large quantities of starch are present. Very often the adjacent tissues contain much starch also. After fertilization, when the endosperm and the embryo begin to develop, the starch in the sac disappears entirely. Many of the grains found in the sac are large, reaching 30 ju or more. Although starch is to be found in either end of the embryo sac it is always much more abundant in the micropylar end. Schmid has found this to be true also. False polyembryony The fusion of 2 ovules appears to be a much more uncommon occurrence than the formation of 2 or more embryo sacs in a single ovule. Miss Mitchell discusses a single case which she found in Striga lutea. The only other plants in which it has been reported are Pyrus Malus, Loranthus europaeus, and Viscum aibum (4). In the course of this study several cases in which 2 ovules had fused were noted. In some the fusion was quite complete, in others the ovule could be seen to be double. The presence of 2 micropyles as well as integumentary tissue between the 2 ^mbryo sacs indicated that 2 ovules had fused. In one instance noted the igig] EVANS— PENTSTEMON 43 1 egg apparatus had formed and polar fusion had occurred in both embryo sacs. False polyembryony seems to be quite common in this species. Fertilization In several cases the pollen tube with the tube nucleus and the 2 sperms were observed. While in the pollen tube the sperms are more or less capsule-shaped, but after reaching the embryo sac they become quite spherical. The pollen tube seems always to enter the embryo sac a little to one side, its entrance usually destroying one of the synergids, the other synergid disappearing soon after- ward. The sperms are readily distinguished from the egg nucleus and the polar fusion nucleus on account of their much smaller size. Fertilization of the egg and the triple fusion always occur in the micropylar end of the sac and in a normal manner. Both fusions occur at approximately the same time. Several cases of double fertilization were observed (fig. 7). Previously double fertilization has been announced as occurring in Digitalis purpurea, Linaria vulgaris, Melampyrum sylvaticuniy Lathraea squamaria, Pedicularis foliosa, and Striga lutea of the Scrophulariaceae. This adds Pentstemon secimdiflorus to the list. Formation of endosperm Without resting after the fusion with the sperm, the endosperm nucleus by a series of divisions forms a large number of nuclei, which migrate to the chalazal end of the sac and there become peripherally placed. Simultaneous with the formation of the free endosperm nuclei the narrowed end of the sac begins to increase in size very rapidly, so that it soon surpasses the micropylar end in diameter (fig. 8). By the time the first endosperm walls have formed this end of the sac is much the larger. During all this increase in size a certain restricted area between the 2 ends remains very narrow, so that the embryo sac comes to be dumb-bell-shaped, with the chalazal end the larger. Endosperm walls continue to form in this end until the whole is completely filled (figs. 8, 10). Although endosperm nuclei are occasionally found in the micro- pylar end of the sac, no cell walls were observed to form. During 432 BOTANICAL GAZETTE [may endospemi formation the tapetum appears to be very active. Integumentary cells in contact with it are broken down and the tapetal cells are always filled with a dense protoplasm. Haustoria With the formation of the endosperm 2 large haustoria are formed: one in the neck which connects the micropylar and chalazal ends of the sac (fig. 8), the other as an outgrowth from the chalazal end of the sac (fig. 9). The former is formed by the growth of 2 endosperm cells forward through the narrowed neck and just into the micropylar end of the sac where growth stops. In the case of the chalazal haustorium there is an outgrowth of the sac in the region not covered by the tapetum. Into this bulbous pocket 4 endosperm cells grow. This brings the endosperm cells well into connection with the vascular tissue, the cells of which are gorged with nutritive material. The protoplasm of each haus- torium is very dense. The cells of the chalazal haustorium are binucleate. The active tapetal layer covers only the chalazal end of the sac (fig. 8), ending abruptly at its junction with the micropylar end. ScHMiD found that the tapetum might cover all of the embryo sac or only part of it as in P. secundiflorus. The latter condition seems the more common occurrence. Miss Mitchell found that no tapetum is formed in Striga lutea. She believes that this may be accounted for by the semi-parasitic habit of the plant. Development of embryo After fertilization the egg rests for a time, often even until endosperm cell walls have begun to form. It then divides, the first division being at right angles to the axis of the embryo sac. The segment nearest to the micropyle forms the suspensor, the other forming the embryo. By a series of divisions, coupled with rapid growth, the suspensor is transformed into a bulbous basal portion, and a number of smaller narrowed cells which lengthen rapidly in such a manner as to push the i -celled embryo through the micro- pylar end of the sac (fig. 8) and into the center of the endosperm beyond (fig. 10). It is usually pushed from one-third to one-half iqiq] EVANS—PENTSTEMON ' 433 of the way through the endosperm, where further progress is prob- ably stopped by the division and growth of the embryo. The first 2 di\isions of the embryo are at right angles to each other and in the plane of the long axis of the sac. The next division is at right angles to the first two and forms the 8-celled stage of the embryo. The i6-celled stage is formed by the pericHnal division of the cells of the octant. The further division of the embryo was not followed. After the embryo becomes imbedded in the endosperm the micropylar end of the sac, together with the suspensor, collapse and disappear. Their disappearance is accounted for by the pres- sure within the ovule, due to the increase in amount of endosperm which eventually comes to occupy all the space inside the seed coat. Discussion In the formation of the embryo sac of P. secundijiorus there is nothing strikingly different from that of other species of this family which have been studied, but the shape of the mature embryo sac is peculiar. The very bulbous micropylar end, with the long, narrowed chalazal end, gives the whole embryo sac a club-shaped appearance. The chalazal part of the embryo sac is never more than half as wide as the micropylar end at the time of fertilization. The drawings of the embryo sac of other Scrophulariaceae by Balicka-Iwanowska, Schmid, and Mitchell show that there is a tendency toward this shape of embryo sac in the family, but none of those drawn are so striking in shape as that of the species under consideration. The distance between the end of the embryo sac and the end of the vascular system is at first marked. As the sac later derives a large part of its nourishment through the vascular system, this may account for the necessity of lengthening the sac until the end comes in contact with this source of food supply. During the development of the embryo sac traces of starch can be seen within it, and in all cases, by the time fertilization occurs, large quantities of starch are present. Often it is so abundant that the nuclei within the sac are partially or entirely obscured. By the time the embryo has reached the endosperm the starch has all dis- appeared. 434 BOTANICAL GAZETTE [may D 'Hubert (6) has made a study of fleshy plants with regard to the formation of starch in the embryo sac. He finds that starch is always present in the sacs of fleshy plants such as the Cactaceae, Mesembrianthaceae, Crassulaceae, Portulacaceae, etc. He has also found, however, that some non-fleshy plants show starch in the embryo sac. According to D 'Hubert this latter case seems to be the exception rather than the rule, and he beheves that there is a relationship between the fleshiness of the plant and starch in the embryo sac due to the slowness of the phenomena before fertiliza- tion. This, however, receives very Uttle attention from him; nevertheless it seems the more plausible theory. P. secundiflorus is not a fleshy plant, but, judging from the drawings which D 'Hubert has made of several fleshy plants, it has more starch in its embryo sac than any of those figured. It appears that while there is activity in the embryo sac very little if any starch is stored up. As soon as the embryo sac matures and becomes inactive just before fertilization, possibly due to delay in poUination, the stream of nourishment which has been coming in cannot be checked suddenly but keeps passing more and more nutrition into the inac- tive sac, where it is stored in the form of starch. Such a conclusion seems to be substantiated by the fact that activity in the sac brought about by fertiUzation soon reduces the amount of stored-up starch. Balicka-Iwanowska (2) has also investigated the deposition of starch in "the embryo sacs of several plants, and has concluded that starch is only found in the embryo sac when the tapetum is cutinized. This does not seem to be the case in P. secundiflorus, however, as the tapetum is undoubtedly not cutinized. Moreover, it covers only half of the embryo sac, as has been explained before. ScHMiD (12) has found starch present in the integuments as well as the embryo sacs of a number of the Scrophulariaceae. He states that the starch is found throughout the embryo sac, but that sooner or later it is all translocated to the micropylar end, "wo die lebhaftesten Teilungen stattfinden." The function of the tapetum seems to be one of nutrition, as has been suggested by Balicka-Iwanowska (2). That it may have a protective function, as has been suggested by Hegelmaier (7), seems rather doubtful. This seems all the more questionable 1919I EVANS— PENTSTEMON 435 when one considers that it covers only the chalazal end of the sac in a number of species. Surely the micropylar end would be as much in need of protection. In P. secundiflonis the integumentary cells border on the micropylar end of the sac. The two haustoria function in passing nourishment to the endo- sperm cells which are farther from the supply of food. By the time the embryo has reached the endosperm the micropylar haus- torium becomes inactive and is lost. The chalazal one, however, functions until the endosperm is formed. The nuclei in this haustorium are very pronounced. On account of the large size and seeming activity of haustorial nuclei some authors have attributed to them si considerable role in nutrition. Balicka- IwANOWSKA (2) has always found them near the point where nutrition is most abundant. In this work a similar tendency was noted. The growth of the suspensor in such a manner as to push the proembryo through the micropylar end of the embryo sac and to imbed it in the endosperm is rather unique. Sharp (13) in his study on Physostegia has recorded a similar situation, but the method of endosperm formation in Physostegia and Pentstemon is different entirely. During the growth of the suspensor which imbeds the embryo in the endosperm, nutrition is derived from the starch stored up in the micropylar end of the sac. Summary 1. The embryo sac is developed from a single megaspore. Its antipodals disorganize early. The micropylar end becomes bul- bous, while the chalazal end becomes long and narrow and is covered by a distinct tapetum. 2. The mature embryo sac is found to be constantly gorged with starch, due to the non-utilization of the nutritive materials which pass into the sac at a time of inactivity just before^ fertilization. 3. The endosperm nucleus immediately divides and free nuclei migrate into the chalazal end of the sac, where wall formation begins. The proembryo is pushed into this endosperm by an 436 BOTANICAL GAZETTE [may extreme growth of the suspensor. The micropylar end of the sac disintegrates. 4. Two haustoria are formed, the micropylar by the growth of endosperm cells from the chalazal end into the micropylar end, and the chalazal by a growth of endosperm cells from the chalazal end out into the vascular system. The cells of the latter haustorium are binucleate. 5. False polyembryony occurs rather commonly in this species. Purdue Agricultural Experiment Station Lafayette, Ind. LITERATURE CITED 1. Bachmann, E. Th., Darstellung der Entwicklungsgeschichte und der Baues der Samenschale der Scrophulariaceen. Nova ActaLeop. 43:1-177. ph. 1-4. 1888. 2. BalickaTwanowska, G., Contribution a I'etude du sac embryonnaire chez certaines Gamopetales. Flora 86:47-71. pis. 3-8. 1899. 3. Chatin, J., Etudes sur la developpement de I'ovule et de la graine. Ann. Sci. Nat. Bot. V. 19:1-107. pis. 1-8. 1874 . 4. Coulter, J. M., and Chamberlain, C. J., Morphology of angiosperms. New York. 1903. 5. Deecke, Th., Nouvelles recherches sur la developpement du Pedicularis sylvatica. Ann. Sci. Nat. Bot. IV. 4:48-63. 1855. 6. D'HuBERT, M. E., Recherches sur le sac embryonnaire des plantes grasses. Ann. Sci. Nat. Bot. VIII. 2:37-128. pis. 1,2. 1896. 7. Hegelmaier, Fr., tJber den Keimsack einiger Compositen dessen Umhul- lung. Bot. Zeit. 47:805-812, 821-826, 837-842. 1889. 8. HoFMEiSTER, W., Zur Entwicklungsgeschichte des Embryo der Personaten. Neue Reihe. 9:449-457. pis. 8, g. 1851. 9. , Neuere Beobachtungen iiber die Embryobildung der Phanero- gamen. Pringsh. Jahrb. 1:1858. •10. Mitchell, Margaret, Embryo sac and embryo of Striga lutea. Bot. Gaz. 59:124-135. pls.8,g. 1915. 11. Schacht, H., Sur I'origine de I'embryon vegetale. Ann. Sci. Nat. Bot. IV. 3:180-208. pis. 7,8. 1855. 12. ScHMiD, E., Beitrage zur Entwicklungsgeschichte der Scrophulariaceen. Zurich. 1906. 13. Sharp, L. W., The embryo sac of Physostegia. Bot. Gaz. 52:218-223. pis. 6, 7. 1911. 14. Vesque, M. J., Developpment de sac embryonnaire. Ann. Sci. Nat. Bot. VI. 8:261-390. pis. 12-21. 1879. BOTANICAL GAZETTE, LXVII PLATE XII EVANS on PENTSTEMON 1919] EVANS— PENTSTEMON 437 EXPLANATION OF PLATE XII All the drawings were made at table level with the aid of an Abbe camera lucida. The labels are: a, sperm; b, polar fusion nucleus; c, megaspore; d, micropylar end of embryo sac; e, embryo; h, haustorium; s, starch; /, tapetum. Fig. I. — Synapsis in the megaspore mother-cell; X880. Fig. 2. — Row of 4 megaspores showing fourth developing into embryo sac; X880. Fig. 3. — Row- of 4 megaspores showing third developing into embryo sac; X880. Fig. 4. — An 8-nucleate embryo sac before maturity, showing character- istic shape of sac and egg apparatus beginning to form; X880. Fig. 5. — ^Micropylar end of mature embr>'o sac showing egg apparatus well formed and migration of polar fusion nuclei; X460. Fig. 6. — IMicropylar end of mature embryo sac showing egg apparatus, entrance of pollen tube, 2 sperms, polar fusion, and starch; X880. Fig. 7. — Mature embr>'o sac showing double fertilization; X880. Fig. 8. — Stage showing growth of suspensor pushing young embryo between micropylar haustoria w-hich have already begun to disintegrate; section a little to one side of chalazal haustorium; tapetum present; X460. Fig. 9. — Stage showing binucleate chalazal haustorium; X880. Fig. 10. — Late stage of endosperm formation showing embryo imbedded; micropylar end of sac already beginning to collapse; X460. BRIEFER ARTICLES DEPRESSED SEGMENTS OF OAK STEMS (with four figures) In a recent paper, Miss Langdon' questions certain statements of the writer' in regard to the deeply depressed or sunken segments which occur commonly in stems of Quercus. She states (p. 321): From observations of transverse sections of twigs from Quercus alba, Q. bicolor, and Q. macrocarpa I find that there is evidence of retardation in growth of the tissues in the immediate vicinity of the wide rays, especially noticeable in the marked dipping in of the annual rings where they cross the large rays. However, aside from a few extreme cases, this checking influence of the wide foliar rays does not explam the 5 conspicuous depressions so char- acteristic of the wood of Quercus. In discussing the topographical features of the stem of the oak, it is essential to distinguish between two different factors which have modi- fying effects upon the outline of the secondary xylem. I refer to the arrangement of the primary elements and the development of multi- seriate rays. The effects of these two factors, or complexes of factors, may be studied most satisfactorily in plants where they occur independ- ently. For example, in Castanea dentata and Populus halsamifera, which normally have only uniseriate rays, the primary elements have the stellate arrangement that is characteristic of Quercus. In the inter- nodes of normal stems of these plants, the first formed secondary ele- ments form a layer of undulating or stellate outline, but at the end of two or three growing seasons, frequently earlier, the outer periphery of the secondary xylem tends to be circular (fig. i). In other w^ords, the early lobed or stellate form of the cambium soon becomes evanescent, and its effects upon the shape of the stele are quite transient. The modifying influences of the second set of factors, acting inde- pendently of the first, may be seen quite clearly in the stems of certain 'Langdon, LaDema M., The ray system of Quercus alba. Bot. Gaz. 65: Z'^Z- 323. 1918. 2 Bailey, I. W., The relation of the leaf trace to the formation of compound raj-s in the lower dicotyledons. Ann. Botany 25:225-241. 1911. 438 iqiq] BRIEFER ARTICLES 439 species of Amphilophiiim. Fig. 2 illustrates a cross-section of the stem of one of these plants. There are 4 pairs of approximated multiseriate rays in the fourth growth layer. The narrow segments of xylem be- tween these wide rays are deeply depressed below the general outline of the stem. They obviously are not correlated with a lobed or stellate arrangement of the primary elements, but are due to differences in the number of xylem elements formed by different arcs of the cambium during the fourth growing season. That the deeply depressed segments which occur commonly in oak stems, having 2-10 or more growth layers, are correlated with the presence of pairs of approximated multiseriate rays rather than the Fig. I Fig. 2 stellate arrangement of the primary elements, is indicated not only by a comparative study of the stems of various arborescent dicotyledons, but also by numerous facts in the anatomy of the genus Querciis. From different species of oaks and from plants grown under different environ- mental or experimental conditions, it is possible to secure a series of stems showing various stages in the disintegration and disappearance of multiseriate rays.'' The segments are most deeply depressed in specimens in which the pairs of multiseriate rays are most conspicuously developed (fig. 3). On the other hand, where the pairs of wide rays or their vestiges ("aggregate rays, " etc.) are entirely absent, the stellate form of the early cambium, which may be conspicuous during the first growing season or two, quickly becomes circular, as in Castanea and Populus. Where there 3 Bailey, I. \V., and Sinnott, E. W., Anatomical evidences of reduction in certain of the Amentiferae. Box. Gaz. 58:36-60. 1914. 440 BOTANICAL GAZETTE [may is a marked retardation in the development of the pairs of rays the ap- pearance of the deeply depressed segments is coincident with that of the rays (fig. 4). Particularly significant are those stems in which one ray of a pair fails to develop. Under these circumstances, the nar- row segment of xylem tends to be unsymmetrically depressed (fig. 4). Furthermore, the fact that depressed segments may occur between pairs of rays, which are opposite the projecting lobes of the pith (fig. 3), and between approximated " secondary " rays, suggests that the stellate outhne of the early cambium is not an indispensible factor in the pro- duction of the sunken wedges of xylem in oak stems. Fig. 3 Fig. Miss Langdon offers a physiological explanation for the stellate form of the stele in young twigs (p. 3 2 1 ) : Since the principal function of the xylem is the conduction of water from the soil to the outer parts of the plant, it is obvious that the maximum upward movement of solutions in the stem would be through the tracheidal tissues and vessels in direct line with the leaf traces. This would cause an acceleration in growth and the consequent outward projection of those five regions of the woody cylinder associated with leaf traces, while the neighboring conducting tissues, namely, the so-called depressions from which the main conducting streams had been diverted to the petioles of the leaves, would fail to maintain their normal rate of growth. It is to be emphasized, in this connection, that the projecting wedges of the first annual rings of Castanea, Populus, and Quercus, when devoid of wide rays, are not due to an acceleration of growth. This can readily be determined by measuring the depth of the convex and concave arcs of 1919] BRIEFER ARTICLES 44 1 xylem in one or two year old stems. The average depth of the latter almost always equals and usually exceeds that of the former (fig. i), indicating conclusively that there is no growth acceleration in the convex arcs of the cambium which form the projecting wedges. As has been indicated by the writer, the undulating outline of the first formed secondary xylem is due to the stellate arrangement of the primary elements, and consequently the stellate outhne of the first formed cam- bium. However, this originally lobed cambium rapidly takes on a cir- cular outline, owing to the slower growth of its convex projecting arcs, except in stems which have a hereditary tendency for the formation of pairs of approximated multiseriate rays. — I. W. Bailey, Bussey Institu- tion, Jamaica Plain, Mass. IMPORTANCE OF EPIDER^LAL COVERINGS' (with two figures) In making tests of the relative resistance of some herbaceous plants to freezing, it was observed that inoculation from ice formed on the leaf surface was a factor of great importance in determining the temperature at which ice formation occurred in the leaf tissue. In testing cabbages it was observed that the greatest undercooling of the tissue below its freezing point occurred in those plants which had the greatest amount of " bloom " on the leaf surface. Plants well covered by wax could be main- tained for hours at a temperature 5°C. below their freezing point without the formation of ice in the tissues. Similar conditions were found to occur in the common Cineraria and other such plants which are densely covered with a mat of epidermal hairs. This condition suggested that inoculation of the undercooled leaf tissue by ice formed on the leaf surface was an important factor in frost resistance. The object of this study was to determine the amount of undercooling which can occur in such tissues, and the importance of the epidermal coverings in preventing surface inoculation of the undercooled tissues. The thermoelectric method was used to measure temperatures, since this method allows one to determine the temperature inside rather thin leaves. A copper-constantan couple of no. 40 B. and S. gauge which had a thermal coefficient of 3 .33 milhvolts per degree Centigrade was used. Using such a couple the delicacy of the potentiometer arrange- ment determines the accuracy of the temperature measurement. Al- though with the arrangement used much smaller changes could be ' Published by permission of the Secretary of Agriculture. 442 BOTANICAL GAZETTE MAY determined, measurements to o . i°C. or less were found to be sufficiently accurate for this work. Fig. I shows the arrangement for temperature measurement as well as the means for securing undercooling of the tissues. Directions for constructing the thermal junctions are to be found in the pubUcations of White from the Carnegie Insti- tution of Washington.^ The con- stant temperature junction of the thermocouple was placed in a thin- walled tube filled with oil and immersed in a slush of clean snow in distilled water contained in a Dewar beaker. It was found that the temperature of this junction could be kept constant for a long time within a few thousandths of a degree Centigrade, as shown by a standard Beckmann thermometer. In calibrating the couple, the junction to be placed within the leaf was placed in a Dewar beaker similar to the former arrangement, with solutions having freezing points a few degrees below zero. A number of tissues were tested with this apparatus, including car- nation stem, cabbage \ea.i,Echeveria leaves. Cineraria petiole, tomato petiole, and others. In general the results show that plants with heavy epidermal coverings of wax or Fig. I trichomes can undergo a much greater undercooling than such plants as tomatoes, in which the epidermal walls are thin and non- resistant. * White, W. P., The thermoelement as a precision thermometer. Physical Review 31:135. 1910. , The constancy of thermoelements. Physical Review 23:449. 1906. White, W. P., Dickinson, H. C, and Mueller, E. I. The calibration of copper-constantan thermoelements. Physical Review 31 : 159. 1910. I9I9] BRIEFER ARTICLES 443 To make certain that all of the samples should have ice formed on the surface, a small drop of water was placed on each. To test the effect of the epidermal covering in preventing inoculation, samples of the same material were cooled with the coverings intact, and also after they had been removed. Precautions were taken to prevent the inoculation of the tissue from juices exuded at cut surfaces. In all cases the cut sur- faces were dried and covered with vaseline, which procedure was found to prevent such an inoculation of the tissue. In all the tests it was 0 2 £ h 6 8 1 TTllnutes 0 12 14 IG 1 8 20 22 24 ^ t\ ■^ — •• — "^x. \\ -1 -2 ^N ^---, \j \ V \ \ '-/■■: 1 ^ \ \ I » •. 1 i \ h -7 -8 \ / \ 1 i \ \ \ f \ \ ^.^ 1 \ -E :hev( iXii.. Wd; [Terr oved 1 \ -E :hev erU Wa> not nemo ^j \ inera inera ruL. 1 ria. 1 Idlrs IdLrs rem not T wed. •£am \ 1 \ Fig. 2 found that when thick epidermal coverings were unbroken, undercooling occurred, while in those cases where they were removed but little undercooling occurred. Fig. 2 shows typical cooling curves, indicating the importance of the epidermal coverings in preventing ice formation within the tissue. It appears that the tissue without the protective covering is unable to undergo any great amount of undercooling. This depends in part upon the amount of shaking of the tissue. The osmotic concentration or the presence of colloidal substances in solution in the cell sap are of relatively minor importance in determining the undercooling. 444 BOTANICAL GAZETTE [may It has been observed during spring frosts that in blooming apricots those buds which have opened and are turned upward to collect snow are frozen, while those turned downward may not be injured. West and Edlefsen^ attempted to prevent frost injury to apricot buds by spraying the trees with water. Instead of the desired effect of preventing freezing, this procedure evidently killed the tissue by allowing inoculation from ice formed on the surface. Trees which were not sprayed with water were not injured, although subjected to the same temperature. These examples serve to illustrate the importance of surface inoculation in producing frost injury. The amount of undercooling in plants is not generally very great, nor is it sufficient to account for true frost hardiness. Such herbaceous plants as cabbage, kale, turnips, however, which ordinarily can with- stand a considerable degree of freezing, acquire hardiness quite rapidly. In the case of cabbage, a temperature of 3°C. was found to harden the plants sufficiently in 5 days to allow them to be frozen stiff at — 3°C. without injury. The principal importance, therefore, of the epidermal coverings for the frost resistance of such plants appears to be that they allow the plants which possess them to withstand temperatures somewhat below zero until the cells are able to adapt themselves physiologically to the changes incident upon freezing. Summary Undercooling of the tissues occurs to a greater degree in such her- baceous plants as possess protective epidermal coverings than in plants not so protected. The undercooHng in such plants is not due to sub- stances in the cell sap, but mainly to the prevention of inoculation from ice formed on the surface of the tissue. A method is given for determin- ing electrically the temperatures within leaf tissues. — R. B. Harvey, Bureau of Plant Industry, Department of Agriculture, Washington, D.C. 3 West, F. S., and Edlefsen, N. E., Orchard heating. Bull. 161, Utah Agric. Coll. Exp. Sta. 1917. CURRENT LITERATURE NOTES FOR STUDENTS Conditions affecting flower development. — Klebs' divides the process of flower formation by the rosettes of Sempcrvivum Funkii and 5. albidum into 3 distinct, successive steps: (i) production of the condition of ripeness to flower (bliihreife Zustand), (2) formation of flower primordia, and (3) develop- ment of flower clusters and elongation of the axis. Light is the dominant factor in determining all 3 of these stages of development. In the first and third, light is effective entirely through its photosynthetic action, and its effectiveness rises with its energy value. Higher temperatures counteract light by favoring dissimilation. Accordingly, the effect of tempera- tures can in part be annulled by increased light intensities. It is the balance of assimilation over dissimilation that furthers the development of these 2 stages. Klebs finds that at lower temperatures (about 6°C.) both these stages can be attained in darkness, although in the last it gives a far less exten- sive inflorescence. He thinks this is likewise tied up with a balance in favor of available carbon synthate. The low temperature gives low respiration and leads to the accimnulation of soluble sugars by the hydrolysis of insoluble carbohydrates. In the second step, formation of flower primordia, light has 2 distinct and antagonistic effects. The one which favors the process is due to the photo- synthetic activity of the light and is a function of the less refrangible rays of the spectrum. The other, which inhibits the process or even annuls the ripe to flower condition, must at present be termed a stimulus effect, and it is a function of the less refrangible blue rays. Diffuse daylight is relatively injuri- ous to primordia development because of the high percentage of blue violet rays it contains. The Osram light and direct sunhght favor this development because of the dominance of the red rays. Klebs says it is still an unanswered question whether inflorescence develop- ment in other forms and in plants in general can be divided into these 3 distinct steps with similar light effects in each step. He suggests some facts as evidence that such may be the case. His past work has done much to show that the formative effects of conditions on plants is largely through the nutrient effects of these conditions. Thus the formative effect of light is explained in a large part by its effect on carbon assimilation, but Klebs points out here, as in his ' Klebs, George, tjber die Bliitenbildung von Sempervivum. Festschrift zum Ernst Stahl. pp. 128-151. Jena. 1918. 445 446 BOTANICAL GAZETTE [may older work, that there is also a specific formative action of the blue rays as yet unexplainable on the nutrient basis. He has often distinguished between the amount of carbon synthate and the amount of salt nutrients as formative factors in the plant, especially in con- nection with reproduction; and now Fischer^ makes this more definite by considering the nitrogen supply as the most important formative factor furnished by the salts, and by speaking of the carbon nitrogen ratio (C/N) of plants. He probably would not deny that the supply of other nutrient ele- ments, phosphorus, calcium, potassium, etc., have at least minor formative effects and often of an opposite nature from nitrogen. This ratio can be increased by increasing the photosynthesis of the plants or by decreasing the nitrogen supply. The ratio can be decreased by decreasing photosynthesis or by increasing the nitrogen supply. Fischer comes to this important conclu- sion. Very high C/N in plants favors flowering, while a low C/N favors veg- etation. His conclusions are largely based on his own work on the effect of increased partial pressure of carbon dioxide upon the development of plants, but not upon chemical analysis of the tissues. Kraus and Kraybill^ have recently worked upon the tomato, varying the C/N in it by varying its nitrogen supply. On the basis of extensive cultures and chemical, microchemical, and anatomical studies, they come to the follow- ing conclusions: (i) a very high C/N gives little vegetative growth and poor reproduction with a high percentage of dry matter; (2) medium C/N gives moderate vegetation growth, good reproduction, and a medium percentage of dry matter; (3) very low C/N gives ver>^ vigorous vegetative growth, little reproduction, and a low percentage of dry matter. Kraus's extensive horticultural investigations enable him to give much evidence that the C/N ratio is a factor of great significance in determining fruitfulness in many economic plants. The contribution apparently puts into the hands of pro- ducers one of the important means of controlling fruitfulness. Fischer's less extensive and one-sided attack caused him to miss the fact that a very high C/N not only reduces vegetative growth but diminishes reproduction. These papers have thrown much light on some of the nutrient factors modifying vegetation and reproduction in plants. — Wm. Crocker. Loss of chlorophyll. — Meyer'' notes that in Tropaeoliim majus, growing in pots in a greenhouse, the young leaves at the top of the stem are dark green, while the progressively older ones down the stem are green, bright green, yellow 2 Fischer, H., Zur Frage der Kohlensaure-Einahrung der Pflanzen. Garten- flora 65:232-237. 1916. 3 Kraus, E. J., and Kraybill H. R., Vegetation and reproduction with special reference to the tomato. Oreg. Agric. Exper. Sta. Bull. 149. pp. 90. 1918. -I Mey-er, Arthur, Eiweiszstoffwechsel und Vergilben der Laubblatter von Tropaeoliim majus. Festschrift zum Ernst Stahx. pp. 85-127. Jena. 1918. iqiq] CURREXT literature 447 green, yellow and bright yellow, and finally the oldest ones on the plant are wilting. Meyer points out that this change in color is due to the gradual decomposition of the chlorophylls, while the carotin and xanthophyll remain constant. As this change progresses the chloroplasts become smaller, and in later stages are shriveled granular masses with balls of excreted material about them. With the gradual loss of chlorophyll goes a similar decomposition of the proteins of the chloroplast. It should be mentioned that Meyer adduces evidence for the view that the chloroplast is the main organ for the storage of the proteins manufactured in the foliage leaf, if indeed not the very seat of protein manufacture. The amount of carbohydrates in the leaves also falls with age. Meyer found that when leaves are placed in darkness no reduction occurs in the proteins until the carbohydrates are greatly reduced by respira- tion. The decomposition of the proteins then begins, he believes, as a source of carbon chains for respiration. He states that there is no loss of nitrogen from the leaf during this change, but that the nitrogen residue remains in the leaf, while the carbon chain of the protein is used for respiration. He appar- ently gives the following interpretation of the process: As the leaves become older they become weakened; in this weakened condition the photosynthetic power falls; this leads to a great reduction in the amount of carbohydrates in the leaf, and finally to the decomposition of the proteins of the chloroplasts as a carbon source for respiration; this decomposition of the proteins is accompanied by the decomposition of the chlorophyll and the change in color. Schertz, in an unpublished work from this laboratory, finds in many respects parallel behavior in Coleus Blumei. He finds that shortage of nitrates leads to the decomposition of the chlorophyll, and that old leaves can be maintained green by addition of nitrogen fertilizer. He also finds the phos- pholipine content of the leaf greatly reduced as yellowing progresses. His evidence seems good that shortage of nitrogen initiates all of the decomposition of nitrogen compounds (chlorophyll, phospholipines, and proteins), and that it must be looked at as the immediate cause of the loss of chlorophyll. Plants grown in pots are likely to become pot bound and limited in their supply of soil nutrients. There are many incompletely worked phases in Meyer's paper; he has filled in some gaps by drawing data from other workers on very different materials; and his work leaves much to be desired in quantitative determina- tions and cultural experiments. All these leave interpretation to bridge broad chasms, and it is therefore not strange if he has missed the initiating cause of loss of chlorophyll. If Schertz is right, that the decomposition of chlorophyll in Coleiis Blumei is due to shortage of nitrogen as a building material, it is also conceivable that a great excess of nitrogen may sometimes lead to the decomposition of chloro- phyll due to the dearth of carbon chains produced by the excess of nitrogen. Shortage of magnesium as a building material may sometimes act in a similar way. — Wm. Crocker. 448 BOTANICAL GAZETTE [may Knop's solution. — Toole and Tottinghams find that additions of FeCOH)^ to Knop's solution greatly increases the growth of barley tops in it (21 day cultures), while additions of carbon black depress the growth of tops and additions of HzSiO^ have no effect. None of these additions affect the growth of the roots. Part of the beneficial action of the Fe(0H)3 may be due to its neutralizing action on the acids of the solution. It is interesting to note that the higher additions of Fe(0H)3 removed 90 per cent of the phosphorus from solution. In another piece of work, Tottingham* has shown that he can displace more than 90 per cent of the MgS04 of Knop's solution with Mg(N03)2 without interfering with the growth of red clover, a rather heavy sulphur requiring plant. It is evident that some of the nutrients in the commonly used nutrient solutions are far beyond the minimum concentration necessary to give the plant its optimum supply, and that the so-called optimum concentration of the solution is determined by other factors than optimum supply. The conditions of and the mechanism for absorption from the soil (root hairs with their acid pectic layer in contact with soil particles bearing certain nutrients in compounds of low solubiHty) are quite different. Work with water cultures has established some very fundamental principles in soil fertiUty (essential nutrient elements, necessity of balanced solutions, etc.). It is a question how much more this method alone is capable of adding to our knowledge of soil fertihty. In the present concentrated nutrient solutions with which we are working we may be mainly playing the toxic concentration of one salt against the toxic concentra- tion of another in a way to get the least possible injury. — Wm. Crocker. Chromosomes in Carex. — Oogenesis and spermatogenesis have been studied by Heilborn^ in several species of Carex, special attention being given to chromosome numbers, which vary greatly in this genus. The gametophyte numbers in the forms investigated are as follows: Carex pilulifera 8, C. erice- tortim 16, C. digitata 24, C. caryophylla and C. flava 32. Juel had already reported 52 for C. acuta, and Stout 37 for C. aquatilis. It is interesting to note that C. pilulifera has the largest chromosomes, and that in species with higher numbers the chromosomes are correspondingly smaller. Attempts to cross . the various species have not yet proved successful, but the work is still in progress. — C. J. Chamberlain. 5 TooLE E. H., and Tottingham, W. E., The influence of certain added solids upon the composition and efficiency of Knop's nutrient solution. Amer. Jour. Bot. 5:452- 461. 1918. 6 Tottingham, W. E., Sulfur requirement of red clover plant. Jour. Biol. Chem. 36:429-438. 1918. 1 Heilborn, Otto, Zur Embryologie und Zytologie einiger Carex-P^xten. Svensk Botanisk Tidskrift 12:212-220. _^"g^. 7-J.^. 1918. PLANT GENETICS By JOHN M. COULTER Head of the Department of Botany in the University of Chicago and MERLE M. COULTER Instructor in Plant Genetics in the University of Chicago ^ This book has been written to meet an increasing need among botanical students. Such students in these days, in whatever phase of botany they may be specialising, find it neces- sary to read with understanding much of the literature of plant genetics, because it is becoming increasingly significant in all botanical problems. This means that teachers and investigators must be able to command th.e literature of plant genetics, much of which has been so complex as to be a closed book for the uninitiated. Plant Genetics is an attempt to open this subject to botanical students. fl The book is not intended to be a thorough, authoritative text, but a relatively simple presentation of the more significant investigations on plant genetics which will initiate the student into the subject. Material dealing with some highly specialized phases of genetics and material that is very complex has been purposely omitted for pedagogical reasons. In short, the book is an easy introduction to plant genetics. X+214 pages, i2mo, cloth; $1 .50, postage extra (weight i lb. 2 oz.) THE UNIVERSITY OF CHICAGO PRESS CHICAGO ILLINOIS A SOURCE BOOK OF BIOLOGICAL NATURE-STUDY By ELLIOT R. DOWNING Associate Professor of Natural Science in the School of Education The University of Chicago For teacher-training classes, school libraries, and general reading by every- one who is interested in the common animals and plants. The book is illustrated with more than three hundred sketches and photographs, and the descriptions and keys further faciHtate the identification of plant and animal forms. The author gives suggestive discussions of the hfe-histories and the problems that confront animals and plants, and includes some discussion of their economic value and their conservation. Octavo size, cloth; $3.00, postage extra The Source Book is the second published volume of The University of Chicago Nature-Study Series The first volume is A Field and Laboratory Guide in Biological Pfature-Study Loose leaf; $i .00, postage extra THE UNIVERSITY OF CHICAGO PRESS CHICAGO ILLINOIS FS2 Microscope -Price $40.75 ^auscK'|omb Microscopes are again in the market — backed by the prestige which attaches to successful war service in any significant capacity. Our release from this service is of course gradual. As the demands made upon us by the government are lessened from month to month, however, we are enabled to increase deliveries to our other patrons. Write for complete catalog of our revised microscope line and inform us regarding your requirements. gausch ^ Ipmb Optical ©. 41 2 ST. PAUL STREET ROCHESTER, N.Y. New York Washington Chicago San Francisco Leading American Makers of Photographic Lenses, Microscopes, Projection Apparatus (Balopiicons) , Ophthalmic Lenses and Instruments, Pholomicrographic Apparatus, Range Finders and Gun Sights for Army and Naoy, Searchlight Reflectors, Stereo-Prism Binoculars, Magnifiers and Other High-Grade Optical Products. FINE INKS AND ADHESIVES For those who KNOW Higgins' Drawing Inks Eternal Writing Ink Engrossing ink Taurine Mucilage Photo Mounter Paste Drawing Board Paste Liquid Paste Office Paste Vegetable Olue, Etc. Are the Finest and Best Inks and Adhesives Emancipate yourself from the use of corrosive and iil-smeiline inks and adhesives and adopt the Higgins Inks and Adhesives. They will be a revelation to you, they are so sweet, clean, well put up, and withal so efficieDt. At Dealers Qenerally • CHAS. M. KIGQINS & CO., Mfrs. Branches: Chicago, London 271 Ninth Street Brooklyn, N.Y. Men of Science The signs and symbols used in your everyday work are provided for in special keyboards prepared for the Home Typewriter, the Self-Starting Remington ^ Tell us your needs and we will sub- mit a keyboard layout for your work. No obligation, of course, on your part. Remington Typewriter Company, Inc. 374 Broadway, New York City Volume LXVII Number 6 THE Botanical Gazette Editor: JOHN M. COULTER JUNE 1919 Structure, Development, and Distribution of So-called Rims or Bars of Sanio --------- Irving W. Bailey 449 (With Plates XUI-XV) Apospory in Pteris sulcata L. ------ W. N. Steil 469 (With Plates XVI, XVII, and four figures) Hydrogen Cyanide Fumigation. Contributions from the Hull Botanical Laboratory 249 E. E. Clayton 483 (With two figures) Studies of Some Porto Rican Fungi - - - - Leo R. Tehon 501 (With Plate XVra) Current Literature Book Reviews ----------- 5^2 The living cycads Notes for Students ---------- 513 The University of Chicago Press CHICAGO, ILLINOIS, U.S.A. THE CAMBRIDGE UNIVERSITY PRESS, London and Edinburgh THE MARUZEN-KABUSHIKI-KAISHA. Tokyo. Osaka. Kyoto. Fukuoka, Sendai THE MISSION BOOK COMPANY, Shanghai Volume LXVIl Number 6 The Botanical Gazette A MONTHLY JOURNAL EMBRACING ALL DEPARTMENTS OF BOTANICAL SCIENCE EDITED BY JOHN M. COULTER With the assistance of other members of the botanical stafif of the University of Chicago Issued June 21, 1919 The Botanical Gazette is published monthly by the University of Chicago at the University Press, 5750 Ellis Avenue, Chicago, 111. IfThe subscription price is $7.00 per year; the price of single copies is 75 cents. Orders for service of less than a half-year will be charged at the single-copy rate. ^[Postage is prepaid by the publishers on all orders from the United States, Mexico, Cuba, Porto Rico, Panama Canal Zone, RepubUc of Panama, Hawaiian Islands, Philippine Islands, Guam, Samoan Islands, Shanghai. Tf Postage is charged extra as follows: For Canada, 35 cents on annual subscriptions (total $7.35), on single copies, 3 cents (total 78 cents); for all other countries in the Postal Union, 84 cents on annual subscriptions (total $7.84), on single copies II cents (total 86 cents). If Patrons are requested to make all remittances payable to The University of Chicago Press in postal or express money orders or bank drafts. The following are authorized to quote the prices indicated: For the British Empire: The Cambridge University Press, Fetter Lane, London, E.C. 4. Yearly subscriptions, including postage, £1 12s. 6d. each; single copies, including postage, 35^ 6d. each. . , ,.- For Japan and Korea: The Maruzen-Kabushiki-Kaisha, 11 to 16 Nihonbashi Tori Sanchome, Tokyo, Japan. Yearly subscriptions, including postage. Yen 15.75 each; single copies, including postage. Yen 1.75 each. For China: The Mission Book Company, 18 Peking Road, Shanghai. Yearly sub- scriptions, $7.00; single copies, 75 cents, or their equivalents in Chinese money. Postage extra, if mailed direct outside of Shanghai, on yearly subscriptions 84 cents, on single copies II cents. Claims for missing numbers should be. made within the month following the regular month of publication. The pubUshers expect to supply missing numbers free only when losses have been sustained in transit and when the reserve stock will permit. DeHveries are not guaranteed within the war zones of Europe, nor in other countries wherfe transportation is disturbed by con- ditions incident to war. Business correspondence should be addressed to The University of Chicago Press,j Chicago, lU. Communications for the editor and manuscripts should be addressed to the Editor of the Botanical Gazette, The University of Chicago, Chicago, lU. Contributors are requested to write scientific and proper names with particular care, to use the metric system of weights and measures, and in citations to follow the form shown in the| pages of the Botanical Gazette. Papers in excess of thirty-two printed pages are not accepted unless the author is walling to pay the cost of the additional pages, in which case the number of the pages in the volume is correspondingly increased. Illustrations are furnished without cost to author only when suitable originals are supplied J A copy of the suggestions made in the January number, 1907, will be sent on application. It is advisable to confer with the editors as to illustrations required in any article to be offered. Separates, if desired, must be ordered in advance of pubHcation. A table showing approxi- mate cost of separates is printed on an order blank which accompanies the proof; a copy wil' be sent on request. Entered as second-class matter August 21, 1896, at the Post-Office at Chicago, Illinois, under the Act of March 3, 1879. Acceptance for mailing at sr'''''a' rate of postage provided for in Section 1103, Act of October 3, 1917, authorized onj July IS. 1918. VOLUME LXVII NUMBER 6 THE Botanical Gazette JUNE igig structure, development, and distribution of so-called rims or bars of sanio Irving W. Bailey (with plates xiii-xv) Introduction In recent years a number of botanists and paleobotanists have given considerable attention to the study of the distribution of certain bandhke thickenings of the middle lamella, so-called rims or bars of Sanio, in the g}innosperms, and their significance in discussion concerning the relative antiquity of the Abieteae and Araucarieae.^ Before considering the distribution of these bandlike thickenings of the middle lamella, it is desirable to outline the conclusions of various investigators concerning their structure and development. The work of Sanio is particularly significant in this connection, as it is also in a discussion of the controversy that has arisen in regard to the true meaning of the term "bars of Sanio." Historical Structure and development of Sanio's Querleisten. — It was stated by Sanio (13), in his comprehensive paper upon the anatomy of Piniis silvestris Linn., that in the cambium of young stems the radial and tangential partitions separating adjacent protoplasts are of equal or nearly equal thickness, but in that of ' The terminology of Engler and Gilg (2) is used in this paper. 449 4SO BOTANICAL GAZETTE [june old stems the radial walls are considerably thicker. He showed that each of these thick radial partitions is distinctly stratified, and consists of a central cellulose septum, Zwischensubstanz, overlaid by cellulose layers belonging to the protoplasts on either side of the wall. Furthermore, he believed that at an early stage in the development of tracheids thin spots appear in the radial primary walls; and that, as the tracheids increase in size, these areas become larger, the Zwischensubstanz is gradually absorbed, and the outer layers of the wall are fused together and stretched to form a thin, more or less homogeneous membrane. Inasmuch as bordered pits are subsequently laid down over portions of these attenuated areas, Sanio considered them to be Primordialtupfeln. He described them as follows (p. 74) : Wahrend diese Verdiinnungen in der Membran seitlich. d. h., in horizon- taler Richtung allmahlich in den starken verdickten Theil iibergehen grenzen sie sich Oben und Unten scharf ab, und erscheinen hier sogar zuletzt mit dop- pelten Umrissen. Haufig liegen diese Verdiinnungen so nahe an einander, dass die sie trennenden verdickten Stellen als Querleisten erscheinen. Unter- sucht man diese Bildung in Tangentialschnitte, so erscheinen diese Querleisten als knotenformige Verdickungen der Membran zweier Nacharzellen, wahrend die Verdiinnungen als zarte Scheidwande sich ausweisen. It is evident from Sanio's figures and descriptions that he considered the primary pit areas, Primordialtupfeln , to be separated by others in which the Zwischensubstanz is not entirely absorbed, and in which the outer cellulose layers are less attenuated. It should be emphasized, in this connection, that Sanio used the word Querleisten, cross-pieces or cleats, in referring to transverse thicker strips of the middle lamella between closely approximated primary pit areas, and the word Umrissen, contours, in referring to the upper and lower outlines of these areas. It is difficult to determine with certainty whether Sanio understood the real significance of the doppelten Umrissen, which partially surround the more isolated primary pit areas in certain. of his drawings. That he probably considered them to be contours, outlining the top and bottom of a sloping surface or escarpment, is indicated by his illustration (pi. 10, fig. 2) of a tangential section of a young tracheid of Pinus silvestris. I have found no conclusive evidence in his text or figures iqiq] bailey— bars of SANIO 451 to indicate that he considered these parallel curved lines as outHnes of a heavily embossed or thickened rim. Strasburger (15) showed very clearly that in Pinus and Larix Sanio's doppelten Umrissen may be the outhnes of embossed portions of the middle lamella. In other words, he made it evident that when the primary pit areas are close together they are sep- arated by a single transverse thickening, but that when they are not closely approximated there may be two curved thickened strips, separated bv a less heavily embossed area, between them (fig. 5). Miss Gerry (3), in discussing the distribution of bandlike thickenings of the middle lamella in the Gymnospermae, referred to them as follows (p. 119): The "bars" or "folds" of cellulose which when stained with haematoxylin are especially obvious as horizontal or more or less semicircular markings in the tracheide walls of a radial section from such a conifer as Pinus silvestris L. were described by Sanio in 1872 These structures were named "Bars of Sanio" from him Groom and Rushton (5), who have studied the structure and chemical composition of the bandUke thickenings of the middle lamella in Indian species of Pinus, state: According to Samo's work it is these unoccupied margins of the primary areas that coincide in position with the above-mentioned bands that are seen in radial section. Consequently the name " Sanio's rims " may be given to the structures causing the bandlike appearance When the primary pit areas are in contact, the two contiguous Sanio's rims are naturally "fused" and form a band that is transverse and single, except possibly at the two lateral edges where the natural curv^ature of each original boundary of the area causes the band to fork In radial sections with iodine and sulphuric acid the "rims" stain yellow; with ordinary haematoxylin they remain unstained; leav- ing sections in cupra-ammonium to dissolve out any cellulose, their staining properties are not changed materially. They are not composed of cellulose. .... When young the actual marginal portion of the primary pit area does not thicken by deposits of lignified wall as soon as it does elsewhere (except on the pit-closing membrane), but thickens by successive deposits of pectic substance until a stage is reached when lignified wall-substance is deposited even over the now thickened rims of the primary pit area. Sanio's rims repre- sent a system of rodlike or bandlike pectic thickenings of the middle lamella running transversely in the radial walls and linked here and there by slightly 452 BOTANICAL GAZETTE [june curved longitudinal bandlike similar thickenings (representing the lateral mar- gins of the primary pit areas) C. MtJLLER (1890) was the author of the name "Sanio's Bars" and, as he explicitly stated, he coined the term to designate these structures,^ as first discovered by Sanio in Piniis silvestris (1873-74). Jeffrey (7), however, still maintains that the bandlike thicken- ings of the middle lamella in the Ginkgoales, Abieteae, Taxoideae, Cupresseae, and Taxaceae are typical ''bars of Sanio." Sifton (14) considers that the rims of neighboring primary pit areas unite to form bars and uses the latter term in referring to bandlike thick- enings of the middle lamella that occur in Araiicaria and Cycas. Bars of Sanio vs. trabeculae. — The fact that the term bars of Sanio is used by certain investigators in referring to trabeculae and by others in describing entirely different structures is unfortu- nate and likely to lead to considerable confusion. In a paper, published in 1863, Sanio (12, p. 17) described trabeculae as: ''Quer- balkens quer durch den Zellenraum von einer Wandung zur andern verlaufen." In 1890, Muller (9) referred to these structures as Sanio'sche Balken; and later Penhallow (id) called them Sanio's bands. It is evident that Sanio and Muller used the word Balken (beams) to designate rodlike structures that are attached at their ends and cross the lumens of cells. In view of this fact, and that Sanio used the word Querleisten (cross-pieces or cleats) in referring to bandlike thickenings of the middle lamella, the terms "bars of Sanio" and Sanio'sche Balkeji are not necessarily synonymous. As has been pointed out by Groom and Rushton, however, Jeffrey and his students were undoubtedly mistaken in supposing that the bandlike thickenings of the middle lamella had been named after Sanio. This conclusion is strengthened by the fact that the phrase "diese scheibenformige Verdickung der Scheidewand ist bisher iibersehen" (used by Sanio [13, p. 78] in referring to the torus) was interpreted as indicating that Sanio considered him- self the discoverer of the bandhke thickenings of the middle lamella; whereas, as a matter of fact, he expressly stated (13, p. 74) that ''Deartige Bildungen hat bereits linger gesehen, aber nicht zu deu- ten gewusst." Although the use of the term "bars of Sanio" was undoubtedly unfortunate, Jeffrey and his students do not appear ever to have » Trabeculae. iqiq] bailey— bars of sanio 455 actually confused Sanio's Querleisten with trabeculae. Further- more, it is to be emphasized that Groom and Rushton consider the rodlike thickenings, between closely approximated primary pit areas, as fusions of two thickened rims. The word rims, therefore, is not an entirely satisfactory substitute for the word bars in referring to Sanio's Querleisten. Distribution and supposed phylogenetic significance of Sanio's Querleisten in Coniferae. — There are considerable dif- ferences of opinion among various investigators concerning the distribution and phylogenetic significance of these bandlike thick- enings of the middle lamella. Jeffrey (7) and his students (3, 6) maintain that they are conspicuously developed in the older wood of Ginkgo and all of the Coniferae except the Araucarieae. Gothan (4) assumes that they are absent in the Araucarieae because the pits are so closely packed together that there is no room for such struc- ture. Thomson (16), on the other hand, considers that they are present in rudimentary form in the Araucarieae, and are closely applied to the margins of the bordered pits. Whatever view is taken in regard to the relative antiquity of the Abieteae and Araucarieae, it must be admitted that there is a very striking difference between the older wood of the Araucarieae and that of the Abieteae, Taxodieae, Cupresseae, Taxaceae, and Ginkgo. "Alternate'' pitting (fig. 13) is stereotyped in the Araucarieae; whereas "opposite" pitting and Querleisten (pi. XV) are firmly fLxed in the Abieteae, Taxodieae, Cupresseae, and Taxaceae. There appear to be no true transitional series between these two types of secondary xylems that may be considered to indicate conclusively that the latter type of pitting is a modification of the former. In so far as tracheary pitting is concerned, the principal argu- ments in favor of deriving the Abieteae and Taxaceae from the Cordaitales or Araucarieae are based upon the anatomy of the young wood of seedlings, cone axes, and the first annual rings of stems and roots, so-called conservative regions. Thus Jeffrey (7) maintains that the presence of alternate pitting and the absence of bandlike cellulose thickenings of the middle lamella in the young wood of reproductive axes, leaf strands, and the first annual rings of Ginkgo 454 BOTANICAL GAZETTE [june and Pinus, are evidences of the Cordaitean ancestry of these genera. He considers the opposite pitting and thickenings of the middle lamella which occur in the cone axes of Araucarians to indicate that the Araucarieae are descended from forms resembling Ginkgo and Pinus. Furthermore, although admitting that "bars of Sanio" are absent or "evanescent" in the young wood of seedlings and the first annual rings of the stems and roots of Araucarieae, he interprets the absence of approximation and consequent flattening of the bordered pits in such tissue as evidence for deriving the Araucarians from pinelike ancestors. The accuracy of these conclusions has been questioned by Thomson (i6) and Sifton (14), who have figured and described bandlike thickenings of the middle lamella in the tracheids of the petiole of Cycas and the cone axes, seedlings, and first formed secondary xylem of the stems and roots of Pinus and other Abieteae. Thomson (16) interprets the rimlike thickenings and alternate pitting that occur at times in the cone axes and first annual rings of stems and roots of Pinus as indicating that the Abieteae are descended from the Araucarieae. It is to be emphasized, in this connection, that in dealing with other structural characters Jeffrey interprets the anatomy of selected conservative regions or organs of the Abieteae and Arau- carieae as indicating that the latter are descended from the former ; whereas Thomson, by applying the same laws to similar material, proves the reverse to be true. Such discrepancies as these suggest that there may be a con- siderable element of danger in placing too much emphasis upon "laws" of recapitulation, reversion, and retention in arguments concerning the phylogeny of plants. Even the most ardent advo- cates of these doctrines admit, in certain cases at least, that ceno- genetic characters do occur in seedlings, roots, traumatic tissue, cone axes, etc. So long as this is acknowledged to be so, it must be extremely difficult, in the absence of reliable collateral evidence, to determine with certainty whether a given structure in a given region is palingenetic or cenogenetic. In other words, even if it should be proven, by means of careful statistical and experimental investigations, that certain organs or regions of plants are inherently iqiq] bailey— bars OF SANIO 455 somewhat more conservative or slower to change than others, considerable difficulty must inevitably be encountered in formulat- ing such facts as these into laws for use as "short cuts" in the study of phylogeny. That this is likely to be the case in dealing with the Ginkgoales and Coniferae is indicated by a number of facts in the comparative anatomy and ecology of the Pteridophyta and Gymnospermae. In the evolution of these groups the primary, as well as the secondary, tissues appear to have been considerably modified. For example, the more primitive vascular plants were characterized by having relatively wide zones of primary xylem; whereas the Coniferae have usually only a relatively limited amount of this tissue, which is correspondingly specialized in structure. Structure and distribution of bandlike thickenings of middle lamella in Pteridophyta, Gymnospermae, and Angiospermae In view of the fact that much emphasis has been placed upon bars of Sanio in the identification of fossil woods of the Mesozoic, and that these structures have been used as the basis for important but conflicting generalizations in regard to the phylogeny of the Coniferae and the relative conservatism of different organs or regions of plants, the structure and distribution of bandlike thickenings of the middle lamella in the Pteridophyta, Gymnospermae, and Angiospermae deserve more careful consideration than they have received heretofore. » As is well known, the metaxylem of most Filicales is composed largely of scalarifomi tracheids. The bordered pits in these tracheary elements are much elongated horizontally, at right angles to the long axis of the tracheids, and are closely approximated in vertical series (fig. 2) . The elongated bordering areas of the second- ary walls are exactly superimposed over attenuated areas of the middle lamella; and the outlines of these areas are more or less effectively concealed by the margins of the bordering areas. The primary pit areas are separated by narrow, bandlike, thicker portions of the middle lamella, which, in carefully stained^ longi- tudinal sections of the xylem, appear as fine dark lines between the ^ Haidenhain's iron-haematoxylin and safranin. 456 BOTANICAL GAZETTE [june bordered pits. Owing to the approximation of the bordered pits and the thickness of the secondary walls, however, these Querleisten are usually more conspicuous when seen in section (fig. 19) than in surface view. This scalariform type of tracheary pitting becomes at times considerably modified. Thus the elongated bordering areas of the secondary wall may be replaced by two or more shorter elongated or oval bordering areas (fig. 2). Under these circumstances the pri- mary wall frequently retains its typical scalariform pitting after the secondary wall has lost it; that is to say, each horizontal row of smaller bordering areas is laid down over a single elongated primary pit area (fig. 6). In other cases the elongated bordered pits of the secondary walls may become contracted to form smaller bordered areas which cover only a portion of the surface of the elongated primary pits, and the Querleisten project beyond the outlines of the bordering areas. This process of reduction in the pitting of the secondary wall may even be carried to a point where the primary pit areas have no superimposed bordered pits; or the primary pit areas become less closely approximated, of oval or circular outlines, and separated by relatively wide biconcave thickenings with forking ends (fig. 7). Scalariform pitting also grades into types in which there is less unconformity between the primary and secondary walls. The elongated bordered pits become replaced by vertical rows of smaller pits which are staggered so that the pits in one row alternate with those in the next series. These pits are usually superimposed over nearly the whole surface of similar primary pit areas, and the thicker portions of the middle lamella tend to anastomose or form a reticulum, as is shown in fig. 6a. Such transitions between scalariform and derived types of tracheary pitting occur in other groups of vascular plants. In certain of the paleozoic and lower mesozoic plants, which had "open" bundles, the metaxylem and secondary wood were com- posed of scalariform tracheids; whereas, in others, the scalariform bordered pits were more or less completely replaced by horizontal or diagonal rows of smaller pits, except in the tracheids of the younger wood of the stele. In the latter types, in passing from the 1919] BAILEY— BARS OF SANIO 457 younger to the older metaxylem or secondary wood, there were transitions betweerl typical scalariform and opposite and alternate multiseriate pitting. Such transitional stages between scalariform -and multiseriate pitting have been observed in a number of Sphenophyllales, Calamariales, CycadofiHcales, Cordaitales, and Bennettitales. In Protopitys, Cycadeoidea Dartoni (Coulter and Chamberlain) Wiel. (figs. 14, 20), and other forms whose secondary jcylems show indications of zonation, such transitions occur periodi- <:ally in the older wood of the stem, as they do in the stems and roots of the vesselless angiosperms, Tetracentron and Trochodendron (figs. 15,16). In these transitional regions the elongated primary pit ^reas and Querleisten that underlie the scalariform secondary walls tend to persist in tracheids having horizontal rows of bordering areas. It is such transitional types of tracheary pitting that have been figured by Jeffrey in the cone axes of Araucarians, and by Sifton in the petioles of Cycas. In Araucaria Bidwillii Hook., owing to the fact that the middle lamella is often relatively thick and the pits not closely approximated, the Querleisten are frequently broad and conspicuous. In transitional tracheids they may break into fragments which cling to the margins of the bordered pits, even after the latter have shifted to the alternate arrangement. Eventu- ally, however, the more or less circular primary pit areas of the older tracheids appear to become surrounded on all sides by equally thickened portions of the middle lamella. Scalariform and transitional types of bordered pitting occur in the lateral walls of the vessels of many dicotyledons. In fig. 3 is illustrated the typical scalariform bordered pitting that occurs in the radial and tangential walls of the vessels of certain Magnoliineae. This type of pitting is in marked contrast to the multiseriate pitting shown in figs. 4 and 9. As these figures indicate, the bordered pits which form the horizontal rows may be closely packed together and have flattened sides, or they may be more loosely arranged and have oval outhnes. Transitional stages, between these typical scalariform and multiseriate types of pitting, are of frequent occurrence in the vessels of certain of the Magnoliineae (figs. 4, 11). In photomicrographs of carefully stained sections there are thin dark colored lines between the elongated or scalariform 458 BOTANICAL GAZETTE [june bordered pits. These transverse strips are deeply stained portions of the middle lamella that stand out in sharp contrast to the thin pit membranes. They are, in fact, very narrow, bandlike thicken- ings {Querleisten) which separate the elongated primary pit areas. In the case of vessels withmultiseriate pitting, these transverse ridges or Querleisten tend to occur between the elongated primary pit areas that underlie the horizontal rows of bordered pits. In other words, a single elongated bordered pit may be laid down over the whole surface of the elongated primary pit area, or one or more smaller bordered pits may be laid down over portions of its surface. Under the latter circumstances, the outlines (Umrissen) of the primary pit areas become more conspicuous (figs. 4, 9, ii). In certain cases these Querleisten become more or less completely divided into shorter, rodlike thickenings, which lie between the upper and lower margins of contiguous bordered pits of the vertical series. It was shown by Strasburger (15) that bordered pits are not laid down over all the primary pit areas in the tracheids of Pinus and Larix. In the Magnoliineae and other dicotyledons the lateral primary walls of the vessels frequently have elongated attenuated areas, which have no bordered pits superimposed over them, or only a comparatively limited part of their surface so covered (figs. I, 8, 12). In the Magnoliineae, Trochodendrineae, and other groups of dicotyledons there is much evidence to indicate that scalariform pitting is a relatively primitive feature in the structure of vessels. That is to say, those vessel segments which most closely resemble tracheids in general form and structure tend to have scalariform or scalariform and opposite multiseriate pitting; whereas the larger and more specialized conducting passageways are characterized by having alternate multiseriate pitting in their lateral walls (fig. 10). In the evolution of larger and more specialized vessels the modi- fication of the primary walls does not appear, in many cases, to have kept pace with that of the secondary walls. Thus in primitive types of vessel segments the elongated bordered pits are exactly superimposed over similar elongated primary pit areas, but in more iqiq] bailey— bars OF SANIO 459 specialized types the primary wall (the first formed portion of the vessel member) tends to retain its primitive elongated type of primary pit areas after the scalariform bordered pits have become locally constricted (fig. 4), or divided into horizontal rows of smaller pits (fig. 9). Similarly, in those wa;lls where there is a tendency to eliminate the bordered pits, the elongated primary pit areas persist after the bordered pits have partially or completely disappeared (figs. I, 8, 9). In the highly specialized vessels of the Anonaceae and Laura- ceae (fig. 10) there are numerous circular or oval bordered pits in the lateral walls of the vessels. Usually they appear to be laid down over similar circular primary pit areas. In other words, in the most highly specialized types of vessels, in which the bordered pits are not arranged in horizontal rows, even the elongated primary pit areas are more or less completely obhterated, and replaced by attenuated areas with circular outlines. Vestiges of Querleisten, however, are sometimes present near the upper and lower margins of the bordering areas. In the dicotyledons, with increasing specialization of the vessels, there is a corresponding reduction in the pitting of the remaining tracheary elements. Thus tj^ical tracheids are replaced by fiber tracheids, which are in turn replaced by libriform fibers. The fiber tracheids of certain dicotyledons have elongated or oval primary pit areas that are separated by wide dark colored bands. Furthermore, it is not uncommon to find that many of the attenu- ated areas of the middle lamella have no superimposed bordered pits (fig. 22). Unconformity of the type that occurs in the tracheids of various primitive vascular plants, and in the vessels of certain dicotyledons, has been observed in the wood of Ginkgo, and certain of the Abieteae and Taxodieae. Fig. 23 illustrates a type of tracheary pitting that is of frequent occurrence in the older secondary wood of vigorous mature specimens of Ginkgo and Taxodium. It is most character- istically developed in large thin-walled tracheids of the so-called spring wood. The numerous, uniformly narrow, elongated primary pit areas and thin, straight, narrow Querleisten are typically scalariform in structure. Each primary pit area has superimposed 46o BOTANICAL GAZETTE [june over it 2-4 bordered pits. The latter are somewhat elongated in many cases, and frequently are so closely approximated as to be flattened by mutual contact and to cover nearly the whole of the primary pit areas. This type of tracheary pitting grades into a second type in which the surfaces of the elongated primary pit areas- are only partially covered by circular bordered pits (figs. 24, 25). The latter type, in turn, grades into a third type in which the primary pit areas are not typically scalariform in structure. Cer- tain of the attenuated areas appear to increase in size at the expense of intervening areas, which are either eliminated entirely or persist as constricted areas that are not overlaid by bordered pits (figs. 26^ 29, 30). By this process of specialization certain of the primary pit areas become oval or biconvex and less closely approximated. Certain of the Querleisten tend to widen and to become biconcave with forking ends; whereas others are crowded together and appear to fuse to form similar biconcave thickenings (figs. 26, 29, 31). This type of pitting grades into others in which the attenuated areas that are overlaid by bordered pits become more circular and more widely separated, the intervening primary pit areas and Querleisten (except the curved bands that commonly persist above and below the bordered pits) becoming vestigial or obhterated (figs. 27, 28). In certain cases the portions of the middle lamella between widely separated primary pit areas may be uniformly thickened so that the pits appear to be separated by a single, wide dark colored band. In other cases, for example, in the thick- walled fiber-like cells of the so-called summer wood, and in the small tracheids which occur in seedhngs and the first annual rings of stems and roots, the bordered pits frequently tend to be super- imposed over nearly the whole surface of the circular primary pit areas, and the curved Querleisten cling to the upper and lower margins of the bordering areas, or are completely obhterated. Such transitions in tracheary pitting have been observed in Larix, Pinus, Abies, Sequoia, and other genera of the Abieteae and Taxodieae, as well as in Taxodium and Ginkgo. The more elongated primary pit areas and the narrower, straighter Querleisten tend to occur in the larger, thinner- walled, heavily pitted tracheids; and are therefore most characteristically developed in the first formed igig] BAILEY— BARS OF SANIO 461 portions of the growth rings of the older secondary xylem. The particular t>'pes of unconformity and bandlike thickenings of the middle lamella which occur in a given species vary considerably in plants grown under different environmental influences and in different organs or regions of a single plant. This is as true ot the first formed as the older secondary xylem. Discussion It is evident that bandlike thickenings'' of the middle lamella, separating more or less elongated primary pit areas, are not confined to the tracheids of certain Coniferae, but are widely distributed among the Pteridophyta, G>Tnnospermae, and Angiospermae.s Any interpretation of the so-called rims or bars of Sanio in the Goniferae, therefore, should be in general harmony with the structure and distribution of these bandlike thickenings in other groups of vascular plants. Since bandlike thickenings of the middle lamella and transitions between scalariform and "alternate" and "opposite" multiseriate pitting are of common occurrence in the younger xylem of many paleozoic and mesozoic (as well as less primitive) plants, the occur- rence of opposite (as well as alternate) pitting and more or less rudimentary Qiierleisten in the transitional tracheids of the cone axes of Araucarians does not indicate conclusively that the Arau- carieae are descended from the Abieteae. Similarly, the more or less sporadic occurrence of alternate pitting, as well as opposite pitting and "bars of Sanio," does not indicate necessarily that the secondary xylem of the Ginkgoales, Abieteae, Taxodieae, Cupres- seae, and Taxaceae is a modification of that which occurs in the Araucarieae or Cordaitales. Furthermore, there are a number of facts in the comparative anatomy and ecology of the Pteridophyta, Gymnospermae, and Angiospermae which suggest that unconformity between the pits in the primary and secondary walls of tracheids and vessels may be " WMch stain dark blue in sections treated with Haidenhain's iron-haematox^din. 5 The bandlike thickenings are usually inconspicuous in surface views of the facets of tracheary elements owing to the fact that they are concealed by the thick, super- imposed secondary walls. 462 BOTANICAL GAZETTE [june a phenomenon that is concomitant of processes of modification or reduction in tracheary pitting. In the metaxylem of ferns and in the secondary xylem of a number of primitive vascular plants, the primary wall frequently tends to retain its scalariform structure after the scalariform bordered pits in the secondary wall have been replaced by horizontal rows of smaller pits. It may even retain its elongated pit areas and bandlike thickenings after the bordering areas of the secondary wall have been considerably contracted, or have disappeared entirely from certain portions of a facet. On the other hand, when scalariform pitting is replaced by alternate multiseriate pitting, the bandlike thickenings of the middle lamella tend to anastomose and form a reticulum. Similar phenomena occur in the metaxylem of many of the higher vascular plants, in the tr&cheids of the secondary xylem of the vesselless dicotyledons Tetracentron, Trochodendron, and Drimys, and in the lateral walls of the vessels of many of the angiosperms. This general tendency for the persistence of scalariform pitting in the middle lamella, after it has disappeared more or less com- pletely from the secondary wall, raises an interesting question in regard to the probable significance of the scalariform pitting which occurs so commonly in the middle lamellae of certain of the Abieteae, Taxodieae, Taxaceae, Cupresseae, and Ginkgo, but appears to be entirely absent in the later formed secondary tra- cheids of the Araucarieae. It is important to note in this connection that the more primitive vascular plants, which possessed relatively wide zones of primary xylem, were characterized by having numerous closely approxi- mated pits in the radial facets of their relatively large tracheids. In the evolution of the Ginkgoales and Coniferae there appears to have been a more or less pronounced reduction in the amount of primary xylem, in the size of the first formed secondary tracheids of the stele, and in the number of bordered pits in the walls of the tracheary elements. The large, thin-walled, heavily pitted tracheids which occur in the' spring wood of the older secondary xylem of mesophytic Conif- erae, resemble the primitive types of tracheids more closely than do the thick-walled, highly specialized elements of the summer wood, igig] BAILEY— BARS OF SANIO 463 or the relatively small tracheids of the first formed portion of the secondary xylem. It is in these larger thin-walled tracheids that the most typical scalariform primary pit areas tend to occur. Occasionally, where the tracheary pitting is very strongly developed, scalariform bordering areas of the secondary wall are superimposed over portions of these elongated primary pit areas (fig. 23). This is likely to occur in Taxodium and Ginkgo and roots of certain of the Abieteae. As has previously been shown, the scalariform structure of the middle lamella and narrow straight Querleisten become gradu- ally modified with increasing reduction in the number of bordered pits. Furthermore, it has been shown that a similar widening of the bandlike thickenings of the middle lamella may occur in certain of the Pteridophyta, as well as the Angiospermae, when the bordered pits tend to become more or less isolated. The occurrence of these interesting structures in the Abieteae, Taxodieae, Cupresseae, Taxaceae, and Ginkgo, and their absence in the secondary wood of Araucarieae, are difficult to explain upon the assumption that the former groups are descended from ancestors having "alternate multiseriate" pitting. On the other hand, from analogy with similar phenomena in other groups of vascular plants, their occurrence is easily accounted for if the microphyllous and relatively xerophytic Coniferae are descended from forms having scalariform tracheary pitting. Primary pit areas of cambium and their relation to the pitting of xylem and phloem The important observations of De Bary, Janczewski, Russow, Strasburger, Kienitz-Gerloff, KRUGER,-and others upon the occurrence of bandlike thickenings of the middle lamella in the cells of the cambium, and their relation to similar structures in the ele- ments of the xylem and phloem, have been overlooked entirely in discussions concerning the phylogenetic significance of the so-called rims or bars of Sanio. Kruger (8) found that, in the cambia of all plants (stems and roots of gymnosperms, dicotyledons, monocoty- ledons, including trees, shrubs, herbs, and succulents) investigated by him, there were leistenjormige Verdickungen in the radial parti- tions (figs. 17, 18). These ridgelike thickenings separated roundish 464 BOTANICAL GAZETTE [june thin spots or pit areas and were more conspicuous in the winter than in the spring and summer, a fact previously noted by De Bary (i). Furthermore, Krxjger traced these structures to the "procambial strands" and through the developing daughter cells of the cambium to the highly differentiated cells of the xylem and phloem. They appeared to be least modified in the walls of phloem parenchyma. In the development of tracheids, vessels, and sieve tubes, owing to increase in the size of the individual cells, they tended to become more or less modified. KriJger concluded, however, that the thin spots became enlarged and modified to form sieve plates (fig. 21), and to form the primary pit areas of tracheary elements. He noted that during this process the ridgelike thickenings tended to become more or less modified in shape. Russow (11) and Strasburger (15) emphasized the fact that in the Abieteae many of the pit areas become more or less obliterated in the development of tracheids and sieve tubes; that is to say, many of the primary pit areas become vestigial, since they have no superimposed bordered pits or do not become modified to form sieve plates. It is evident, therefore, that not only are bandhke thickenings of the middle lamella of frequent occurrence in the tracheary elements of Pteridophyta, Gymnospermae, and Angiospermae, but also in the cells of the cambium and phloem. In discussing the phylogenetic significance of the bandlike thickenings of the middle lamella in tracheids of Coniferae, therefore, it is essential, not only to compare these structures with similar structures which occur in other types of cells and other groups of plants, but also to contrast the various stages in their development. This must inevitably be the case, since there is a considerable element of danger in basing generalizations concerning relationships upon comparisons between the structure of end products. Of course similar structures may be attained through entirely different developmental stages. Unfortunately, comparati^vely little is known in regard to the detailed structure of the cambium and the various stages in the development of tracheary pitting in different groups of vascular plants. It is to be hoped that this gap will be filled in the near future, since information of this character promises to throw con- iqiqI bailey— bars of sanio 465 siderable light upon a number of interesting problems, especially upon the structure and true significance of the so-called rims or bars of Sanio. Summary 1. BandUke thickenings of the middle lamella and scalariform primary pit areas are characteristic of tracheids which have scalari- form bordered pits. They are widely distributed among the Pteridophyta, G>Tnnospermae, and Angiospermae. 2. The middle lamella frequently retains its typical scalariform structure after the secondary wall has lost it. 3. In the G>Tnnospermae, as well as in the Pteridophyta and Angiospermae, there appear to be transitions between primary membranes of this t\pe and others in which the scalariform structure is profoundly modified, 4. The comparative anatomy and ecology of the Pteridophyta, Gj-mnospermae, and Angiospermae afford considerable evidence which suggests that the tv-pes of unconformity and peculiar band- like thickenings of the middle lamella (so-called bars or rims of Sanio) which occur in certain Pteridophyta and Angiospermae, as well as in many G>Tnnospermae, are concomitants of processes of modification or reduction in tracheary pitting. 5. The structure of the walls of the cambium and the develop- ment of the pitting in the elements of the xylem and phloem in Pteridophyta, G>Tnnospermae, and Angiospermae deserve more careful consideration in discussions concerning the phylogenetic significance of the so-called rims or bars of Sanio, than they have received heretofore. BussEY Institution Jamaica Plain, Mass. LITER.\TURE CITED Bary, a. de, Vergleichende Anatomic der Vegetationsorgane der Phanero- gamen und Fame. Leipzig. 1877. Engler, a., and Gilg, E., Syllabus der Pflanzenfamilien. Berlin. 1912. Gerry, Eloise, The distribution of the "Bars of Sanio " in the Coniferales. Ann. Botany 24:119-123. 1910. 466 BOTANICAL GAZETTE [june 4. GoTHAN, W., Die fossilen Holzreste von Spitzbergen. K. Svenska Vetensk. Akad. Handl. 45. 1910. 5. Groom, P., and Rushton, W., The structure of the wood of East Indian species of Pinus. Jour. Linn. Soc. (London) Bot. 41:457-490. 1913. 6. HoLDEN, Ruth, Some fossil plants from eastern Canada. Ann. Botany 27:243-255. 1913. 7. Jeffrey, E. C., The history, comparative anatomy, and evolution of the Araucarioxylon type. Proc. Amer. Acad. 48:531-571. 1912. 8. Kruger, F., tjber die Wandverdickungen der Cambiumzellen. Bot. Zeit. 50:633-640, 649-657, 665-673, 681-688, 702-708. 1892. 9. MiJLLER, C, tJber die Balken in den Holzelementen der Coniferen. Ber. Deutsch. Bot. Gesells. (Gen.-Versamml.) 8:17-46. 1890. 10. Penhallow, D. p., a manual of the North American Gymnosperms. Boston. 1907. 11. Russow, E., tJber die Entwicklung des Hoftiipfels der Membran der HolzzeUen und des Jahresringes bei den Abietineen in erst en Linie von Pinus silvestris L. Sitzber. Naturf. Gesells. Dorpat. 6:109-158. 1881. 12. Sanio, K., Vergleichende Untersuchungen liber die Elementarorgane des Holzkorpers. Bot. Zeit. 21 : 85-91 , 93-98, loi-i 1 1 , 1 13-1 18, 1 2 i-i 28. 1863. 13. , Anatomic der gemeinen Kiefer {Pinus silvestris L.) IL Jahrb. Wiss. Bot. (Pringsheim) 9:50-126. 1873-74. 14. SiFTON, H. B., On the occurrence and significance of "bars" or "rims" of Sanio in the Cycads. Bot. Gaz. 60:400-405. 1915. 15. Strasburger, E., tJber den Bau und das Wachsthum der Zellhaute. 38-64. Jena. 1882. 16. Thomson, R. B., On the comparative anatomy and affinities of the Arau- cartnese. Phil. Trans. Roy. Soc. London. B. 204:1-50. 1913. EXPLANATION OF PLATES XIII-XV PLATE XIII Fig. I. — Magnolia acuminata Linn.: pitting in facet of small vessel seg- ment, showing scalariform primary pit areas and isolated circular bordered pits; X300. Fig. 2. — Todea hymenophylloides A. Rich.: pitting in lateral facet of tracheid, showing typical scalariform and transitional types of pitting; X300. Fig. 3. — Magnolia acuminata: scalariform pitting in lateral facet of vessel segment ; X400. Fig. 4. — Magnolia acuminata: pitting in lateral facet of vessel segment, showing scalariform and transitional types of pitting; X450- Fig. •,.— Pinus Strobus Linn.: tangential longitudinal section, showing knotenformige thickenings of middle lamella; X500. Fig. 6. — Todea hymenophylloides: pitting in lateral facet of tracheid, showing pairs of elongated bordering areas superimposed over single scalariform 1919I BAILEY— BARS OF SAN 10 467 primary pit areas; portions of bandlike thickenings of middle lamella visible between rows of bordered pits; X300. ■ Fig. 6(2. — Todea hymenophylloides: showing alternating arrangement of bordered pits and anastomosing of bandlike thickenings of middle lamella; X300- Fig. 7. — Acrostichum sorbifolium Linn.: lateral facet of tracheid, showing small, circular, bordered pits and biconcave thickening of middle lamella^ latter somewhat indistinct owing to thickness of superimposed secondary wall; X8cx>. Fig. 8. — Cercidiphyllum japonicum Sieb. and Zucc: lateral facet of vessel segment, showing persistence of scalariform primary pit-areas after more or less complete disappearance of bordered pits; Xsoo- Fig. 9. — Magnolia macrophylla Michx.: lateral facet of vessel segment, showing unconformity between primary and secondary pitting. Fig. 10. — Asimina triloba (Linn) Dun.: lateral facet of vessel segment, showing alternate multiseriate pitting; X300. Fig. II. — Cercidiphyllum japonicum: lateral facet of vessel segment, showing transitional types of bordered pitting and imconformity between primary and secondary walls ; X400- Fig. 12. — Cercidiphyllum japonicum: lateral facet of vessel segment, showing unconformity between pitting of primary and secondary walls; X40o- PLATE XIV Fig. 13. — Agathis auslralis Steud.: pitting in radial facet of tracheid, showing "alternating" type of arrangement; X3SO- Fig. i4.—Cycadeoidea Dartoni (Coult. and Chamb.) Wiel.: radial longi- tudinal section of secondary xylem, showing transitional types of pitting; Xioo. Fig. 15. — Trochodendron aralioides Sieb. and Zucc: radial facets of 2 tracheids, showing transitions between scalariform and opposite bordered pitting; Querleislen appear as narrow dark lines separating scalariform pits or pairs of smaller bordered pits; X350. Fig. 16. — Trochodendron aralioides: radial longitudinal section of second- ary xylem, showing transitional types of pitting and (right) persistence of scalariform primary pit areas with reduction in number of bordered pits; X200. Fig. 17.— Ulmus americana Linn.: tangential longitudinal section of cambium (winter condition), showing knotenjormige thickenings of middle lamella; X230. Fig. 18. — Ulmus americana: radial longitudinal section of cambium and young phloem cells, showing sculpture of middle lamella; X200. Fig. 19. — Pteris aquilina Linn.: adjacent walls of 2 tracheids in sectional view, showing knotenjormige thickenings ("bars of Sanio") of middle lamella; X300. 468 BOTANICAL GAZETTE [june Fig. 20. — Cycadeoidea Dartoni: radial longitudinal section of secondary xylem, showing transitions between scalariform and opposite and alternate pitting; Xioo. Fig. 21. — Juglans nigra Linn.: radial longitudinal section through cam- bium (left) and young phloem (right), showing relation between sieve plates and primary pit areas; Xi8o. Fig. 22. — Kayea paniculata Merrill: radial longitudinal section of second- ary xylem, showing primary pit areas and "bars of Sanio"; X200. PLATE XV Fig. 23. — Taxodium distichum (Linn.) Richard: radial facet of tracheid, showing scalariform structure of primary wall and crowded bordered pits; X400. Fig. 24. — Larix occidentalis Nutt.: radial facets of 3 tracheids, showing scalariform structure of primary wall; X350. Fig. 25. — Taxodium distichum: radial facet of tracheid, showing scalari- form structure of primary wall and less crowded bordered pits than fig. 22,] X400- Fig. 26. — Larix occidentalis: radial facets of 2 tracheids, showing numerous primary pit areas having no superimposed bordered pits; X220. Fig. 27. — Pinus Strobus: radial facets of 3 tracheids, showing modified t3T)es of primary wall structure ; X 220. Fig. 28. — Pinus Strobus: radial facets of 3 tracheids, showing reduction in bordered pitting and persistence of primary pit areas; X220. Fig. 29. — Larix occidentalis: radial facet of tracheid, showing curving and fusing (or widening) of bandlike thickenings of middle lamella; X240. Fig. 30. — Pinus Strobus: radial facets of several tracheids, showing scalariform structure of primary wall; X220. Fig. 31. — Taxodium distichum; radial facet of tracheid, showing curving of bandlike thickenings of middle lamella with reduction in number of bordered pits; X400. Figs. 14, 17, 18, 20, and 21 were made from sections loaned to the writer by Drs. G. R. Wieland and L. H. MacDaniels. BOTAXICAL GAZETTE, LXVII PLATE XIII BAILEY on BARS OF SANIO BOTANICAL GAZETTE, LXVII PLATE XIV BAILEY on BARS OF SANIO BOTANICAL GAZETTE, LXVII PLATE XV BAILEY on BARS OF SANIO APOSPORY IN PTERIS SULCATA L. W. N. Steil (with plates XVI, XVII, AND FOUR FIGURES) Historical Although apospory was discovered in the mosses by Pringsheim (i8) and Stahl (19) a short time after apogamy had been found in Pteris cretica albo-lineata by Farlow (ii), apospory remained unknown in the Pteridophytes until Druery (6, 7) reported the phenomenon in Athyrium Filix-Joemina var. darissima Jones, Prothalha of this fern were observed to form either from the head or the stalk of the sporangium which was arrested in its develop- ment. The prothallia of aposporous origin produced antheridia and archegonia. Druery (8) also reported the first case of apical apospory, namely in Polystichum angulare var. pulcherrimum Wills. The tips of the leaves of this species of fern produced the game- tophyte as a direct vegetative outgrowth. Bower (2) published a brief summary of his investigations of apospory in Athyrium Filix-foemina var. darissima Jones and Polystichum angulare var. puldierrimum Wills, the material for the investigation having been placed at his disposal by Druery. Sex organs were also observed to develop on the prothallia of the Polystichum angulare variety. The main portion of tl^e paper, however, is not concerned with original studies, but with a discussion of "short cuts" in the life history of the fern. Apospory was also discovered by Bower (3) in Trichomanes pyxidiferum and T. alatum. In the former the gametophyte generation was produced from aborted sporangia ; in the latter there was soral and apical apospory. Sex organs were formed in T. pyxidiferum, but in T. alatum archegonia were absent and antheridia were never developed to maturity. Only, in T. alatum were sporophytes observed to develop from the aposporously produced prothallia, and these sporophytes were of apogamous origin. Bower did not seem to be convinced that apospory and 469 470 BOTANICAL GAZETTE [june apogamy in these cases were not induced by cultural conditions. He believes that if apospory might be expected to occur anywhere in the plant kingdom, it is in the Hymenophyllaceae, the game- tophyte and sporophyte of which are more nearly alike in some respects than in any other homosporous ferns. Bower (4) later attempted to induce apospory in more than 46 species and varieties of ferns by placing the fronds with immature sporangia on moist sphagnum. In not a single instance, however, were gametophytic outgrowths obtained. In regard to the difhculty with which apos- pory is induced, he makes the following statement: "There is a marked disability on the part of ferns to bridge over the limits of the two generations by other means than by the formation of spores; the phenomenon is by no means a promiscuous one occur- ring readily and often, but a rare process, which seems to appear spontaneously under conditions not yet understood and is not readily induced." Farlow (12) found apospory in Pteris aquilina L. Prothallial growths were produced from sporangia which had aborted in their development. No advanced stages in the development of the prothalHa were observed. The first case of apospory from the young sporophyte of a fern was discovered by Druery (9) in a Lastrea variety (probably Lastrea pseudomas var. cristata). Later Druery (id) also induced gametophytes aposporously by placing on moist soil portions of the leaves of Scolopendrium vulgar e var. crispum Drummondae. The prothallia of this species developed both sex organs. Apical apospory was reported by Druery in Athyrium Filix-foemina var. darissima Bolton. He grew sporophytes from the aposporously produced prothallia of Lastrea pseudo-mas var. cristata. These were of apogamic origin and of interest since they possessed characters of both sporophyte and gametophyte. Stansfield (20) brought into contact with the soil portions of the fronds of Athyrium Filix-foemina var. uncoglomeratum and thus induced gametophytes aposporously. When " pinnule ts'' and "leaflets" of young sporophytes, produced by the prothaUia of aposporous origin, were pinned down to the soil, he again obtained readily the gametophyte generation. From the young sporophytes igig] STEIL—PTERIS 47 1 of 4 forms of Polystichum angulare, one of Laslrea, and 3 of Athyrium Filix-foemina, Stansfield in a similar manner induced apospory. He is inclined to believe that apospory may be induced by the same method in many other species of ferns. GoEBEL (14) discovered apospory in Asplenium dimorphum. The ends of a finely divided leaf of a plant of this species produced prothallia, tx-pically heart-shaped, and bearing both archegonia and antheridia. The nuclear history was not investigated, but GoEBEL suggests that reduction theoretically occurs when the prothallia are formed. Goebel (15) successfully induced apospory in Aneitnia Dregeana, Alsophila van Geertii, Ceratopteris thalictroides, Gymnogramme chrysopJiylla, Polypodium aiireum, and Pteris longi- folia. In Marsilia Drummondii and two Adiantum species the results were negative. The primary leaves of young sporophytes were removed and placed on sterihzed loam and peat. From the lamina and petiole of the leaf, thus treated, there were produced gametophytes, sporophytes, or forms intermediate in character, since such outgrowths in some cases bore both antheridia and stomata. The aposporously produced prothalha of Pteris longi- Jolia developed antheridia and archegonia. Sporophytes were not, however, observed to develop from the prothallia. Since there was found no great difference between the nuclei of the two generations, Goebel concludes that there is no sharp line of demarcation between gametophyte and sporophyte. As a result of a series of experiments, Goebel found that young sporophyte tissue possesses greater power of regeneration than old tissue. Contrary to Bower's view that apospory is induced with difficulty and is rare, Goebel is convinced that the phenomenon can be produced readily and is widely distributed in ferns. Considered from the phylo- genetic point of view, Goebel regards the prothallium of the fern as a rudimentary leaf, bearing sexual organs. In a preliminary note on the cytology of apospory, Miss Digby (5) described her results in inducing apospory in Lastrea pseudo-mas var. cristata, the aposporal nature of which first had been reported by Druery (9). As a result of the study of the nuclear condition of the fern, it was reported that 50 chromosomes were retained throughout the life cycle. 472 BOTANICAL GAZETTE [june Farmer and Digby (13) published some very interesting results of their cytological studies of apospory and apogamy in ferns. Five of the 7 species which they investigated produced game- tophytes aposporously. The aposporous nature of the 4 following was observed first by Druery: Athyrium Filix-Joemina var. clarissima Jones, A. Filix-foemina var. clarissima Bolton, Scolo- pendrium vulgare var. crispum Drummondae, and Lastrea pseudo- mas var. cristata apospora. The fifth, Athyrium Filix-foetnina var. uncoglomeratum, was first observed as an aposporous form by Stansfield (20). The origin of the aposporously produced prothalKa in the different ferns was studied with special reference to changes in the chromosome number, and it was discovered that either the haploid or the diploid number is retained throughout the Ufe cycle. In Athyrium Filix-foemina var. clarissima Jones, the embryo formed by the prothaUia of aposporous origin is apogamic. Ninety chromosomes, the diploid number, are found in both genera- tions. The embryo of Lastrea pseudo-mas var. cristata apospora is also of apogamic origin, but the chromosome number, between 60 and 78, is probably the reduced number. The other 3 species were found to be parthenogenetic. In Athyrium Filix-foemina var. clarissima Bolton, 84 chromosomes were counted, in A . Filix-foemina var. uncoglomeratum about 100, and in Scolopendrium vulgare var. crispum Drummondae 70 chromosomes were found in the game- tophyte, and between 80 and 100 in the embryo sporophyte. Since 64 chromosomes were present in the sporophyte of this species, Farmer and Digby were inclined to beheve that the diploid number of chromosomes are present. WoRONiN (12, 24) studied apogamy and apospory in the follow- ing species of ferns: Trichomanes Kraussii, Pellaeaflavens, P. nivea, P. tenera, Notholaena Eckloniana, and N. sinuata. In Trichomanes Kraussii sporangia were not produced, but prothallia were formed in large numbers from portions of leaves which were brought into contact with the soil. According to Woronin, antheridia were in some instances produced directly from the leaves of the sporophyte. It appears, however, that there is not an omission of all the pro- thalHal portion, but that there is formed in each case a short filament which may not be considered as the stalk of an antheridium. iqiq] STEIL—PTERIS 473 Apospory was induced in Pellaea flavens by growing prothallia in continued darkness. When prothallia were transferred to sand cultures there were produced some aposporous growths, but the tendency to apospory was not pronounced. The aposporously produced prothalHa in both species in turn produced apogamous embryos and antheridia. Aposporous prothallia were also induced as a result of regeneration experiments. When a portion of the sinus of the prothalHum with a young embryo was removed and maintained in culture, aposporous prothallia were occasionally formed. When primary leaves were cut off and similarly placed under cultural ct)nditions, prothallia were produced which devel- oped antheridia but no apogamous embryos. Materials and methods A large number of cultures of Pteris sulcata L. were made during the past 3 years by sowing the spores on steriUzed sphagnum saturated with a one-tenth of i per cent Knop's solution or Beyer- inck's solution as modified by Moore (17). The spores for the cultural work were obtained from Dr. A. B. Stout, Bronx Botanical Garden, New York, and Dr. E. B. Copeland, Los Banos, Philippine Islands. An abundant supply of spores was also obtained from a plant grown in the university greenhouse. In one of the cultures made March i, 191 6, the aposporous developments to be described in this paper appeared. Many young apogamously produced sporophytes were found in the culture and on January i, 191 7, some of the young embryos pre- sented a somewhat unusual appearance. As a result of microscop- ical examination it was discovered that prothalloid portions were present in certain parts of the sporophytes. In June 1917 a number of aposporously developed prothallia were found to be present. When some of the prothallia produced by the germination of the spores were transferred to a new culture, made in a similar manner to that of the original one, more aposporous developments were produced. The prothallia and embryos from which the drawings and photo- graphs were made were fixed in chrom-acetic acid solution diluted with 5 parts of water, stained with Haidenhain's iron haematoxylin, 474 BOTANICAL GAZETTE [june and mounted in glycerine jelly. Some of the drawings were made from living material. The sporophyte tissue is represented in the drawings by the darker shading; the gametophyte by the lighter shading. Observations The gametophyte of Pteris sulcata produced by the germination of a spore is in many respects similar to* that of Pteris cretica var. alho-lineata described by Farlow (ii) and DeBary (i). • When the prothallia of both species were grown under the same cultural conditions, it was observed that those of the former were somewhat larger. Archegonia were never found on any of fhe large number of prothalha carefully examined with the microscope. Antheridia occur commonly and frequently in great abundance. The anthero- zoids are apparently normal in every respect. The embryo of Pteris sulcata is always produced apogamously. Such a development of the embryo was first described by the writer (21). Suminski (22) and later Mercklin (16) observed tracheids in the prothalha of this species just posterior to the apical notch. DeBary grew the prothalha of the same species, but observed no sporophytic tissue elements or any other indications of a sporophyte of apogamous origin. Accordingly he reported the fern as non-apogamous. The writer, however, has found the fern to be constantly apogamous under normal cultural conditions. From the sporophytes apogamously produced the gametophytic developments herein described appeared. In every instance in my culture the prothallial portions or prothalha were produced in connection with the lamina or the petiole of the primary leaf. Usually the terminal portion of the leaf was in part or wholly prothalloid. Occasionally forms were observed which were intermediate in character, the cells partaking of the nature of both sporophyte and gametophyte. Sometimes root, lamina, and petiole of the sporophyte were formed (fig. 13). In some instances the root was absent (figs. 6, 8), and in still other cases only a well developed petiole was present (fig. 8) , the lamina of the ordinary leaf being displaced by a prothalloid portion. The vascular system in the primary leaf as a rule was well developed. iqiq] STEIL—PTERIS 475 Prothallia were found, in one instance, growing from both surfaces of the leaf (fig. 5). These closely resembled the prothallia grown from the sp>ore. The largest one of these was ribbon-like, but the others were mere filaments, each consisting of a single row of cells. On the former an antheridium {a) was produced. Rhizoids had been formed from the same prothallium. Two antheridia (a, a') had also been developed by a filament produced from the other side of the leaf. From one of these the antherozoids were discharged when the living prothallia, still attached to the lamina of the leaf, were examined under the microscope. The antherozoids were normal in appearance and actively motile. The prothallial growths were surrounded at their points of origin by normal sporophyte cells. The gametophytic portions, present on the lamina of the primary leaf, may be large, as represented by figs. 6 and 7, in each case of which the terminal portion is distinctly prothalloid. In addition to the large gametophytic portions, a number of smaller regions of the same nature were present on the lamina of the sporophyte represented by fig. 7 {c, d, e,f) and a single one by fig. 6, b. These were almost wholly surrounded by sporophytic cells. A highly magnified view of b is represented by fig. 9. There is, as the figure shows, a sharp line of demarcation between the cells of the two generations. The cells of the gametophytic portion slightly project above the surrounding sporophyte cells. The smaller areas (fig. 7,e,f) were partially surrounded by cells of a somewhat intermediate character (n). The chloroplasts in these cells were less numerous and smaller than those in the neighboring gametophyte cells (fig. 10). Even in the Hving condition this area was almost colorless as compared with the dark green gametophytic areas, and could be readily distinguished with a hand lens. A cell of the same character as those in the paler region just described was found, in one instance, in the larger gametophytic portion represented by fig. 7. This cell was wholly surrounded by ordinary gametophyte cells (fig. i). From the lower part of the petiole, represented by fig. 6, a pro- thallial portion (p) had been formed. A leaflike and sporophytic growth (s) had also been produced in this case. Three projections 476 BOTANICAL GAZETTE [JUNE Fig. I . — Cell of somewhat intermediate character sur- rounded by ordinary gameto- phyte cells; X210. {m, m', m"), of the nature of secondary prothallia, had also been produced from the tenninal gametophytic portion. The lamina of the leaf in some cases was wholly absent, the petiole being, however, well developed and resembling that of an ordinary leaf (fig. 16). From the ventral surface of each of the two large terminal gametophytic portions {b, c) numerous antheridia ia) had been formed. From the dorsal surface of one of the prothallial portions {h) two small secondary pro- thallia {m, m) had begun their develop- ment. The vascular system, which was generally well developed in this portion of the embryo sporophyte, extended for some distance into the terminal gametophytic portion. Between the gametophytic and sporophytic portions the cells were of an intermediate character, as shown in fig. 2, w^hich is a highly magnified portion taken at t. From the petiole of the same sporophyte a prothallial portion (p) had been produced. This outgrowth bears no relation to the vascular system of the petiole. The much modified primary leaf just described was devel- oped in connection with a root and normal secondary leaves. A structure similar to the preceding one is represented by fig. 8. The petiole ap- peared to be well developed but both lamina and root were absent. The terminal portion was also distinctly gameto- phytic in nature and on both surfaces numerous antheridia {a) of different sizes and secondary prothalHa had been produced. An interesting form is represented by fig. 13, since the single large gametophytic portion has been produced in connection with Fig. 2. — Peculiar cells, intermediate in character, of Pteris sulcata; X162. iqiq] STEIL—PTERIS 477 both petiole and lamina of the leaf. The development of the gametophytic portion was observed for several months and during this period grew rapidly. In the meantime the first secondary leaf had been produced. From both surfaces of the prothallium pro- jections appeared, one of which resembled a young sporophyte {s). In this case a root (r) had been produced. The two projections (o) on the same surface of the prothallium (fig. 15) were similar to tho^e appearing on the other surface, which with a small portion of the prothallium are shown highly magnified in fig. 14. While there is no marked differ- ence between the form of the cells of the out- growth and the pro- thallium, those of the latter are much larger. Whether any or all such projections pro- duce embryos was not determined. If the aposporously pro- duced prothallia of Pteris sulcata are like those developed from the germination of a spore, and I am in- clined to think that they are, such growths may produce prothallia, cyUndrical in form, sporophytes normal in every respect, or forms intermediate in character between gametophyte and sporophyte. These were frequently observed to develop from the prothallia in the culture. On the surface of the prothallium, from which the single projection appeared, a number of antheridia had also been formed. The nature of the cells of the two generations is shown in a highly magnified portion of the region between the lamina of the leaf and the prothalloid part id). There is also in this case a sharp line of demarcation between the two generations (fig. 3). Fig. 3. — Cells of Pteris stdcala between gametophyte and sporophyte; X170. 478 BOTANICAL GAZETTE [JUNE A number of forms were found in the culture which were inter- mediate in character, as shown by fig. ii. This growth resembled the leaf of an ordinary sporophyte. Typical epidermal cells including stomata, always present on the lamina of a leaf, were absent in this instance. No structure resembhng a root was present. Fig. 4 shows the terminal portion of the lamina-like part. An effort was made to study the nature of the development of the prothalloid portions from the earliest stages. The earHest stage found was one composed of 4 or 5 cells, formed in connec- tion with the petiole of the primary leaf. For several days the growth was followed and during that time a number of cells had been produced, as shown in fig. 1 2 . When the earliest stages were observed the leaf had already emerged from the pro thallium, and distinctly gameto- phytic cells could then be seen. Since the number of instances of apospory appearing in the culture was not large, as compared with the number of sporophytes, a favorable opportunity to study the stages in sec- tions was not offered. It is certain that such forms represented by figs. 8 and 16 were never wholly sporophytic and later became gametophytic. From such an instance as shown in fig. 5, how- ever, it could readily be conceived that the prothallia developed from the primary leaf which was distinctly sporophytic in character. Fig. 4. — Terminal portion of leaflike structure of Pkris sulcata; X322. iQiq] STEIL—PTERIS 47g It seems not improbable that the game tophy tic cells were present at the earliest stages in the development of the embryo. Since the sporophyte of Pteris sulcata is of apogamous origin, there is an intimate connection between the cells of the gametophyte and the sporophyte. These cells may be carried upward by the sporo- phyte, and, retaining their power to divide, they may give rise to the gametophytic portions which have been described. It has not so far been possible to state the conditions under which apospory occurred in the culture of the prothalHa of Pteris sulcata. All attempts to induce the phenomenon have failed. Young sporophytes grown in subdued light produced no game- tophytes aposporously. Portions of the leaves of young and old sporophytes when placed on moistened sphagnum also failed to develop prothalHa. The nuclear history of Pteris sulcata was not followed. It is very probable, from studies so far made, that there is no change in the chromosome number when the apogamous embryo originates. It is also likely that when the gametophyte is formed in connection with the embryo sporophyte there is no change in the chromosome number. On account of the limited number of aposporous develop- ments in the culture no favorable opportunity was presented to count the chromosomes at a point in the hfe history when the gametophyte originates. It is believed, however, that the game- tophyte thus produced and one formed by the germination of the spore have the same number of chromosomes. The changes which are involved in the formation of an embryo of apogamous origin, except in the two Lastrea pseudo-mas varieties described by Farmer and Digby, are unknown. In these ferns, according to their description, fusion of adjacent prothallial cells and their nuclei initiate the formation of the embryo with the diploid number of chromosomes. It is certain from studies already made that such changes are not involved in any of the apogamous species which I have had an opportunity so far to investigate. Until the exact nature of the changes which are involved when the apogamous embryo originates are known, however, the origin of the aposporous developments in Pteris sulcata cannot be explained in a satisfactory manner. 48o BOTANICAL GAZETTE [june Summary 1. The gametophyte generation of Pteris sulcata L. is ordinarily produced by the germination of a spore. 2. The embryo sporophyte is of apogamous origin. 3. The gametophyte generation of Pteris sulcata under certain conditions was produced aposporously. 4. The gametophytic portions or gametophytes were formed in connection with the lamina or the petiole of the primary leaf. In one instance a prothallium was produced from both lamina and petiole of the primary leaf. A sharp line of demarcation usually exists between the cells of the gametophyte and the sporophyte. • 5. The prothallial portions developed antheridia, secondary prothallia, and in one instance a sporophyte-Hke outgrowth. 6. The antherozoids, produced by the aposporously developed prothallia, were actively motile and normal" in appearance. 7. Occasionally forms intermediate in character between game- tophyte and sporophyte were formed. 8. It seems probable that the origin of the aposporously pro- duced gametophyte may be traced to an early stage in the development of the embryo. Since the embryo, on account of its apogamous origin, is intimately connected with the prothallium, it is not impossible that in some way cells of the prothallium may be embodied in the developing embryo. These cells, retaining the power to divide, may produce such outgrowths as have been described. LITERATURE CITED 1. Bar"^, a. de, tjber Apogamie Fame und die Erscheinung der Apogamie im Allgemeinen. Bot. Zeit. 36:449-487. pi. 14. 1878. 2. Bower, F. O., On apospory in ferns. Jour. Linn. Soc. 21:360-368. pis. 10, II, figs. 5. 1884. 3. , On some normal and abnormal developments of the oophyte of Trichomanes. Ann. Botany 1:269-305. pis. 14-16. 1888. 4. , Attempts to induce aposporous developments in ferns. Ann. Botany 4:168-169. 1889. 5. DiGBY, L., On the cytology of apogamy and apospory. II. Preliminary note on apospory. Proc. Roy. Soc. 463-468. 7?g-J- J- 1905. 6. Druery, C. T., Observations on a singular development in the lady-fern. Jour. Linn. Soc. 21:354-360. 1884. 19 ig] STEIL—PTERIS 481 7. Druery, T. C, Further notes on a singular mode of reproduction in Athyrium Fiiix-foemina var. clarissima. Jour. Linn. Soc. 21:358-360. jigs. 2. 1884. 8. , On a new instance of apospory in Polystichutn angulare var. pul- cherrimum Wills. Jour. Linn. Soc. 22:437-441. fig. i. 1886. 9' , Notes upon an aposporous Lastrea {Nephrodium). Jour. Linn. Soc. 29:479-483. pi. 34. 1892. 10. , Notes upon apospory in a form of Scolopendrium vidgare var. crispiim and a new aposporous Athyrium, also an additional phase of aposporous development in Lastrea pseudo-mas var. cristata. Jour. Linn. Soc. 30:281-284. pi. 17. 1893. 11. Farlow, W. G., An asexual growth from the prothallus of Pteris cretica. Quar. Jour. Micr. Sci. 14:226-273. pis. 10, 11. 1874. 12. , Apospory in Pteris aquilina. Ann. Botany 2:383-386. 1889. 13. Farmer, J. B., and Digby, L., Studies in apogamy and apospory in ferns. Ann. Botany 21:161-199. pl^- i6~2o. 1907. 14. GoEBEL, K., Kleiner Mitteilungen 3. Aposporie bei Asplenium dimor- phum. Flora 95 : 239-244. 1905. 15' , Experimentell-morphologische Mitteilungen. I. Kiinsthch hervor- gerufeue Aposporie bei Farnen. Sitz. Konigl. Bayer. Akad. Wiss. 37: 119-137. 1907. 16. Mercklin, C. E., Beobachtungen an Prothallium der Farnkrauter. St, Petersburg. 1850. 17. Moore, G. T., Methods for growing pure cultures of algae. Jour. Appl. Microscopy 6:2309-2314. 1903. 18. Pringsheim, N., Vegetative Sprossung der Moosfriichte. Monatsbericht der koniglichen Akademie der Wissenschaften zu Berlin. 425-429. 1876. 19. Stahl, E., tJber kiinsthch hervorgerufene Protonemabildung an dem Sporogonium der Laubmoose. Bot. Zeit. 34:689-695. 1876. 20. Stansfield, F. W., On the production of apospory by environment in Athyrium Filix-foemina var. uncoglomeratum. Jour. Linn. Soc. 34:262- 268. 1899. 21. Steil, W. N., Studies in some new cases of apogamy in ferns. Bull. Torr. Bot. Club 45:93-108. pis. 4, 5. 1918. 22. Suminski-Leszczyc, Zur Entwickelungsgeschichte der Farnkrauter. Berlin. 1848. 23. WoRONiN, H., Apogamie und Aposporie bei einigen Farnen. Ber. Deutsch Bot. Gesells. 25:85, 86. 1907. 24. , x\pogamie und Aposporie bei einigen Farnen. Flora 98:101-162. 1908. 482 BOTANICAL GAZETTE [june EXPLANATION OF PLATES XVI, XVII All of the drawings were made with the aid of a camera lucida. Figs. 5, 6, 7, 8, II, 13, and 16 represent a magnification of about 30 times. AH the other figures, with the exception of fig. 10, were drawn with a magnification of about 325. Fig. 10 represents a slightly greater magnification. The drawings were reduced one-half in reproduction. PLATE XVI Fig. 5. — Lamina of primary leaf of sporophyte of Pteris sulcata from both surfaces of which a number of prothallia appear as outgrowths of leaf blade; a, a', antheridia. Fig. 6. — Sporophyte of Pteris sulcata with large terminal prothallial portion formed in connection with primary leaf; m, m', and m", young secondary prothallia; 6 and ^, prothallial portions; .y, leaflike outgrowth. Fig. 7.— Another sporophyte with large terminal prothallial portion; c, d, e, and /smaller prothallial portions; n, "light" area, cells being neither characteristically sporophyte nor gametophyte. Fig. 8. — Sporophyte with well developed petiole of primary leaf; lamina displaced by prothalloid portion ; />, old prothallium; o, antheridium. Fig. 9.— Highly magnified portion of gametophytic region (b) and neigh- boring sporophyte cells shown in fig. 2. Fig. 10. — Highly magnified portion of n and neighboring cells shown in fig. 3 ; chloroplasts are less numerous and smaller in paler region than in pro- thallial cells; cells of paler region are somewhat intermediate in form, partaking of nature of both generations. Fig. II. — ^Leaflike portion; epidermal cells of lamina not typically sporo- phyte, being regular in form and lacking stomata. Fig. 12. — Early stage in development of prothallium (p) in connection with petiole of primary leaf. PLATE XVII Fig. 13. — Sporophyte with prothallium produced from both lamina and petiole of primary leaf; 0, outgrowth from one surface of prothallium. Fig. 14. — Highly magnified view of outgrowth (0) shown in fig. 9; cells in outgrowth much smaller than those of prothallium. Fig. 15. — Three outgrowths {0 and s) on other surface of prothaUium represented in fig. 9; r, rootUke portion of outgrowth (s). Fig. 16.— Primary leaf of sporophyte; b and c, prothallial portions; m, m' , secondary prothallia developed from dorsal surface of b; p, prothalliuiil developed from petiole; a, antheridium. BOTANICAL GAZETTE, LXVII PLATE XVI STEIL on PTERIS BOTANICAL GAZETTE, LXVII PLATE XVII STEIL on PTERIS HYDROGEN CYANIDE FUMIGATION CONTRIBUTIONS FROM THE HULL BOTANICAL LABORATORY 249 E. E. Clayton (with TWO figures) The object of this work was to secure data on the manner in which green plants are affected by exposure to hydrocyanic acid; with particular emphasis on the resistance of the plants to this gas and the modification of this resistance by various factors, external and internal. A number of articles have been published concerning the effect of cyanide on animals. More recently its action as an enzyme paralyzer has been brought forward prominently. Information as to the action of cyanide on plants is of scientific interest, and certainly of practical value, for hydrocyanic acid finds important use as an insecticide in orchard and greenhouse practice. Circumstances have forced the discontinuance of this work, with many phases of it incomplete, but enough facts have been estab- Hshed, and enough new lines suggested, to warrant its pubh'cation in this incomplete form. Historical Literature bearing on this problem is not abundant. The Department of Agriculture and several of the experiment stations, notably the CaUfornia station, have pubHshed a number of bulletins dealing with fumigation as a commercial process; but the work done is of a kind which assists little in answering the fundamental physio- logical questions involved. The action of cyanide on the animal and in connection with various chemical processes has been thoroughly investigated, and from a consideration of these data we can gain much. ScHONBEix (8) first called attention to the inhibitory effects of hydrocyanic acid. He worked with the leaves of plants, and also with animal blood, and found that the presence of the acid prevented each of these materials from decomposing hydrogen peroxide. He concluded that the extremely poisonous action of cyanide on the 483 484 BOTANICAL GAZETTE [june animal was due to an incapacitation of the red blood corpuscles, which caused suffocation, a conclusion still generally accepted. Geppert (3), and still later Vernon (id), who worked with animals, found this same inhibitory effect of hydrocyanic acid on respiration. They showed that if a lethal dose has not been given the organism recovers completely; that is, if the organism is not killed it is not injured in any way. Schroeder (9), using the fungus Aspergillus niger, made a long series of determinations on the effect of potassium cyanide on respiration. He confirmed the previous work and emphasized the fact that, if cells were not exposed too long, recovery was complete. He also did some work with ether, and decided that the inhibition of respiration caused by treatment with this anesthetic was quite different in character from inhibition caused by treatment with cyanide. He characterized diminution of respira- tory rate through treatment with ether as a secondary effect, brought about by previous death of the tissue; while diminution of respiratory rate through treatment with cyanide is a primary effect, the cyanide directly inhibiting respiration and killing the tissue. Mathews (5) has offered strong arguments favoring a contrary view, namely, that hydrocyanic acid and anesthetics both act primarily on the respiratory processes, each affecting these in exactly the same manner. Another activity of cyanide, which has come into prominence in recent years, is its role as an "enzyme paralyzer '' and more specifi- cally its abihty to check certain chemical reactions carried on in the laboratory. Bredig and Ikeda (i), also Loevenhart and Kastle (4), have found that in the catalytic decomposition of hydrogen peroxide the addition of a very small amount of potassium cyanide completely stopped the reaction. They concluded that the cya- nide was acting on the catalytic agent. Mathews and Walker (6), who worked on the spontaneous oxidation of a very reactive amino acid, cystein, found that a very small amount of potassium cyanide checked this action. Their statement of the probable explanation of this is interesting, since the same reasoning is equally appKcable to other inhibitions of oxidation processes by cyanide. "The proportion of cystein molecules in a condition to be oxidized igig] CLAVTON— FUMIGATION 485 at any given time is extremely small ; while the proportion of active potassium cyanide molecules is large. The number of active oxygen atoms is also small. If we further assume that the cyanide unites with the cystein at the same point where the oxygen ordinarily combined, then the results obtained are easily understood." In plant respiration, with cyanide present, we have the same general condition; that is, the oxygen of the air is not able to oxidize the plant compounds, and we may suppose that the cyanide has acted in the same way. In the case of respiration, of course, we may have a chain of several oxidations and reductions, and the exact point at which the cyanide intervenes is not known. Material With the data at hand, while there is much on which to base inferences, there is nothing which tells of the behavior of an auto- troph and the factors influencing this. Consequently, since this was pioneer work, it seemed highly desirable that it be done with normal green plants under normal conditions. The tomato was selected because it is easy to grow, sensitive to cyanide, and a plant commonly requiring fumigation in the greenhouse. Work under conditions such as those indicated is made difficult by the presence of factors not under control, which vary conditions with the different periods of time. In consequence of this it seemed best, in the case of most of the material experimented upon, not to try to carry on the work to its final expression, but merely to open up the way. The writer hopes at a later date to complete the work upon some of the more important phases with conditions carefully controlled. Responses The response of the plant to varying concentrations of hydro- cyanic acid was observed as indicated by the subsequent growth of the plant. Two main points were considered in deciding the effect of a given concentration of gas: (i) the growth rate of the plants after fumigation; (2) the appearance and growth character of the plants after fumigation. The method was to select 7 or 8 groups of plants, similar in all respects, and to fumigate these on successive nights. Suitable checks were kept, and the growth of the treated 486 BOTANICAL GAZETTE [j0NE plants was compared with the growth of the check plants during the same days. Many series, made up as indicated, were run. the experiments covering a period of 5 months. It is obvious that the slow growing plants of January are not wholly comparable with the rapid growing plants of April. This matter cannot be controlled, TABLE I Average increase or decrease in growth during 12 days FOLLOWING fumigation KCN per cubic feet Plants measured Percentage I St to 4th days Percentage Sth to Sth days Percentage 9th to 1 2th days O.OOI gm 0 . 002 gm 0.003 gm 0.004 gm 0.005 gm 0 . 006 gm 0.007 gm 0 . 008 gm 6 10 20 19 25 30 20 16 -16 -19-5 - 5-5 - 4.2 - 5-7 + 17-7 +33 + 13 - 1-5 — 10. 2 + 21 + 11. 4 + 16 + 7 + 15-3 - 8 -"■5 + 9-3 - 7 + 11. 6 + 13 + 6.5 -18 TABLE II Percentage of 4 day periods showing decreased growth rate ant) also percentage showing increased growth rate KCN Percentage showing rate decrease Percentage showing rate increase 0 001 gm 66.6 83.6 63s 50.0 33-3 33-3 16.6 33-3 0 002 gm 0 0 oo"? firm 36.4 0 OOA 2m 41.6 0 00? sm. 66.6 0 006 arm 55.5 0 007 em 750 and introduces an element of error. The time of exposure, length of exposure, temperature, and moisture conditions were similar in all cases. Using the growth of the check lots as 100 per cent, the loss or gain of the treated lots was computed, and from these values the averages shown in tables I and II were secured. Consideration of these data shows that with concentrations of cyanide up to o .004 gm. per cubic foot of air space, the effects were I9I9] CLA YTON—FUMIGA TION 487 somewhat detrimental to the growth of the plant. With o . 004, stimulation and retardation effects seem to be about balanced ; but at 0.005, 3.nd continuing up through 0.006 and 0.007, stimulation eiTects plainly dominate. The plants giving these responses retained after the treatment all appearances of normal growth and metabolism. With the concentration of 0.008 gm. KCN the first injury appeared, although there was some variation in this regard. Thus in one series no injury appeared before o.oigm. KCN. The harmful effect of this injury was very marked. Growth records were taken for concentrations up to 0.016, but unfortunately, with these damaging strengths (0.008-0.016), growth records do not always indicate the true condition of the plants. False stimulation TABLE III Percentage increase or decrease in growth following fumigation KCN per cubic foot Percentage I St to 4th day Percentage 5th to 8th day Percentage gth to 12th day Percentage 13th to i6th day 0.007 gm.: no injury o 008 gm.: injury to parts of plants o 009 gm. : injury to all plants o.oi gm 0.012 gm + 14 +36 -37 -27 -16 +32 +44 + 2 -29 -40 -33 -58 -63 -66 — 12 -16 -44 -58 was frequently shown. This is an increase in growth rate of plants ob\dously in poor condition. The explanation of this false stimula- tion is that the growing tips of plants are rarely killed, and in con- sequence of the loss of considerable leaf area through injury these growing tips are forced into renewed activity. Thus badly injured plants in the course of a month's time form entire new tops, although old injured leaves drop off, for the most part. Usually, however, when injury occurred the depressing effects were so strong that this increased growth rate was very transient in character, or did not show at ail, as is seen in table III. A characteristic feature of this injury is the duration of the depressive effects. The plants were still far below normal at the 488 BOTANICAL GAZETTE [juxe end of the 12 or i6 day period recorded. The following results were secured: f Gram kcn per cubic foot of air space o.ooi 0.002, temporary depressive effects 0.003 0.004, intermediate effects 0.005 0.006, temporar>' stimulative effects 0.007 0.008, relatively permanent depressive effects o . 009 o.oi It is interesting to note that the greenhouse white fly (Aleyroides) was killed at a concentration of 0.006 gm. KCN. Another point worthy of emphasis is that the actual difference in amount between a concentration of cyanide which brings about noticeable injury to the plants, and one which does no harm, may be very small indeed. This compares very well with previous statements (3, 9) that cyanide either killed or that recovery was complete. There is not much work which can be cited in substantiation of the mixed stimulative and depressive effects found. In the work with hydrogen peroxide (4) very low concentrations of cyanide accelerated the reaction. Again it was discovered (7) that cyanide hastened the oxidation of an amino acid when the amino acid was present in an impure state. With these two exceptions the cyanide hterature bearing on this work deals with inhibitory actions. Factors EXTERNAL FACTORS Moisture.— In considering moisture as an external factor we deal only with the effect of free water on the leaf surface of the plants fumigated. It is common knowledge among greenhouse men that if plants go into the fumigation wet, considerable injury is often induced. In testing this I have found that some species are made more susceptible to injury by wetting the leaves, while other species are not visibly affected. The tomato is in this latter class. On closer observation it was found, in the case of the tomato, that 1 9 1 9] CLA YTON—FUMIGA TION 489 not only did the free water have no detrimental action, but actually, under certain conditions, gave beneficial results. This was true in the cases where the strength of fumigation was not sufficient to injure the plants (below 0.008 gm. KCN): Growth during 12 days following fumigation Dry plants Wet plants Lot I : . . . ICX3 per cent 127 per cent Lot 2 100 per cent 115 per cent Lot 3 100 per cent 139 per cent With the following data, however, the fumigation was strong enough to injure the plants: Dry plants Wet plants Lot I 100 per cent 115 per cent Lot 2 100 per cent qq per cent Lot 3 100 per cent 100 per cent Thus wetting the leaves had a beneficial eft'ect if the fumigation was not strong enough to cause injury. With the appearance of injury this beneficial action ordinarily disappeared. The results just given are from experiments performed at wide intervals, and the plants used varied greatly as to size and age. Moisture present in pans as free water surfaces or saturated soil gave negative results. Temperature. — The effect of temperature on plant resistance was determined quahtatively only. Excessive temperatures during the period of fumigation very materially increase the amount of injury. A variation of 20° F. has a marked effect; thus plants fumigated at 75° F. suffered much less injury than others which w^ere fumigated at 95°. In what way temperature affects the plant, to cause these changes, has not been determined. One possible explanation is that the effect is through changes in per- meability of protoplasm, such as Rysselberghe, Lepeschkin, and EcKERSON have shown to occur, or even in permeability of wall structures. That it may be due to a change in the rate of chemical reaction is of course possible. Light. — Light undoubtedly exerts a direct action on plant resistance, but the conditions did not permit accurate observation on this point. Here permeability changes as well as catalytic effects of Light appear possible. The indirect action of Light as a regulator 490 BOTANICAL GAZETTE [june of stomatal aperture and in connection with photosynthetic activity will be discussed later. PLANT FACTORS Protective membranes, — Most plants which are highly resistant to cyanide are characterized by having thickly cutinized or suberized epidermal membranes which serve as a protection. As further evidence of the protective power of these, Tradescantia zebrina is made perceptibly more resistant by coating the upper surfaces of the leaves with Blackman's wax. The Tradescantia leaf has no stomata on the upper surface, and the reduction in injury must be due to the wax covering, thus making the thin epidermis relatively impervious. This increased resistance, when unobscured by a large amount of stomatal activity, is very marked. It is by no means possible to explain all differences in resistance on the basis of protective membranes, however. The radish endures without injury 3 times the strength of fumigation which a tomato endures, yet microchemical examination reveals but little difference in the cuticular development. Stomata.^ — The stomata seem to be the most important single factor in determining the amount of injury resulting from hydro- cyanic acid fumigation. To ascertain the extent of their influence experiments were conducted in the following manner. Fumigations were run at various times of day and night, using like strengths of cyanide. Each lot of plants exposed included tomatoes and Tradescantia zebrina, the leaves of the latter being painted in various ways with Blackman's wax. After the beginning of each fumiga- tion samples of epidermis were taken from several species with large stomata {Geranium and Tradescantia), and the amount of stomatal opening determined under the microscope. Fumigations conducted on a very dark rainy day (0.02 gm. kcn PER CUBIC foot) Exposure 1 : 30 TO 3 : 30 P.M. Tomatoes Badly injured Tradescantia Leaves under surface coated . . Uninjured (stomatal surface closed) Leaves upper surface coated . . Badly injured (stomatal surface open) Leaves untreated Killed (stomatal surface open) Average stomatal opening i : 30 p.m., 3.5/1 1919] CLAYTON— FUMIGATION 491 Exposure 5:30 to 7:30 p.m. Tomatoes Slightly injured Tradescantia Leaves under surface coated . . Uninjured (stomatal surface closed) Leaves upper surface coated . . Uninjured (stomatal surface open) Leaves untreated. Slight injury (stomatal surface open) Average stomatal opening 5:30 p.m., almost all stomata were closed. Exposure 11:30 p.m. to i 130 a.m. Tomatoes Occasional slight injury Tradescantia Leaves under surface coated. .No injury (stomatal surface closed) Leaves upper surface coated . . No injury (stomatal surface open) Leaves untreated No injury (stomatal surface open) Average stomatal opening, none. It was evident that the closing of the stomata greatly increased the resistance to fumigation. Approximately speaking, if 100 per cent were to represent the injury at 1:30 P.M., by 5:30 P.M. it had dropped to 10 per cent, and by 11 : 30 to 2 per cent. Why the plants were most resistant at the very late period was not entirely apparent, the stomata at 5 : 30 p.m. being, with very few exceptions, as tightly closed as at 1 1 : 30 p.m. Miss Eckerson, however, has found that stomata in certain portions of some leaves lag behind as to closing time. During sunny weather there are great variations of light and temperature in the course of a day and night. This is in contrast with conditions during dark weather. In a sunny period, however, considerable work was done between 6:30 P.M. and midnight. It was found that stomatal activity, as indicated by injury to the hypostomatous Tradescantia, continued on a gradually diminish- ing scale till 8 : 45 p.m. (about 2 hours after sundown). The stomata at this time were found to be almost closed. A noticeable fact was that at the later periods the correspondence between the amount of stomatal opening and the injury was not absolute. The injury at 8 : 00 p.m. was not so great as would have been expected from the size of the openings. In sunny weather, as in dark, there was a gradual increase in resistance of the plants during the night. The maximum of resistance was reached during the period between 1 1 : 30 P.M. and i : 00 a.m. 492 BOTANICAL GAZETTE [june In conclusion it may be said that the amount of injury follows the stomatal movement rather closely. A fumigation at 5 : 30 p.m. on a dark day was about equivalent in injury to one at 8:30 p.m. on a bright day. CHEMICAL FACTORS Water. — It was early observed that there were rather wide differences in the resistance of tomato plants grown under various conditions. It was found that when plants grew rather slowly, with a high chlorophyll content per unit area, they were very resistant to the hydrocyanic acid. Plants growing rapidly, with a low chlorophyll content per unit area, were very susceptible to injury from the hydrocyanic acid. Thus, judging solely by intensity of color, it was possible to select from a large group of plants 2 lots differing widely in their ability to withstand injury. Variations in water supply seemed to be the underlying cause of these differences, although other conditions will produce similar characters. To test this the following experiment was conducted : Twenty-four plants in vigorous growing condition were selected and divided into 2 lots. Lot I was watered only enough to keep growing, while lot 2 was watered abundantly. After 10 days, plants from lot i, now dark green, were fumigated and found very resistant. Plants from lot 2, which were light green, were very easily injured by fumigation. An exhaustive chemical analysis of these plants was not made, but preliminary tests revealed one significant fact: the resistant plants (lot i) had a greatly increased carbohydrate content. The reducing sugars claimed attention as being the most reactive of these sub- stances, and also the ones most concerned in plant respiration. Determinations of the reducing sugar content of the leaves of lots I and 2 gave the following results (the samples were taken at 5:15 P.M.): lot I, non-resistant leaves, 0.108 per cent calculated as dextrose per unit weight of green tissue; lot 2, resistant leaves, 0.57 per cent dextrose per unit weight of green tissue. Thus the resistant plants had much more reducing sugar. The actual amount is not large in either case, but the relative difference is great, lot 2 being 5 times as rich in reducing sugars as lot i. The dry weight of the plants of lot 2 averaged little more than i . 5 times the dry weight of lot i. 1 9 1 9] CLA YTON—FUMIGA TION 493 A possible role of reducing sugars as determiners of resistance was further tested by arbitrarily changing the sugar content of plants and observing the effects on resistance. (i) Plants were placed in a dark box for 48-72 hours. This treatment brings the reducing sugars in the leaves practically to zero at the end of these periods. These plants, when fumigated, were found to be very easily injured. (2) Plants treated exactly the same were taken from the dark box 12 hours before the fumigation and infiltrated with a glucose solution. They were returned immediately after the infiltration to the dark box, the exposure to light not exceeding 20 minutes. Plants thus made rich in glucose were highly resistant to cyanide injury. The following are extracts from data collected on this point : Fumigation with 0.012 gm. kcn per cubic foot, infiltrated at 9:30 a.m. at 6 cm. mercury pressure Dark box plants Checks Bad injury Infiltrated plants 0 . 25 per cent glucose Slight injury o. 50 per cent glucose No injury 1 . 00 per cent glucose No injury (from fumigation) Fumigation with o. 14 gm. kcn, infiltrated at 10:00 a.m. at 5 cm. MERCURY pressure Dark box plants Checks Very great injury Infiltrated plants 0.90 per cent glucose No injury (from fumigation) o. 75 per cent glucose No injury o . 60 per cent glucose Very slight injury With concentrations of glucose o . 9 per cent or above there was some injury to the plants from plasmolysis; such injury was distinct from that due to cyanide and caused no confusion. These experi- ments were repeated many times with uniformly decisive results. Frequently the check plants (not infiltrated) had their entire tops killed, while the treated plants showed little or no injury. With these data in hand it is possible to say that for plants to be "normally" resistant they must have a fair content of reducing 494 BOTANICAL GAZETTE [june sugar. Meyer has made determinations of the amounts of car- bohydrates during the course of the day and night. Working with the leaves of Tropaeolum, he found that reducing sugars were at a minimum during the day but started to increase with nightfall. Analyses of tomato leaves yielded similar results ; when the weather was bright the maximum reducing sugar content was found at 1:00-2:00 A.M. During dark weather this gradation was found to be much disturbed, as would be expected. The following figures were secured from the tomato leaves during such a period: Reducing sugars; weather very dark and rainy 1:30 P.M., 0.414 per cent, calculated as dextrose (net weight) 5:30 P.M., 0.335 per cent, calculated as dextrose 11:30 P.M., 0.285 per cent, calculated as dextrose Presence of other factors made it impossible to ascertain what effect these variations had on resistance. There seems to be little doubt that the glucose in the plant acts as a protective agent 'against injury by cyanide. Considering glucose to have a direct effect, there are several possibihties concerning the manner of this action, (i) It may protect the plant by supplying an excess of molecules to unite with the cyam'de entering. Cyanide does unite readily with glucose. (2) There is much evidence in physiological experimentation, with both plants and animals, showing that an excess of glucose present will temporarily take the place of missing oxygen. Asphyxiated animals produce glucose in excess amounts, other compounds being broken down. Plants, in absence of oxygen, behave normally for a time when glucose is supphed. It may very readily be that the protective action of the glucose is an indirect one, working through other channels. Thus it may possibly modify stomatal action. What goes to make up a resistant plant, and under what con- ditions is it most resistant ? We have given but Httle attention to the first of these questions ; hence the broad problem of why certain species of plants are much more resistant to cyanide than others will be left without attempting an answer. Rather exhaustive comparative studies seem to be the only possible way of solving this. iqiq] CLAYTON— fumigation 495 Concerning the conditions under which plants are most resistant, the following statements are possible : The resistance of the tomato to cyanide is increased by the presence of water on the leaf surface, but with some species wet leaves increase injury. Temperature should be moderately low. Light intensity should be low during the day preceding the fumiga- tion. The plants should go into the fumigation with the stomata closed. A large amount of reducing sugars in the plant is correlated with maximum resistance. Checking growth by reducing water supply, during the period preceding the fumigation, increased resistance. The action of cyanide on entering the plant is a very interesting question, and all evidence favors the conclusion that cyanide combines readily with the substances there. In certain species of plants cyanides in the form of glucosides occur, and on these naturally occurring cyanides a great deal of work has been done. The investigators (12, 13) are unanimous in considering that this cyanide is never in an uncombined state. Granting then that the hydrocyanic acid, on entering the plant, unites quickly with plant compounds, there may still be variations in the type of union. Thus possibly there may be adsorption under certain conditions and under other conditions chemical combination. Methods for the determination of cyanide in plants have been devised (2, 11), but tKey are rather laborious. One method is based on the fact that a cyanide in a picric acid-sodium carbonate solu-' tion gives a red color. The depth of the color, as compared with a standard range, is the basis of estimation. The coloration is due to the strong reducing action of the cyanide, but unfortunately the cyanide is not the only substance present in the plant which gives the reaction. For accurate quantitative work it is necessary to isolate the cyanide compounds. This was not done at this time; instead, estimations were made on unit weights of leaf tissue. The leaves were put in the picric acid-sodium carbonate solution, and observations were made after 24 hours by means of a colorimeter. The phrase ''reducing substance content" is used because, as already stated, the cyanides are not the only things reacting. The variation in values secured, during the course of a series of 496 BOTANICAL GAZETTE [june fumigations, is very interesting. In the data T represents an empirical measure. Fumigation strength in cm. Reducing substance KCN per cubic foot CONTENT o . 003 T7.5 o . 004 T 8 o . 005 T 8 o . 006 T8.S 0.007 ^ 0.008 point of first injury T 4 o . 009 T 3.5 o.oio ^35 0.012 T2.S 0.016. T 2 0.020 ^ ^ ■ 5 The sudden drop at the point of first injury was very noticeable; thus from T 9 to a value less than half. Considering this drop as an index of chemical changes, it agrees well with the results of the growth experiments recorded at the beginning of the paper. Every indication was found, also, that the action of the cyanide, at a concentration just below the point of external injury, was radically different from its action at and beyond this point. There are two main types of injury resulting from cyanide fumigation: (i) the killing of definite areas of leaf and, much more rarely, stem tissue. This injury is always locaHzed on the younger portions of the plant. (2) Injury in the form of an epinastic response. This is frequently found in cases where the fumigation was just a trifle too strong. As a rule, under these circumstances, there is no apparent injury to the plants for a period of 5-7 days after the fumigation, and then a twisting of certain leaves (epinasty) becomes apparent. This kind of injury is quite distinct from the distortion which arises through an excessive kilhng of tissue. The drawings in fig. i were secured after fumigating, with a damaging strength, a house containing tomato plants of different ages. The movement of the injury from the inner portions of the leaf in the old plants to the outer portions of the leaf in young plants was noticeable. The drawings show all the injured leaves from representative plants. Even with the very large plants the injury did not extend below the third or fourth leaf from the tip. iqiq] CLA YTON—FUMIGA TION 497 Fig. I. — Injury from cyanide fumigation to tomato plants ranging in age from seedlings to mature plants (a, b, c, d); the injury is shown by the tlack areas, and it will be noted that the leaf tissues killed in the small seedlings (a) are peripheral, and also more or less apical. 498 BOTANICAL GAZETTE [JUNE There are several possible explanations of this localization on the younger portions. The epidermal membranes are not so well developed in the young leaves and it is easier for the cyanide to enter. The younger and more rapidly growing portions of a plant are the first to suffer in absence of oxygen. It may be that the very reactive state of the protoplasm in these parts makes them less resistant. Fig. 2.— Older plants, in which there is a steady change in the location of the injured areas, until with old bearing plants the killed area is basal, and is not limited to the margin of the leaf. The rapidity with which cyanide can kill the plant tissue is another point of interest. The plant can live in the total absence of atmospheric oxygen for some time. However, 5 hours after the beginning of a damaging fumigation it was possible to see the dead leaf areas, dark and water-soaked in appearance. It may be said that there are several things which stand out prominently as causes for the unsatisfactory results often secured from greenhouse fumigation with cyanide. First, there is a lack of 1919] CLAYTON— FUMIGATION 499 appreciation of the necessity for delicate regulation of amounts of cyanide. Every separate greenhouse varies in its ability to retain the gas, and it is not alone the amount of cyanide which is put in that counts, but rather the amount which is retained. Thus definite recommendations cannot be given more than it is possible to give a universal fertilizer formula. Second, it is the usual thing to start the fumigation during the latter part of the afternoon when the stomata are still open. This is sure to induce excessive injury. It is safe to start fumigation 2 . 5-3 hours after sundown on a bright day, or at sunset on a very dark day.. Summary 1. Different concentrations of hydrocyanic acid gas gave effects ranging from stimulative to depressive. The maximurn of bene- ficial results was secured from concentrations deadly to insect life, but just a little below the point of first injury to the plant. 2. External factors having important action on the resistance are as follows : (a) wetting the leaves had a beneficial effect on the tomato; {b) reduced temperature and low light intensity during the day preceding fumigation increased resistance. 3. Injury closely paralleled the stomatal movement, increasing as the size of stomatal aperture increased. 4. A higher or lower water supply in the soil affected resistance, through hastening or retarding the growth rate. Rapid growing plants were susceptible to injury, while slow growing plants were more resistant. 5. High reducing sugar content seemed to be correlated with maximum resistance. I wish to acknowledge the assistance given by Dr. William Crocker and Dr. Sophia H. Eckerson, under whose direction this work was conducted. University of Wisconsin Madison, Wis. 500 BOTANICAL GAZETTE [june LITERATURE CITED 1. Bredig, G., and Ikeda, K. I., Anorganische Fermente II. Zeitschr. Physikal. Chem. 37:8-16. 1901. 2. Chapman, A. C, The colorimetric estimation of hydrogen cyanide. Analyst 35:469-477. 1910. 3. Geppert, J., tJber das Wesen der Blausaurevergiftung. Zeitschr. Klin. Medicin 15:208-307. pi. i. 1889. 4. LoEVENHART, A. S., and Kastle, J. H., On the catalytic decomposition of hydrogen peroxide and the mechanism of induced oxidations. Amer. Chem. Jour. 29:397-436. 1903. 5. Mathews, A. P., The action of ether on an anaerobic animal tissue. Jour. Pharmacol, and Exper. Therapeutics 2:231-238. 1910. 6. Mathews, A. P., and Walker, S., The action of cyanides and nitrites on the spontaneous oxidation of cystein. Jour. Biol. Chem. 6:29-37. 1909. 17. , The spontaneous oxidation of cystein and the action of iron and cyanides upon it. Journ. Biol. Chem. 6:289-298. 1909. 8. ScHONBEiN, C. F., iJber das Verhalten der Blausaure zu dem Blutkorper- chen und den Ubrigen organischen das Wasserstoffsuperoxide kataly- sirenden Materien. Zeitschr. Biol. 3:140-144. 1867. 9. ScHROEDER, H., Uber den Einflusse des Cyankaliums auf die Atmung von Aspergillus niger. Jahrb. Wiss. Bot. 44:409-481. 1907. ID. Vernon, H. V., The effect of hydrocyanic acid. Jour. Physiol. 35: 7o-73- 1906. 11. ViEHOVER, A., and Johns, C. O., Determination of small amounts of hydrocyanic acid. Jour. Amer. Chem. Soc. 37:601-607. 1915. 12. ViEHOVER, A., Johns, C. O., and Alsberg, C. L., Cyanogenesis in plants. Jour. Biol. Chem. 25:141-150. 1916. 13. Willaman, J. J., The estimation of HCN and the probable form it occurs in in Sorghum vulgare. Jour. Biol. Chem. 29:25-45. 1917. STUDIES OF SOME PORTO RICAN FUNGI Leo R. Tehon (with plate xviii) The fungi reported and described in this paper represent a miscellaneous group of specimens from the collections of Dr. F. L. Stevens in Porto Rico. They were selected by him and turned over to the author for study. Most of the numbers present points of special interest to the mycologist, as will be seen later in con- nection with the individual species. Type and co-type specimens are deposited in the herbarium of the University of Illinois, and in the herbarium of the New York Botanical Garden. CoccoMYCES De Notaris CoccoMYCES CLUSiAE (Lev.) Sacc. Syll. Fung. 8:147. 1889. On dead leaves of Chisia rosea Jacq., Maricao, July 19, 1915. nos. 882a, 8765 (figs. 8. 9. 10). It is assumed that these specimens belong to this species. No exsiccati have been available for comparison and the descriptions extant are too meager to allow of a positive determination. The fungus, however, clearly belongs to the genus Coccomyces, and its occurrence on Clusia is regarded as being sufficient to warrant calling the specimens C. clusiae. For the convenience of other workers an emended description of the species foUows: Spots irregular, large, light-colored, amphigenous, thickly dotted with ascomata. Ascomata amphigenous, but chiefly hypophyllous, black, circular, subepidermal, erumpent, 0.5-2 mm. in diameter, rupturing with the epidermis by 4 or 5 radial splits. Paraphyses filiform, numerous, coalescing above into a brown epithecium. Asci long, narrow, cylindrical, 100-135 X 7 m, 8-spored; ascospores filiform, I M in diameter and nearly as long as the asci, multi- cellular, hyaline. Under the microscope the bottoms of empty ascomata present a curious, honeycombed appearance as the result of numerous pits in the hymenial layer (fig. 9). With the hope of finding an explanation for the pitted condition there found, microtome sections were made of the ascomata, some of which contained 501 502 BOTANICAL GAZETTE [june asci and some of which did not. An examination of the sections disclosed the fact that, in certain regions of the hymenium, ascigerous hyphae do not develop. In these places there is, therefore, only a very thin covering formed over the bottom of the ascoma and a corresponding pit in the hymenial surface results. The formation of a large number of these pits causes the honeycombed appear- ance on the floor of older ascomata from which the asci have disappeared (figs. 9, lo). CoccoMYCES MUSAE (Lev.) Sacc. Syll. Fung. 8:752. 1889. On dead leaves of a cultivated species of Musa, Rio Maricao, above Mari- cao, September 10, 1913, no. 3631 (fig. 7). With this, as with the preceding species, no exsiccati have been available for comparison, and the current descriptions are not such as would make possible an accurate determination. The fungus clearly belongs to the genus Coccomyces, however, and its occurrence on Musa is regarded as being sufficient, with other general superficial characters, to place it under the species already described for that host. Since, however, the previous description is so meager, an emended specific description is here given. Spots amphigenous, v^^hitish or straw-colored, roundish, 5-10 mm. in diameter, frequently confluent, uniformly but sparsely dotted with ascomata. Ascomata punctiform, black, 350-750 n in diameter, rupturing irregularly by a 3 or 4-partite radial cleft. Asci cylindrical, short-stipitate, 55-90X10-12 m, 8-spored; asco- spores long, rodlike, with obtuse ends, 50-70X3 M, hyaline and multiseptate at maturity. Paraphyses long, filamentous, numer- ous, exceeding the asci, hyaline. Meliola Fries The following species constitute a few that were overlooked, either because of their resemblance to other fungi, or because of their inevidence on a host collected for the sake of other fungi upon it, when the monograph of Porto Rican Meliolas was compiled by Dr. Stevens.^ Meliola conferta, sp. nov.— Spots amphigenous, irregularly circular, punctiform, 0.5-1 mm. in diameter. Mycelium brown, densely compacted, radiate, branches opposite, filaments 8 a^ in diameter. 'Stevens, F. L., The genus Meliola in Porto Rico. 111. Biol. Monographs 2: no. 4. 1916. igig] TEHON— PORTO RICAN FUXGI 503 Capitate h>Tphopodia opposite, exceedingly crowded, intricately- interlocked with the h>phopodia of adjacent filaments, apical cell ovoid to globular, 12X18 ju in diameter, basal cell 3-6 m long. Mucronate hjphopodia few, opposite, situated mostly near the ends of the h>'phae, bottle-shaped, 18 /jl long. Mycelial setae none; perithecial setae few, 6-8, arising subapically, straight, 7 X8o M, tips obtuse. . Perithecia 120-135 )u in diameter, rough; asci 2-spored, soon evanescent; ascospores brown, 4-septate, obtuse, slightly con- stricted, 40X15 m- On leaves of Rhacoma crossopetalum L., Mona Island, December 20, 1913, no. 6147 (type) (figs. 17, 18, 19). The compact habit of this species is very remarkable. Starting from a single spore, the spot develops into an exceedingly thick black mass. The hyphae are arranged radially and branch repeatedly, so that when the whole colony is seen under the microscope it is almost impossible to distinguish individual filaments. The radial habit, when the fungus is viewed in place on the host leaf, strongly suggests the radial character of an immensely large Microthyriaceous perithecium. ■ This Meliola is probably most closely aUied to M. parethesicola Stevens, but can readily be separated from it because of its denser habit and the fact that the perithecial setae arise subapically instead of basally. What appears to be a constant manner of spore germination has repeatedly been seen. From one of the end cells of the spore there is put out a capitate hj^jhopodium (fig. 18) which acts as an anchor for the spore. Following this, filaments are sent out from the other cells of the spore and begin to branch and rebranch immediately. Thus the characteristic compactness of growth is assumed almost with the germination of the spore. Meliola perexigua Gaill. Le Genre Meliola. 98. 1892. On leaves of Petiveria alliacea L. at Corozal, February- 2i,iQi3,no. 415. This species has not hitherto been reported from Porto Rico, or on this host. The specimen, however, agrees thoroughly with Gaillard's description of the species. Meliola asterinoides Wint. Hedwigia 96. 1886. On the upper surfaces of leaves of Genipa americana L., no. 7135 (figs. 11-16). This specimen is referred to this species with some hesitation. The descrip- tion given by Gaillard^ characterizes the fungus as forming very small colonies, 0.25-1 mm. in diameter, with only a small number of subdimidiate ^ Gaillard, a., Le Genre Meliola. 58. 1892. 504 BOTANICAL GAZETTE [june perithecia (often only one) near the center of the spot, each of which sends out long radial filaments. Although no exsiccati having been available for com- parison, it was impossible to be certain of a correct interpretation of the description, our specimen seems to show certain more or less inconformable characters. '^ The colonies are large, 1-5 mm. in diameter, and give rise to a compara- tively large number of perithecia scattered uniformly throughout the entire colony. The perithecia are dimidiate, rather than subdimidiate, and are made up of radiating hyphae which extend somewl;iat beyond the limits of the perithecium to form an areola (fig. 16) somewhat similar to, but entirely distinct in aspect from that surrounding the perithecia of M. aibonitensis Stevens, M. comodadiae Stevens, etc. Gaillard called attention to the Microthyriaceous aspect of the perithecia of M. asterinoides, but the present specimen may be seen actually to possess Microthyriaceous characters in its dimidiate and radially formed perithecium which opens by a false ostiole. It is possible, therefore, that the specimen may throw some light upon the phylogenetic relationship of the genus. In mycelium and spores, this specimen is characteristically a Meliola; but in perithecial development and characters it is almost typically Microthyriaceous. The development of the perithecium begins by the extrusion of a hyphopodium-like branch, i or 2 cells long, at some point, usually near the growing tip (fig. 12) of a filament. The tip of this special branch then becomes swollen (fig. no), and just beneath the swollen part a small thumb like projection (fig. 11b) is sent out in such a manner that, at its com- pletion, what may be termed the first stage resembles somewhat the profile of a closed fist. The thumbhke portion now grows until it reaches approximately two-thirds the size of the enlarged tip, and lies alongside it (fig. 12a). There now appear, along the outer edges of the 2 prongs thus formed, evident indentations (fig. 13a) that eventually cut the tip of the perithecial branch into 5 or 6 cells, which may be regarded as perithecial mother cells. Simultaneously with the appearance of the marginal indentations, one finds growing out radially from beneath the edges of the perithecial mother cells a number of rather light colored hyphae (fig. 13 ft) which, as growth continues, eventually form a complete circle of elongated radiating cells (fig. 14a) about the cells from which they originated. When iqiq] TEHON— PORTO RICAN FUNGI 505 the circle has reached full development most of its cells will be found to possess terminal indentations (fig. 146), indicative of a dichotomous scheme of branching, and from beneath the outer edges of some cells may already be seen (fig. 14c) the beginning of a new circle similar to the last. Thus by the addition outwardly of circle after circle of radiat- ing, dichotomously branching cells (figs. 15, 16) the complete perithecium is formed. Fig. 16 shows a mature perithecium illus- trating the true dimidiate character. On ripening, the central portion becomes black and opaque, so that it is impossible to see the hyphal structure, but around the edge there is always apparent the areola of radiating close lying h>^hae, some of which extend out- ward rather farther than the others and are capped by a hyphopodium-like head cell (fig. i6a). In some cases these elon- gated hyphae from the perithecial areola may even send out capitate hyphopodia laterally. The false ostiole is a character which, if taken together with the dimidiate character of the perithecium, may also be of some phylogenetic significance. Although Meliola has never been definitely allied to the Micro- thyriaceae, Engler referring it to the Aspergillaceae and Saccardo to the Perisporiaceae, the impression is becoming more and more firmly rooted that Meliola is closely related to the Microthyriaceae. In this connection, the present specimen shows certain characters which strengthen that impression. The specimen may even be regarded as a transitional stage connecting this genus with the typical genera of the Microthyriaceae. Meliola cestri, sp. nov. — Colonies epiphyllous, irregularly circular, 1-3 mm. in diameter. Mycelium dark, straight, forming a rather close network, filaments 9-10 /x in diameter. Capitate h>phopodia alternate, 35 ju distant, head-cell cylin- drical to globose, 16-20X8-13 ju; basal cell 4-5 ix long. Mucronate hyphopodia opposite, bottle-shaped, 24-28X9 ju- Perithecial setae none; mycelial setae numerous, straight, black, 650-850X10-11 ix, tips obtuse. Perithecia numerous, grouped in the center of the colony, sur- rounded by an areola of hyphae when young, smooth, 225-275/4 5o6 BOTANICAL GAZETTE [june in diameter. Asci soon evanescent; ascospores 4-septate, dark brown, cylindrical, 50-55X18-20 )u, definitely constricted at the septa. ■ On leaves of Cestrum species, Mayaguez, June 29, 1915, no. 7576 (type). This species is entirely distinct from M. gesnerii Stevens, which has also been reported on this host. Meliola clusiae Stevens, 111. Biol. Monographs 2:75. 1916. On Clusia gundlachi Stahl, El Alto de la Bandera, July 15, 191 5, no. 8670. Previously reported on Clusia rosea Jacq. Meliola bayamonensis, sp. nov. — Colonies hypophyllous, 2-5 mm. in diameter, mycelium a very loose network of threads; branches alternate, hyphae dark, 4 m in diameter, wavy. Capitate hyphopodia alternate, 30-60 ju distant, the head cell globose to oval, 8-10 n in diameter; the basal cell variable, 6-16 n long. IMucronate h>^hopodia few, alternate, 14 ju long. Mycehal setae none; perithecial setae 5-7, arising basally, decumbent, dark brown to opaque, 225X4-5 M, apex acute. Perithecia scattered, 100-135 ju in diameter, rough, Asci soon evanescent, ascospores 4-septate, 27-30X7 M, slightly constricted at the septa. On Psychotria pubescens Sw., at Bayamon, February 19, 1913, no. 392 (type). This species is separated immediately from the variety described on the same host by Stevens as M. glabra, var. psychotriae by the presence of peri- thecial setae. It is also distinct from M. psychotriae Earle on account of its remarkably looser habit, the characteristic waviness of the mycelium, its lighter color, the shape of the capitate hyphopodia, and the strikingly short basal cell. Meliola marcgraviae, sp. nov. — Colonies epiphyllous, irregular, 3-10 mm. in diameter, branching opposite, hyphae dark to opaque, 5-6 ju in diameter. Capitate hyphopodia alternate, 32 m distant, head-cell globose, II M in diameter; basal cell short, 5 fx long. Mucronate hypho- podia mostly opposite, but frequently alternate, flask-shaped, 12-14 jLt long. Mycelial and perithecial setae none. 1919] TEHON— PORTO RICAN FUNGI 507 Perithecia scattered, small, 65-75 M in diameter. Asci soon evanescent, ascospores 4-septate, light brown, cylindrical, slightly constricted, 40X15 /x. On leaves of Marcgravia rectiflora Tr. and Planch., Porto Rico, July 16, 1915, no. 8722 (type). The colonies formed by this Meliola are so inconspicuous as to pass entirely ■ unnoticed. It is only when the leaf is placed under a lens that the colony is to be seen. Phyllachora Nits. Phyllachora quadraspora, sp. nov. — Stroma variously shaped, mostly oval to linear, 0.5-1X0.5 mm., with epidermal clypeus on either side of the leaf usually bilocular. Locules sub-spherical, 115-125/1 in diameter. Asci cylindrical, short-stipitate, loo-iio X10-12 )U,4-spored; spores hyaline, granular, i-guttulate, elliptical, 20-22 X 8 )U- Paraphyses present. On leaves of Paspalum glabriim Poir, Maricao, no. 8803 (type); P. con- jugatum Bergius, Tanama River, no. 7856 (fig. 4). In naming this species the author has endeavored to follow the monograph of the genus by Theissen and Sydow.^ The fact that the stroma is laid down in the mesophyll of the leaf brings the fungus into the genus Phyllachora; but the occurrence of only 4 spores in the ascus would seem, on the other hand, to exclude it. This, however, need not be true, since Theissen, in characterizing Ph. graniinis as having spores up to 8 ("zu acht") leaves the implication that the spore number may be fewer than 8. Phyllachora graminis (Pers.) Fcl. Symb. Myc. 218. 1869. This species has previously been reported by Carman on an undetermined species of Paspalum from Sabana Grande in 191 5. It is also to be noted on Paspalum glabrum Poir., at Rosario, no. 9495a. Phyllachora ischmaemi, sp. nov. — Stromata appearing on the upper side of the leaf, circular, crowded, and often confluent, laid down in the mesophyll, 0.75-1.25 mm. in diameter. Clypeus 24-26 ju thick. Locules 2-several in a stroma, spherical to flask- shaped, 125-145 jj. in diameter. Asci cylindric to clavate, 105- 150X10-12 M, 8-spored. Spores hyaline, spherical, uniseriate, 8 ju in diameter, with a single guttula. Paraphyses present, filiform. On leaves of Ischmaemum latifolium, St. Pierre, Martinique Island, no. 2972 (type) (figs. 2, 3). 3 Theissen, F., and Sydow, H., Die Dothidiales. Ann. Mycologici 13 1437. 1915. 5o8 BOTANICAL GAZETTE [june Stigmatea Fries Stigmatea guettardae, sp. nov. — Spots large, 0.5-3001, in diameter, irregular, light brown or red in young stages and ashy or white in old stages, always sharply defined by a dark brown border. Perithecia epiphyllous, gregarious, grouped in the center of the spot, small, black, 60-80 fj, in diameter. Ostiole erumpent, small, 6-10 n in diameter. Asci obhquely oblong, short and abruptly stipitate, obtuse above, 2 5-40 X 8-1 1 fx, 8-spored. Spores long- elliptic, 2-celled, the septum at or nearly at the center, shghtly constricted, hyaline, 11-13X2-3/1. Paraphyses present, long, filiform, septate, hyaline to yellowish. On Guettarda ovalifolia Urb., Maricao, January 10, 1913, no. 191 (type); Barros, January 2, 1913, no. 164; Maricao, April 5, 1913, no. 771; July 19, 1915, no. 8804; Monte Alegrillo, no. 4741; Indiera Fria, Maricao, October 8, 1913, no. 3338. On Guettarda scabra (L.) Lam., Tanama River, July 6, 1915, no. 7851. Phl^eosphaerella Karst. Phaeosphaerella paspali, sp. nov. — Perithecia amphigenous, sunken, sub-spherical, 125-137/1 in diameter; ostiole minute, 10-15/1 in diameter. Asci crowded, sub-cylindric, 55-60X8/1, 8-spored; spores brown, oblong to fusoid, unequally i -septate, slightly constricted at the septum, 12-15 X 3-5 /i- Paraphyses none. On leaves of Paspalum glahrum Poir., Maricao, no. 8803a (type) (fig. 5). CoNiOTHYRiuM Corda Coniothyrium marisci, sp. nov. — Spots oval to hnear, yellowish, with a dark brown margin. Pycnidia gregarious, amphigenous, sub-spherical, 120-130 /i in diameter, subepidermal, only the ostiole erumpent. Perithecial wall thick (35/1); ostiole 16-20 /t in diameter. Spores dark brown, ellipsoid or globose, 5-7X2.5/1. Conidiophores not apparent. On Mariscus jamaicensis (Crantz) But., no. 124 (type) (fig. 6). Pestalozzia De Notaris Pestalozzia lucumae, sp. nov. — Spots, epiphyllous, black, irregular, 2-5 mm. in diameter. Acervuli subepidermal, erumpent, iqiq] TEHON— PORTO RICAN FUNGI 509 white or ashy at maturity, o . 5-2 mm. in diameter, crowded on the stroma, circular or elongate, and rupturing irregularly. Conidia elliptic or slightly falcate, 14-18X4-5 ju, 4-septate, slightly con- stricted; central cells fuscous, end cells hyaline, the apical cell conic, ornamented by 2 widely divergent, hyaline, fihform setae 7-10 /x long, basal cell prolonged into a single hyaline seta 3-5 n in length. On leaves of Lucuma muUiflora A. DC, Monte Alegrillo, July 20, 1913, no. 2301 (type). It was a surprise to find this specimen a Pestalozzia. Its superficial characters suggest strongly a Phacidiaceous form, but a close examination reveals the fact that instead of an ascoma, the fruiting body is an acervulus. The acervuh are grouped closely on a large stroma which causes a characteristic black spot on the leaf. Helminthosporium Link Helminthosporium folliculatum Corda var. brevipilum Corda, Icon. 2:13, 1838. On Paspalum conjugaliim Bergius, Tanama River, July 6, 1915. AcROTHECiUM Prcuss. Acrothecium flacatum, sp. nov. — Mycelium light olivaceous, septate, 2 . 5 ju in diameter, attacking the spikelets and thickly covering the glumes and awns of the flowering heads. Conidio- phores dark olivaceous in color, straight, somewhat bulbous at the base, 96-125 X 5 /x. Spores olivaceous, luniform, 3 or 4-septate, borne apically on the conidiophores in fascicles of 3-5, 35X10JU; central cell very much enlarged, dark, and not equilateral; the terminal cells small and nearly hyaline. On Setaria species, Porto Rico, 191 5, no. 9181 (type) (fig. la, b, c). Another fungus belonging to this genus and possessing the same strikingly characteristic spore is generally reported as the cause of the "ringspot" of sugar cane. It was first mentioned by Breda de Haan in 1892, and was thought by him to be the conidial stage of Leptosphaeria sacchari. In 1898 a fungus identical with that of Breda de Haan's was described by Wakker^ on dead leaves of sugar cane under the name of Acrothecium lunatum. The present species, however, is clearly distinct from that described by Wakker, in that the conidia are more frequently 4-septate than 3-septate, distinctly larger, and regularly borne in fascicles of 3-5 on the tips of the ■t Wakker, J. H., and Went, F. A. F. C, De Ziekten van het Suikerriet op Java, pp. 149, 196. Leiden. 1898. 510 BOTANICAL GAZETTE [june conidiophores. Cultures of Acrothecium lunatum frequently show a secondary group of spores borne below the apical group. Furthermore, the 2 hosts being so widely separated phylogenetically, it is probable that they would not be attacked by the same parasite. Cercospora Fries Cercospora personata (B. and C.) Ellis, Jour. Mycol. 1885. On the leaves of Arachis hypogaea L., Trujillo Alto, August 17, 1915, nos. 2506, 2447. These specimens agree in general with the species, comparison having been made with reliable exsiccati. They do present a peculiarity, however, in that the spots are consistently smaller, and much more regularly circular in outhne. Trichostroma Corda Trichostroma axonopi, sp. nov. — Spots oval to linear, yellowish with a definite brownish or purple border. Sporodochia gregarious, black, globular to oval, verruciform, often confluent, 95-125 m long. Setae black, straight or the tips sometimes repand, rigid, base bulbous, 65-85 M long, few-septate. Conidia brown, globular to ovoid, 5 M in diameter. On leaves of Axonopus compressus (Sw.) Beauv., College grounds, Maya- guez, May 30, 1913, no. 924 (type). EXPLANATION OF PLATE XVIII Fig. I. — Acrothecium falcatiim: conidiophore and conidia (a, h, c). Fig. 2. — Phyllachora ischmaemi: section of stroma. Fig. 3. — Phyllachora ischmaemi: ascus. Fig. 4. — Phyllachora quadraspora: ascus. Fig. 5. — Phaeosphaerella paspali: ascus. Fig. 6. — Coniothyrium marisci: section of pycnidium. Fig. 7. — Coccomyces miisae: ascus and ascospores. Fig. 8. — Coccomyces clusiae: habit sketch of single ascoma. Fig. g.^Coccomyces clusiae: old and empty ascoma showing honeycombed appearance of floor. Fig. 10. — Coccomyces clusiae: cross-section of ascoma showing localization of ascus producing hymenium which results in honeycombed appearance seen in fig. 9. Figs. 11-16.^ — Meliola asterinoides. Fig. II.— Beginning of perithecium: a, large swollen tip; b, smaller thumbUke projection. BOTANICAL GAZETTE, LXVII PLATE XVIII 18 TEHON on PORTO RICAN FUNGI iqiq] TEHOX— PORTO RICAX FUXGI 511 Fig. 12. — Further stage in perithecial development: a, thumblike pro- jection grown to approximately two-thirds size of swollen tip. Fig. 13. — Third stage: a, indentations appearing on margins of the 2 prongs; b, radial hyphae developing from under side of perithecial mother cells. Fig. 14. — First circle of radial cells complete: a, radial cells; b, marginal indentations; c, new cells which will form another concentric circle outside first. Fig. 15. — Perithecial development in intermediate stage: pronounced radial character evident. Fig. 16.— -^Mature perithecium: showing areola of radiating hyphae, dimidiate character, and false ostiole; a, elongated perithecial h^'phae capped by hyphopodium-like cell. Figs. 17-19. — Meliola conjerta. Fig. 17. — Ascospore. Fig. 18. — Ascospore germination: showing manner in which compact habit is assumed. Fig. 19. — Portion of myceHum illustrating compact habit. CURRENT LITERATURE BOOK REVIEWS The living cycads In a recent volume of the ''University of Chicago Science Series" Cham- berlain' presents in untechnical language some of the results of his 15 years of study on cycads. This group of 9 genera has several advantages from the point of view of popular treatment, and of these the writer makes good use. It may safely be said that no one is so well fitted as the author to deal with this subject, because he approaches the task not only as an expert student of the group, but also as one who knows the cycads as living plants, on account of his extended travel in regions where cycads occur. The title of the book was selected, so the author tells us, to contrast with Wieland's The fossil cycads. The account is divided into 3 sections, part I dealing with the collecting of the material, part II with the hfe history, and part III with the evolution and phylogeny of the group. The first part recounts the author's journey- ings in the 3 cycad regions, tropical America, Australia, and South Africa, but is much more than a narrative, for it includes important observations taken in the field which help to clear up certain disputed points. This part is rendered attractive by the author's excellent photographs of field speci- mens as well as by his graphic style, and will probably be the most interesting section to the non-botanical reader. The section on life history includes a short chapter on vegetative structures, in which the vascular tissues receive scant attention. The very significant anatomical features were probably considered too technical for presentation to a general audience. The sporoge- nous and gametophytic structures, which have attracted so much attention to the group, are clearly described and figured, and are compared with the corresponding structures of ferns. It is in this part of the book that the author's particular contribution to our knowledge of cycads comes out con- spicuously. One realizes how many gaps in the story have been filled during the past decade. Part III, the shortest in the book, passes on from fact to speculation, and brings the ferns and fossil cycads into relation with the living cycads. Especially satisfying is the account of the evolution of the cone, which presents a remarkable series even in living genera. The point of view throughout this section is a conservative one, as is evinced by the treatment of Bennettitales and their relationships. 'Chamberlain, Charles J., The living cycads. 8vo, pp. xiv-|-i 72.- _/zg5. gi. The University of Chicago Press. 1919. 512 19 iq] current literature 513 Botanists will be glad to have the author's assurance that a more extended and technical account of the living cycads is in preparation. In the present pocket volume of 172 pages the author has done a useful piece of work and has done it well. — M. A. Chrysler. NOTES FOR STUDENTS Inheritance in Pediastrum. — Practically all of our present knowledge of inheritance in the plant kingdom is based upon work done with flowering plants, regularly involving the sex act. The sex act in flowering plants, furthermore, is peculiarly obscured; we cannot be altogether certain what happens between the time of pollination and seed germination. We think that the program followed is a remarkably regular one, but we feel that fre- quently irregularities may occur, and we know that sometimes they do. We wish therefore to know what all of these irregularities are, how they affect inheritance, and how they may be induced or controlled artificially. It has long been felt that a study of inheritance in simple plants would be suggestive, for in them many of the complexities surrounding reproduction are stripped away. The sex act takes place "in the open," so that there is more hope of absolute control; some forms even he "below the level of sex," furnishing unusual material for "pure line" work; and germ plasm seems identical with body plasm. The direct bearing of such a study upon practical genetics may be negligible, but upon theoretical genetics it promises to be profound. Harper^ has been working with cultures of Pediastrum, and has developed some very significant ideas. He considers 3 "degrees of directness" of inheri- tance in Pediastrum: (i) direct transmission, as by division of plastids; (2) the more indirect transmission of those adult cell characters (as cell form) which are not visibly present as such in the germ cell; (3) the entirely indirect transmis- sion of the characters of the many-celled organism as a whole (as colony form). The adult cell characters which Harper observed "do not suggest the working out of influences emanating from elements in the chromosomal organization of the nucleus, but rather the direct expression of the organization of the cell as a whole when it begins to grow," involving specific polarities, surface tension, etc. These cell characters come into expression whether or not the colony is successfully formed. Colony characters, therefore, are dependent upon individual cell characters, rather than the reverse. "If in the swarming period the cell does not achieve its normal position .... the maladjustment is never overcome." Thus the author paints for us two distinct pictures, which should be considered separately. First, the picture of inheritance through specific polarities, etc., of protoplasts as a whole, rather than determiners located on ^ Harper, R. A., Organization, reproduction, and inheritance in Pediastrum. Proc. Amer. Phil. Soc. 66:375-439. pis. 5, 6. figs. 54. 1918. 514 BOTANICAL GAZETTE [june chromosomes. This may be a rather general situation among the simpler plants, where germ plasm and body plasm are merged. Whether it is at all applicable to higher plants is questionable. Perhaps the " phylogenetic age" of the latter has brought this difference of body plasm and germ plasm, involv- ing a rigid chromosome mechanism.' The other picture is "that the swarming period .... is not one of aimless movement .... but a definitely directed effort to achieve for each cell a specific relation to its fellows." Successful achievement means normal colonies; otherwise monstrosities result.. This situation could apply only to a very limited number of cases, even among the lower plants. Among higher plants a vivid imagination might attempt to apply it to the free nuclear stage in the embryo formation of gymnosperms, or in the organization of the embryo sac of angiosperms. The author, however, does not carry his ideas beyond Pediaslrum, where they seem quite appro- priate and well founded. Similarly careful work upon less peculiar types of algae should yield even more profitable suggestions. — Merle C. Coulter. Mendelian inheritance in gametophytes. — One of the most critical tests of the current theoretical mechanism for inheritance lies in the behavior of the gametophyte generation in inheritance. If our Mendelian mechanism is correct, gametophytes should show predictable peculiarities; segregation should take place in the first hybrid generation, and dominance should be out of the question. Such an investigation is not particularly hopeful among angiosperms, owing to the insignificance of the gametophytes. In fact it is a rather general opinion that "the characters which they possess appear to be wholly sporophytic, the factors which they carry functioning only after fertiliza- tion."^ Belling'' ejs^lains semi-sterility in beans on the basis of the germinal equipment of the gametophytes upon the gametophytes themselves, but this merely involves lethal effects. More hopeful material is provided by the lower plants, where the gameto- phyte generation is more prominent and really has characters of its own. Transeaus reports hybridization in Spirogyra, and it is significant that he can give it a Mendelian interpretation. Unfortunately the work is as yet merely observational rather than experimental. Hybridization was observed taking place in nature between S. communis and S. varians, S. varians and S. poHicalis. The 3 species involved showed distinguishing characters in the shape and size of the vegetative cells, and the shape and orientation of the zygotes. The auther looked in the immediate vicinity, therefore, for possible hybrids result- ing from these crosses which should display new combinations of the parental 3 East, E. M., and Park, J. B., Studies on self-sterility. I. The behavior of self-sterile plants. Genetics 2:525-609. 191 7. '^ Belling, John, Lethal factors and sterility. Jour. Heredity 9:161-165. 1918. s Transeau, Edgar Nelson, Hybrids among species of Spirogyra. Amer. Nat. 53:109-119. figs. 7. 1919. iqiq] current literature 515 characters. He was successful in finding practically all the new combinations that were theoretically possible. The Mendelian explanation runs as follows: The character of the hybrid zygote itself is maternal, as is to be expected from the cytological behavior during conjugation. The reduction division takes place during the first 2 nuclear divisions of the germinating zygote, but 3 of the resulting nuclei degen- erate, so that the cells of the mature filament all have a common ancestor in the fourth nucleus; hence segregation appears in the first hybrid generation, but of course all of the cells of a given filament are alike. Such facts would furnish excellent support for our theoretical mechanism of inheritance, but the author could not be positive as to whether he was deahng with an Fi or an F2 generation. It is to be hoped that he will discover how to cultivate this material in the laboratory, and carry the work further under rigid experimental control. — Merle C. Coulter. Enzyme action. — Van Laer* reports some observations on the nature of zymogens, which are claimed to confirm the results of Ford and Guthrie, who had shown that the increase of the amyloclastic activity of papaine with barley meal is not manifested when the infusion is kept in direct contact with the proteo-clastic ferments. The yeast infusions were obtained from yeast pre- pared according to the Lebedeflf method. The addition of papaine to yeast juice destroyed the catalase and zymase. In the state of zymogens, there was shown greater stability and resistance to the factors of inactivation. The hefanal extract of yeast in the presence of antiseptics showed a measurable degree of inverting activity. This inverting agent was amylase. The diastase and papaine had no influence upon the hefanel infusion even after a 24 hours' digestion. Observation is made upon the intensity of autofermentation. After the latter there remains some amylase which is sensitive to papaine. This sensitiveness is expressed in the data as the decrease of the percentage of sugar inverted from 25 .6 to 19 when papaine was added. Certain cellulary materials, as soluble or incoagulable protoplasmic products, decreased the activity of sucrase according to the concentration. In the presence of small quantities of these substances the rapidity of hydrolysis of saccharose is hardly modified. Extracts of yeast inactivated by acetone give a notable increase of inverting power when added to solution of papaine or active anylase, the yeast cells in this respect behaving like cellulary bodies. This increase is due on the one hand to the increase of sucrase, and on the other to the decrease of cellulary substance in the digestion products. — A. M. Gurjar. Buried weed seeds. — Miss BRENCHLEy,^ on the basis of considerable investigation, makes the following statement concerning the longevity of weed seeds in agricultural soils: "The changes in the proportion of arable and n' AX Laer, Henri, Zeits. fiir Garungsphysiologie 6:169-175. 1918. 7 Brenchley, Winifred E., Buried weed seeds. Jour. Agric. Sci. 9:1-31. 1918. 5i6 BOTANICAL GAZETTE [june grassland plants derived from buried seeds are so consistent and so regularly- associated with the history of the land that one is irresistibly forced to the conclusion that when arable land is grassed over, a certain number of the seeds are able to retain their vitality for very many years. Many of the seeds die within a comparatively short time after burial, and as time goes on the number of living seeds gradually becomes less, although the evidence goes to show thaf some seeds will survive burial for at least 58 years. Usually most of the older arable seeds survive in the lower depths of soil where the conditions are less variable, whereas the shorter the time that land has been under grass the greater , the proportion of arable seeds that are found near the surface. While the stock of arable seeds is diminishing with the lapse of time, the supply of grassland seeds is being augmented by fresh seeds that are ripened by the surface vegetation and are gradually carried down into the soil. Naturally enough, the greater number of these seeds are found in the upper inches of soil, comparatively few penetrating below the eighth inch." Miss Brenchley fails to note the much earlier and extensive work (1893- 94) of Peter, which is very similar to hers in method and conclusion. She also fails to mention the well controlled work of Beal and of Duvel on the longevity of buried seeds, which likewise justifies her conclusions.* — Wm. Crocker. Wound callus and bacterial tumor. — Polar difference in wound callus formation has often been observed in stems, and less frequently in root structure. Magnus' finds that segments of the root of a half long carrot with which he worked produced a wound callus on the morphologically apical face, but not on the basal face. This occurred whether the apical face was oriented upward or downward in the moist chamber. The callus starts at the cambium ring and spreads centripetally. When the apical face is infected with Bacterium tumejaciens the callus development is much greater. When the basal face is infected there is a considerable development of tumors on that face, and this acts in a correlative way to inhibit the normal tumor develop- ment in the apical face. Magnus also worked with a long fodder carrot. While infection in this form increased the callus development on the apical face of the segments tenfold, it induced very little tumor development on the basal face, and accordingly showed little correlative effect in inhibiting the normal development on the apical face. Magnus offers evidence for the view that the tumor inducing organism in plants is not identical with that in man. He also suggests that certain con- clusions of Blumenthal and Hirschfield on the effect of Diplococcus in ' See Crocker, Wm., Mechanics of dormancy in seeds. Amer. Jour. Bot. 3:99- 120. 1916. ■'Magnus, Werner, Wund-callus und Bakterien-Tumore. Ber. Deutsch. Bot. Gesells. 36:20-29. 1918. igig] CURRENT LITERATURE 517 tumor formation in plants may be wrong because they failed to recognize the polar disposition to callus formation. He thinks the studies on tumor forma- tion in plants will finally throw much light on cancer development. — Wm. Crocker. Effect of illuminating gas on plants. — Wehmer" has studied the effect of passing continuous streams of illuminating gas through the soil bearing potted herbaceous as well as 3-7-year-old woody plants. There was a great difference in the amount of injury, according to the stage of development. In the spring the trees were entirely killed in a relatively short time. This is in general the sort of reaction given by the actively growing herbaceous forms at all times. In late summer and early fall the injury is less marked and is shown mainly by leaf fall, while in the dormant period of winter the trees are very resistant. In the cress the embryo in the resting seed and the seedling stage proved very sensitive. Cuttings stood in gas-impregnated water shewed, with few excep- tions (Ilex), seasonal variations in sensitiveness similar to the plants rooted in soils. In spite of this the author thinks that injury to parts above the soil is in part a secondary result of root injury. The injury is due to toxic conditions of the gas and not to mere displacement of oxygen by the gas, as Sorauer has suggested. The toxic constituents increase or decrease with the conditions that lead to an increase or a decrease in the odor-producing materials. A later paper on the toxic constituents is promised. The author seems to have over- looked most of the literature on the effect of illuminating gas on plants. — Wm. Crocker. Aeration systems of leaves. — Neger" has earlier spoken of 2 types of leaves on the basis of the nature of their intercellular systems, heterobaric and homo- baric. In a recent article he compares a heterobaric leaf to a house with thousands of rooms lacking communicating doors, and a homobaric leaf to a similar house with communicating doors present and all open. In the first type the intercellular system is divided into many small isolated regions by the smaller veins, with the resulting possibility of different air pressure existing in each; while in the second the whole intercellular system of the leaf is con- nected and therefore the same pressure exists throughout. Most plants with flat leaves have heterobaric leaves, and the size of the individual chambers varies considerably. In various species of Quercus they run from 1/840 to 1/1400 sq. cm., and in Syringa vulgaris from 1/8 to i/io sq. cm. In the same species shade leaves have larger chambers than sun leaves. The following trees and shrubs have homobaric leaves: Evonynms japonica, Ilex aqui^olium, '"Wehmer, C, Leuchtgaswirkung auf Pflanzen. 4. Die Wirkung des Gases auf das Wurzelsystem von Holz-pflanzen; Ursache der Gaswirkung. Bar. Deutsch. Bot. Gesells. 36:140-144. 1918. " Neger, F. W., Die Wegsamkeit der Laubblatter fiir Gaze. Festschrift zum Ernst St.^hl. pp. 152-161. Jena. 1918. 5i8 BOTANICAL GAZETTE [june Prunus Laurocerasus, Hedera Helix, Ardisia crispa, and all needle bearing trees and shrubs. When injured by smoke the homobaric leaves show the injury to the whole leaf due to the gases distributing themselves throughout the whole intercellular system, while the heterobaric leaves show the injury in spots corresponding to individual intercellular chambers. — Wm. Crocker. Nitrogen fixation by Azotobacter. — Hutchinson," in agreement with Koch, Remy, and others, finds that the nitrogen content of sand or soil may be increased appreciably by the activity of Azotobacter when some suitable source of energy is suppUed. Sugars proved very effective as an energy source, and distinct gains were obtained with plant residues. In pot cultures the nitrogen gains ran as high as 9 mg. of nitrogen per gram of plant residue added. Even in field cultures additions of sugar increased crop yields 20-54 per cent when conditions were favorable. Hutchinson beUeves the carbohydrates of plant residues act in a similar way in furthering nitrogen fixation and crop yields. For successful operation of this organism suitable temperature, the presence of phosphates, and a supply of basic material such as calcium carbonate are necessary. Besides these factors, some unknown conditions appear periodically in the soil, interfering with the action of this organism. The effect of the addition of straw or other crop residues to the soil may be very complex. As important among these effects may be mentioned modi- fication of physical condition of the soil, direct addition of nutrients (in the case of straw, considerable potash, little nitrogen as well as other nutrients), and the indirect addition of nitrogen through furnishing an energy source for Azotobacter.— Wu. Crocker. Fucosan vacuoles. — Hansteen noted that granules, as he called them, accumulate about the chromatophores of Phaeophyceae during carbon assimilation. He thought they were produced by the chromoplasts and were the first visible product of carbon assimilation. On this basis he called them fucosan granules. Kylin'^ has made a rather extensive study of these bodies, the results of which are summarized in the article here reviewed. He finds that these bodies are vacuoles rather than granules, and while they are prob- ably formed by the chromoplast in connection with carbon assimilation, they are not made* up in the main of carbon synthate. He thinks he has shown that, dextrose is the first carbon synthate of the Phaeophyceae, and that this is condensed to lamanarin. These vacuoles may be the means by which the synthate leaves the plastid, but it is not stored in them. On the contrary, it rapidly diffuses from them into the cytoplasm. He thinks these vacuoles, especially the older ones, are filled with substances resembhng tannin, but '^ Hutchinson, H. B., The influence of plant residues on nitrogen fixation and on losses of nitrate in the soil. Jour. Agric. Sci. 9:92-111. 1918. " Kylin, Herald, Uber die Fucosanblasen der Phaeophyceen. Ber. Deutsch. Bot. Gesells. 36:10-19. 1918. igig] CURRENT LITERATURE 519 differing from true tannin in some respects. He considers these tannin-like substances as meaningless waste products, which upon oxidation give rise to the brown pigment of this group of plants (phycophaein), which was formerly- considered a pigment of the chromatophore. — Wm. Crocker. Apogamy in Nephrodium. — Steil''' has uncovered an interesting situation in Nephrodium hirtipes, in that the gametophyte never produces archegonia, although antheridia with normal sperms are developed. Secondary game- tophytes were induced readily, but only rarely was apospory induced. The embryo begins to develop early in the history of the gametophyte as a vegeta- tive outgrowth, the apical cells of leaf, root, and stem appearing successively, but no foot is formed. No migrations or fusions of nuclei were observed in connection with embryo development. The diploid number of chromosomes is 120-130, and the haploid number (60-65) was observed both in the gameto- phyte and the apogamous sporoph.yte. Suggestions are offered as to the origin of such persistent apogamy, the most interesting one being that N. hirtipes and other apogamous ferns may be of hybrid origin. The paper is introduced by a very useful summary of our knowledge of apogamy in ferns, beginning with the discovery of tracheids in the gametophyte of Pteris sulcata by Leszcyc- SuMiNSKi in 1848. The first clear recognition of an apogamous embryo, however, as distinguished from one resulting from fertilization, is credited to Farlow, who in 1874 discovered apogamy in Pteris cretica albo-lineata. — J. M. C. Soil acidity. — Hartwell and PEMBER'sfind that rye does well on acid soils, while barley is much injured by them. Aqueous extract of acid soil, residue from distillate of the aqueous extract, and the ash of the residue from such distillate affect the two plants much as does the soil itself. Soluble aluminum salts and not the acid in the soils proved to be the source of injury to the barley plants. Addition of acid phosphates, which renders the soil more acid,*&nd lime reduced the solubility of the aluminum salts in acid soils and rendered them non-toxic to barley plants. The authors think these substances often produce beneficial effects in this way rather than by furnishing more available phosphorus or by neutralizing the acid. The reviewer has noticed that the hydrogen ion concentration found in acid soils by the gas chain method is generally only a fraction of the hydrogen ion concentration necessary to reduce the growth rate of plants in water or sand cultures. This work again suggests the complexity of the apparently simple problem of soil acidity.^ Wm. Crocker. '•* Steil, W. N., A study of apogamy in Nephrodium hirtipes. Ann. Botany 33: 109-132. pis. 5-y. 1919. 's Hartwell, B. L., and Pember, F. R., The presence of aluminum as a reason for the difference in the effect of the so-called acid soil on barley and rye. Soil Science 6:259-279. 1918. 520 BOTANICAL GAZETTE [june Statolith starch. — Miss Zollikofer'* finds that the statolith starch of seedhng organs is relatively readily removed by periods of illumination fol- lowed by periods of darkness. The persistence of the statolith starch is a function of the degree of etiolation. This the writer considers a biological adaptation. By growing seedlings of Tagetes erecla and seedlings of other Compositae in light 1-4 days, followed by 3-4 days of darkness, hypocotyls were obtained that bore no statolith starch. These hypocotyls were still growing and capable of phototropic movement, but incapable of geotropic movement. Light rendered them geo-sensitive only after it had produced statolith starch. Working by similar methods the author shows a close rela- tion between the amount of mobile starch and geo-sensitivity in the coleoptile of grasses. — Wm. Crocker. Fat storage in evergreen leaves. — A number of investigators have stated that there is considerable storage of fats in evergreen leaves during the winter. Meyer'' finds that the droplets that were supposed by these former workers to be fat droplets are not fat, and that the total volume of these does not rise and fall with winter and summer, but that it increases continuously with the age of the leaf. He speaks of the droplets as "mesophyllsekret," and points out that little is known of their origin and composition. Some of the forms studied were Ilex aqiiifolium, Taxus baccata, and Vinca minor. The methods used by Meyer, as well as by former workers, are exclusively microchemical. It is evident that these ought to be checked up by quantitative determina- tions.— Wm. Crocker. Light and germination. — Lehmann'* finds in a germinator at 30° C. o. i second illumination with 730 H.K. is sufficient to cause 50 per cent of the seeds of Lythrum Salicaria to germinate within 24 hours, whereas only 6-7 per cent germinate in similar condition in darkness, and not more than 7 per cent after 10 days. — Wm. Crocker. Osmotic pressure of epiphytes. — Harris"' finds that the epiphytes ^Bro- mehaceae, Orchidaceae, and Peperomia) of the Jamaican montane forest have 37-60 per cent of the osmotic pressure shown by the herbaceous terrestrial vegetation of the same region, and 28-45 P^r cent of that of the ligneous terrestrial vegetation. — Wm. Crocker. 'fi ZoLLiKOFER, Clara, tjber das geotropische Verhalten entstarkter Keimpflanzen und den Abbau der Starke in Gramineen-koleoptilen. Bar. Deutsch. Bot. Gesells. 36:30-38. 1918. '7 Meyer, Arthur, T)ie angebliche Fettspeicherung immergruner Laubblatter. Ber. Deutsch. Bot. Gesells. 36:5-10. 1918. '* Lehmann, Ernst, tJber die minimal Belichtungszeit welche die Keimung der Samen von Lythrum Salicaria. Ber. Deutsch. Bot. Gesells. 36:157-163. 1918. '9 Harris, J. Arthltr, On the osmotic concentration of the tissue fluids of phanero- gamic epiphytes. Amer. Jour. Bot. 5:490-506. 1918. GENERAL INDEX GENERAL INDEX Classified entries will be found under Contributions and Reviewers. New names and names of new genera, species, and varieties are printed in bold face type; synonyms in italics. • Acer floridanum 234; Negundo 237; Negundo arizonicum 240; Negundo californicum 241; Negundo interior 239; Negundo texanum 239; Negundo texarium latifolium 239; Negundo violaceum 238; rubrum 235; rubrum Drummondii 236; rubrum Drum- mondii rotundata 237; rubrum tomen- tosum 235; rubrum tridens 237; sac- charum 233; saccharum glaucum 233; saccharum Rugelii 233; saccharum sinuosum 234 Acrothecium flacatum 509 Actinomyces, morphology of 65, 147 African plants 184 After-ripening and germination 281 Agaricaceae of Michigan 279 Alternation of generations in Padina 278 Alway, F. J. 185 American gall insects 268 Angiosperm wood lacking vessels 279 Apogamy in Camptosorus 280; in Nephrodium 519 Apospory in Pteris sulcata 469 Appleman, C. O. 98, 99, 100, loi; work of 179 Apple trees, blister canker of 105 Aquilegia 184 Arber, Agnes, work of 273 Archangiopteris 84, 89; Henryi 90; Somai92; subintegra 90; tamdaoensis 91 Arthur, J. C. 180, work of 184 Artsch wager, E. F. 373 Atkinson, G. F. 266; biographical sketch of 366 Azotobacter, nitrogen fixation by 518 B . Bacterial tumor and wound callus 516 Bailey, C. H., work of 180 Bailey, I. W. 276, 438, 449; work of 279, 374 Baker, O. E., work of 103 Bars of Sanio 449 Bartlett, H. H., work of 100 Basidium, cytology of 376 \ Baur, E., work of 95 ' Belling, J., work of 514 Bennettitales 375 Betula commijcta 216; Eastwoodae 216 Bews, J. W., "Grasses and grasslands of South Africa" 370 Bexon, Dorothy, work of 280 Blackman, V. H., work of 278 Blister canker of apple trees 105 Boerker, R. H. D., "Our national forests" 370 Bonnier, G., work of 178 Bonns, W. W., work of 277 Borneo, ferns of 104 Brenchley, Winifred C, work of 182, 183, British lichens 268 Brotherton, W., work of 100 Brown, Elizabeth W., work of 183, 280 Bryophytes of Iceland 104 Buried weed seeds 515 Camptosorus, apogamy in 280 Cane sugar, action of neutral salts on acid inversion 98 Cape Breton, coniferous sand dune of 417; vegetation of 370 Carex, chromosomes in 448 Castanea alnifolia floridana 242 Catalase, respiration and vitamines 189 Cavers, F., work of 270 Celtis Douglasii 218; laevigata 221; laevigata anomala 225; laevigata brachyphylla 225; laevigata brevipes 226; laevigata microphylla 225; laevi- gata Smallii 223; laevigata texana 223; Lindheimerii 219; occidentalis 217; occidentalis canina 217; occidentalis crassifolia 217; pumila 227; pumila Deamii 228; pumila georgiana 227; reticulata 220; reticulata vestita 221 Cercospora persona ta 510 523 524 INDEX TO VOLUME LXVII [JUNE Chamberlain, C. J. 93, 270, 278, 375, 376, 448; "The living cycads" 511 Chlorophyll, inheritance of 95; loss of 446 Chondriosomes, in Carex 448; in plants 270 Chrysler, M. A. 512 Clayton, E. E. 483 Coccomyces clusiae 501 ; musae 502 Cockerell, T. D. A. 264 Combes, R., work of loi Compositae, a new genus of 280 Coniothyrium marisci 508 Contributors: Alway, F. J. 185; Apple- man, C. O. 98, 99, 100, loi; Arthur, J. C. 180; Artschwager, E. F. 373; Atkinson, G. F. 266; Bailey, I. W. 276, 438, 449; Chamberlain, C. J. 93, 270, 278, 375, 376, 448; Chrysler, M. A. 512; Clayton, E. E. 483; Cockerell, T. D. A. 264; Cook, M. T. 169, 268; Coulter, J. M. 103, 104, 183, 184, 269, 273, 274, 275, 278, 279, 280, 519; Coulter, M. C. 95, 100, 273, 513, 514; Crocker, W. 174, 179, 182, 184, 269, 272, 275, 277, 278, 279, 445, 448, 515, 516, 517, 518, 519, 520; DeVries, H. i; Dorety, Sister Helen Angela 251; Drechsler, C. 65, 147; Evans, A. T. 427; Faull, J. H. 369; Fink, B. 97, 268; Fuller, G. D. loi, 102, 104, 173, 181, 184, 369, 370, 374; Gurjar, A. M. 515; Haas, A. R. C. 347, 377; Halsted, B. D. 243; Harrington, G. T. 177; Harvey, L. H. 417; Harvey, R. B. 441; Hasselbring, H. 97, 102; Hayata, B. 84; MacDougal, D. T. 405; McDole, G. R. 185; McDougall, W. B. 258; Richards, H. M. 405; Rose, D. H. 105; Rose, R. C. 281; Sargent, C. S. 208; Schneider, C. 27, 309; ShuU, C. A. 276, 376; Sinnott, E. W. 374; Spoehr, H. A. 405; Steil, W. N. 469; Stevens, F. L. 173; Tehon, L. R. 501; Trelease, W. 173; Trumbull, R. S. 185; Walster, H. L. 171; Whetzel, H. H. 366; Wylie, R. B. 271 Cook, M. T. 169, 268 Copeland, E. B., work of 104 Correns, C. E., work of 95 Costa Rica, rusts of 184 Coulter, J. M. 103, 104, 183, 184, 269, 273, 274, 275, 278, 279, 280, 519 Coulter, M. C. 95, 100, 273, 513, 514 Crocker, W. 174, 179, 182, 184, 269, 277, 278, 279, 445, 448, 515, 516, 517, 518, 519, 520; work of 516 Curtis, O. F., work of 100 Cuttings, root growth in 100 Cycads, the living 512 Cytology of basidium 376 D Danilov, A. N., work of 97 Davenport, A., work of 277 DeVries, H. i Dicotyledons, seedling of 103 Disease resistance, breeding for 273 Dorety, Sister Helen Angela 251 Drechsler, C. 65, 147 Duggar, B. M., work of 102, 277 Dutcher, R. A., work of 179 East, E. M., work of 95, 514 Embryo, of Dioon spinulosum 251; sac of Pentstemon 427; sac and fertiliza- tion in Oenothera 275 Emerson, A. I., "Our trees, how to know them" 174 Emerson, R. A., work of 95 Enzyme, action 515; secretion 276 Epidermal coverings, importance of 441 Epiphytes, osmotic pressure of 520; sap concentration in 374 Eupatorium, tropical species of 280 Evans, A. T. 427 Evans, I. B. P., work of 273 Evergreen leaves, fat storage in 5 20 Farmer, J. B., work of 274 Farrow, E. P., work of 181, 182 Fat storage in evergreen leaves 520 Faull, J. H. 369 Felt, E. P., "American gall insects" 268 Fernald, M. L., work of loi Ferns of Borneo 104 Finch, V. C, work of 103 Fink, B. 97, 268 Fischer, H., work of 446 Fitzpatrick, H. M., work of 376 Fixative for paraffin sections 373 Flower development, conditions affecting 445 Forests, our national 369 Fraxinus americana subcoriacea 241 Fred, E. B., work of 277 Fucosan vacuoles 518 Fuller, G. D. 78, loi, 102, 104, 173, 181, 184, 369, 370, 374 Fumigation, hydrogen cyanide 483 Fungi, physiology of 102; Porto Rican 501 Gametophytes, inheritance in 514 Geotropism and phototropism 184 Germination and light 520 igig] INDEX TO VOLUME LXVII 525 Glacial plunge basin, vegetation of 184 Glucosides, physiological role of loi Grasses and grasslands of South Africa 370 Grassland and heath 181 Gurjar, A. M. 515; work of iSo H Haas, A. R. C. 347, 377; work of 182 Halsted, B. D. 243; biographical sketch of 169 Harper, R. A., work of 513 Harrington, G. T. 177 Harris, J. A., work of 374, 375, 520 Hartwell, B. L., work of 519 Harvey, L. H. 417 Har\'ey, R. B. 441 Hasselbring, H. 97, 102; work of 99 Hayata, B. 84 Hazen, T. E., work of 271 Heath and grassland 181 Heilborn, O., work of 448 Helminthosporium folliculatum 509 Hesselbo, A., work of 104 Hibbard, R. P., work of 174 Holden, H. S., work of 280 Holman, R. j\I., work of 376 Host and parasite, relation of 180 Hutchinson, H. B., work of 518 Hybrid perennial sunflowers 264 Hydrogen cyanide fumigation 483 Iceland, bryophytes of 104 Ikeno, S., work of 96 Illuminating gas, effect of 517 Inheritance, in gametophytes Pediastrum 513 Ishikawa, M., work of 275 514; m Java, orchids of 280 Journal of General Physiology 275 Juniperus utahensis megaloca'rpa 208 K Kauffman, C. H., work of 279 Kidd, F., work of 269 Kincer, J. B., work of 102 Klebs, G., work of 445 Knop's solution 448 Kraus, E. J., work of 446 Kraybill, H. R., work of 446 Kylin, H., work of 518 Leaves, aeration systems of 517 Lebert, M., work of 98 Lehmann, E., work of 520 Lichens, conidia of 97 Light and germination 520 Livingston, B. E., work of 175 Loeb, J., work of 103 Luminous moss 278 MacDaniels, L. H., work of 276 MacDougal, D. T. 405 McCall, A. G., work of 176 McDole, G. R. 185 McDougall, W. B. 258 Magnolia acuminata ludoviciana 222; virginiana 231; virginiana australis 231 Magnus, \V., work of 516 Maige, AL A., work of 178 Maize, evolution of 104 Mangin, L., work of 178 Maps of rainfall and crop plants 102 Martin, W. H., work of 176 Meliola asterinoides 503; bayamonensis 506; cestri 505; clusiae 50O; conferta 502; marcgraviae 506; perexigua 503 Metabolism, effect of different oxygen pressures on carbohydrate 99 Meyer, A., work of 44'6, 520 Michigan, Agaricaceae of 279 Monocotyledonous leaves, nature of 273 Monographs on experimental biology 103 Moore, S. LeM., work of 184 Mottier, D. M., work of 271 Murphy, P. A., work of 98 N Neger, F. W., work of 517 Nephrodium, apogamy in 519 Newfoundland, vegetation of loi Nichols, G. E., work of 371 Nicolas, G., work of 177, 178 Nitrogen, assimilating organisms 277; fixation by Azotobacter 518 North American, flora 269; trees 20S Nummularia discreta 105 o Oak stems, depressed segments of 438 Oenothera, embryo sac and fertilization in 275; rubrinervis, a half mutant i Orchids of Java 280 Osmotic pressure of epiphytes 520 Osterhout, W. J. V., work of 182 Ostrya virginiana 215; virginiana glandu- losa 216 526 INDEX TO VOLUME LXVII [JUNE Padina, alternation of generations in 278 Paine, S. G., work of 278, 279 Paraffin sections, fixative for 373 Park, J. B., work of 514 Parker, J. H., work of 181, 273 Parr, Rosalie, work of 272 Payson, E. B., work of 184 Pediastrum, inheritance in 513 Pember, F. R., work of 519 Pentstemon, embryo sac and embryo 427 Permeability 279 Persea pubescens 229 Pestalozzia lucumae 508 Pethybridge, G. H., work of 97 Petry, L. €., work of 184 Phaeosphaerella paspali 508 Phegopteris, regeneration in 183 Phloem, histology of 276 Photosynthesis 182 Phototropism 272; and geotropism 184 Phyllachora graminis 507; ischmaemi 507; quadraspora 507 Phytopathology, history of 173 Phytopathological journal, a new 103 Phytophthora, sex organs of 97 Picea glauca 208; glauca albertiana 208 Piemeisel, F. J., work of 181, 273 Platanus occidentalis 229; occidentalis attenuata 229; occidentalis glabrata 230 Pods, correlations in position of seeds 243 Poisons, organic plant 182 Pontederia, trimorphism of 271 Populus arizonica 210; arizonica Jonesii 211; Fremontii 213; Fremontii pubes- cens 213; Fremontii Thornberii 213; Fremontii Toumeyi 214; Macdougalii 212; Palmeri2ii; Parryi 214; texana 211; tremuloides vancouveriana 208 Porto Rican fungi 501 Prairies, minimum moisture content of subsoil related to hygroscopic coeffi- cient 185 Pritzel, E., work of 280 Protomarattia 84, 88; tonkinensis 88 Pteris sulcata, apospory in 469 R Rankin, H. W., "Manual of tree diseases" 369 Reed, G. M., work of 180 Rehfous, L., work of 274 Respiration, after death 347; and age 177; effect of anesthetics on 377; of stored wheat 180; vitamines and catalase 179 Reviews: Bews' "Grasses and grasslands of South Africa" 370; Boerker's "Our national forests" 370; Chamberlain's "The living cycads" 511; Emerson's "Our trees, how to know them" 174; Felt's "American gall insects" 268; Rankin's "Manual of tree diseases" 369; Russell's "Soil conditions and plant growth" 171; Seward's "Fossil plants" 93; Smith's "British lichens" 268; Weed's "Our trees, how to know them" 174; Whetzel's "History of phytopathology" 173 Rhodosporeae, relations within the 266 Richards, H. M. 405 Richards, P. E., work of 176 Rims of Sanio 449 Robbins, W. J., work of 276 Robinson, B. L., work of 280 Roots, orientation of 376 Rose, D. H. 105 Rose, R. C. 281 Rubus, after-ripening and germination 281 Russell, E. J., "Soil conditions and plant growth" 171 Rusts of Costa Rica 184 Rydberg, P. A., work of 269 Salix, anamesa 62; arctica 56; arctophila 57; argophylla 325; cascadensis 55; chamissonis 63; chlorolepis 60; cordi- folia 62; Dodgeana 54; exigua 328; exigua luteosericea 334; exigua neva- densis 331; exigua tenerrima 335; fluviatilis 322; glacialis 63; glauca acutifolia 60; herbacea 52; hudson- ensis 57; leiolepis 46; longifolia 340; longifolia Wheeleri 342; melanopsis 337; melanopsis Bolanderiana 338; niphoclada 59; nivalis 47; nivalis saximontana 47; obcordata 64; Par- ishiana 323; Peasei 51; phlebophyUa 53; polaris 49; reticulata 44; rotundi- folia 53; sessilifolia 316; stolonifera 57; taxifolia 316; Uva-ursi 50; venusta 64; vestita 45 Sambucus, after-ripening and germina- tion 281 Sand dune, coniferous 417 Sap concentration in epiphytes 374 Sargent, C. S. 208 Saunders, L. M., work of 279 Schmitz, H., work of 102 Schneider, C. 27, 309 Seedling, anatomy 280; of dicotyledons 103; of Dioon spinulosum 251 I9I9] INDEX TO VOLUME LXVII 527 Seeds, after-ripening and germination 281; correlations in position in pods 243; secondary dormancy in 269 Selaginella 183 Severy, J. \V., work of 102 Seward, A. C, "Fossil plants" 93 Shive, J. W., work of 176 Shull, C. A. 276, 376 Sinnott, E. W. 374; work of 103 Smith, Annie L., "British lichens" 268 Smith, J. J., work of 280 Soil, acidity of 519; conditions and plant growth 171; physiological balance in 174 South Africa, grasses and grasslands of 370 Spoehr, H. A. 405 Stakman, E. C, work of 181, 273 Stallard, H., work of 372 Standley, P. C, work of 269 Statolith starch 520 Steil, W. N. 469; work of 519 Stevens, F. L. 173 Stigmatea guettardae 508 Stomata 274 Stopes, Marie C, work of 375 Stropharia epimyces, development of 258 Subsoil of prairies 185 Succulence, basis of 405 Sunflowers, hybrid perennial 264 Szombathy, K., work of 373 Transeau, E. N., work of 514 Transpiration 277 Trelease, W., "Winter botany" 173 Trichostroma axonopi 510 Trimorphism of Pontederia 271 Trumbull, R. S. 185 Tupper, W. W., work of 374 Turgor movements 278 Van Ameijden, U. P., work of 184 Van Eseltine, G. P., work of 183 Van Laer, H., work of 515 Variation, analysis of quantitative 100 Vitamines, catalase, and respiration 1 79 w Walster, H. L. 171 Water conduction in trees and shrubs 275 Weatherwax, P., work of 104 Weed, C. M., "Our trees, how to know them" 174 Wehmer, C, work of 517 West, C, work of 269 Whetzel, H. H. 366; "History of phyto- pathology" 173 Willows, American 27, 309 Wolfe, J. J., work of 278 Wound callus and bacterial tumor 516 Wylie, R. B. 271 Tehon, L. R. 501 Thompson, W. P., work of 279 Tilia, after-ripening and germination 281 Toda, Viscount Y., work of 279 Toole, E. H., work of 448 Tottingham, W. R., work of 175, 448 X Xylem, size variation in 374 Zollikofer, Clara, work of 520 'a A NEW VOLUME OF THE UNIVERSITY OF CHICAGO SCIENCE SERIES The Living Cycads By Charles Joseph Chamberlaix Professor of Botany in the UniTxrsity of Chicago This studv includes investigations extending over fifteen years, that necessitated trips to Cuba, New Mexico, Australia, and Africa. The author studied all the genera and many of the species in the field, and preserved much material for later study in the laboratory. In the first part of the book is given an account of the distribution, general appearance, and field conditions of the cycads, together with some of the experiences of an investigation involving much travel in distant and ^'aried countries. In the second part is presented the life-history of the group based largely on the author's own observations in the field and laborator>\ Part III is devoted to the evolution and phylogeny of the cycads, the opportunity for such a study being exceptionally favor- able because the ancestry can be traced back through geological periods, and because the extinct predecessors of the cycads are the best known of the fossil plants. .v?>+j-2 pages, doth; $1.50, postage extra TO BE PUBLISHED IN MAY Problems of Fertilization By Frank R. Lillie Professor of Embryology and Chairman of //«• Depj^'"""' "^ 7r,nln?\- in the University of Chicago As the problem of fertilization deals with the union of the male and the female sex elements, which forms the beginning of each indi\-iduars life-history, it has important bearings upon the problems of heredity. It also has close relations to several other fundamental biological problems. Cloth; $1.75, postage extra OTHER PUBLISHED VOLUMES IN THE SERIES The Ecolution of Sex in Plants. By John Merle A Chemical Sign of Life. By Shiro Tashiro. Coulter, x-f- 140 pages, umo, cloth; $1.00, post- x-i-142 pages, small i2mo, cloth; $1.00, postage age extra (weight 14 oz.)- "tra (weight 14 oz.). r J. .J I. . ^ T. .- »» ^ The Biology of Twins. By Hokatio Hackett Indmdualitym Organisms. By Ch.xrles Man^jing Xewsl^-. x-|-iS6 pages, small i2mo. cloth; Child, x + 21 > paps, smaJl i2mo, cloth; $1.25, 5, ,. posuge extra (weight i lb.). ' postage extra (weight i lb. 6 oz.). j,^ Poisoning. By Edwin O.^S ' Jord.^n. The Origin of the Earth. By Thomas C. viii -j-i 16 pages, small i2mo, cloth; Si. 00, postage Ch.\mberlin. iii-l-272 pages, small i2mo, doth; extra (weight 11 oz.\ 81.50, postage extra (weight i lb. 6 oz.). ff,g Electron: Its Isolation and Measurement, and i -iteCollineationCroups. ByH.VfsF.BucHFELDT. the Determination of Some of Itj Properties. xii+194 pages, smaU i2mo, cloth; $1.50. postage By Robert Antjrews Millik.a^-. xii+205 pages extTa (weight 1 lb.). • ' = ^ small i2mo, cloth; $1.50, postage extra (weight I lb. 3 oz.). VOLUMES PLANNED OR IN PREPARATION Black Body Radiation. By Ch.\rle5 E. Menden-haxl. Mechanics of Delayed Germination in Seeds. By W. Crocker. The Rigidity of the Earth and of Materials. By A. A. Michelson. Linear Integral Equations in General Analysis. By Ellaktv Hastings Moore. THE UNIVERSITY OF CHICAGO PRESS CHICAGO ...--- ILLINOIS Men of Science The signs and symbols used in your everyday work are provided for in special keyboards prepared for the Home Typewriter, the Self-Starting Remi.,t.. Tell us your needs and we will sub- mit a keyboard layout for your work. No obligation, of course, on your part. Remington Typewriter Company, Inc. 374 Broadway, New York City FINE INKS 45? ADHESIVES For those who KNOW DrawInK Inks eternal Writing; Inic Bngrossing Ink Taurine Mucilage Photo Mounter Paste Drawing Board Paste Liquid Paste Ofnce Paste Vesetable Olue. Etc. Are the Finest and Best Inks and Adhesives Higgins' Emancipate jrourself from the use of corrosive and ill-smellin£ inks and adhesives and adopt the Hlnins Inks and Adhesives. They will bea revelation to ^ou, they are so sweet, clean, well put up, and withal so efficient. At Dealers Qenerally CHAS. M. HIQQINS & CO., Mfrs. Branches: Chicago, London 371 Ninth street Brooklyn, N.Y. FS2 Microscope -Price $40.75 ^auscfi'lomb Microscopes are again in the market — backed by the prestige which attaches to successful war service in any significant capacity. Our release from this service is of course gradual. As the demands made upon us by the government are lessened from month to month, however, we are enabled to increase deliveries to our other patrons. Write for complete catalog of our revised microscope line and inform us regarding your requirements. Bausch ^ Ipmb Optical (5. 41 2 ST. MUL STRBtT ROCHBSTBR. N.Y. New York Washington Chicago San Francisco Leading American Makers of Photographic Lenses, Microscopes, Projection Apparatus {Baloptlcona) , Ophthalmic Lenses and Instruments, Photomicrographic Apparatus, Range Finders and Gun Sights for A rmy and Navy, Searchlight Reflectors, Stereo-Prism Birwculars, Magnifiers and Other High-Grade Optical Products. 97r mmmmmmmmm